Use of Automatically Defined Functions and Architecture- Altering Operations in Automated Circuit Synthesis with Genetic Programming
|
|
- Cori Sutton
- 6 years ago
- Views:
Transcription
1 Use of Automatically Defined Functions and Architecture- Altering Operations in Automated Circuit Synthesis with Genetic Programming John R. Koza Computer Science Dept. 258 Gates Building Stanford University Stanford, California David Andre Visiting Scholar Computer Science Dept. Stanford University Stanford, California u Forrest H Bennett III Visiting Scholar Computer Science Dept. Stanford University Stanford, California fhb3@slip.net Martin A. Keane Econometrics Inc West Grover Chicago, Illinois makeane@ix.netcom.com ABSTRACT This paper demonstrates the usefulness of automatically defined functions and architecture-altering operations in designing analog electrical circuits using genetic programming. A design for a lowpass filter is genetically evolved in which an automatically defined function is profitably reused in the 100% compliant circuit. The symmetric reuse of an evolved substructure directly enhances the performance of the circuit. Genetic programming rediscovered the classical ladder topology used in Butterworth and Chebychev filters as well as the more complex topology used in Cauer (elliptic) filters. A design for a double-passband filter is genetically evolved in which the architecture-altering operations discover a suitable program architecture dynamically during the run. Two automatically defined functions are profitably reused in the genetically evolved 100% complaint circuit. 1. Introduction Human-designed electrical circuits are replete with instances of the hierarchical reuse of previously designed subcircuits. Indeed, design of a large electronic circuit would be impractical if the human designer had to start from first principles and rethink the design of each subcircuit each occasion time it is needed. The previous paper in this volume "Automated WYWIWYG Design of Both the Topology and Component Values of Electrical Circuits Using Genetic Programming" demonstrated that complex structures such as electrical circuits can be evolved by means of natural selection (Koza, Bennett, Andre, and Keane 1996). The regularity, symmetry, and reuse of substructures found in human-designed electrical circuits suggests that automatically defined functions might be useful in the problem domain of circuit synthesis. Accordingly, this paper explores the usefulness of automatically defined functions and architecture-altering operations for the problem of automated circuit synthesis. Section 2 provides background on automatically defined functions and the architecture-altering operations. Section 3 describes the design of a lowpass filter in which an automatically defined function is profitably reused in the genetically evolved electrical circuit. Section 4 describes the design of a double-passband filter in which the architecture-altering operations successfully evolve the architecture of the solution and in which automatically defined functions are profitably reused in the genetically evolved solution. 2. Background Genetic programming is an extension of John Holland's genetic algorithm (1975) in which the population consists of computer programs of varying sizes and shapes (Koza 1992, Koza and Rice 1992). The book Genetic Programming II: Automatic Discovery of Reusable Programs (Koza 1994a, 1994b) describes how to evolve multi-part programs consisting of a main program and one or more reusable, parameterized, hierarchically-called subprograms. This book is based on the premise that no approach to automated programming is likely to be successful on non-trivial problems unless it provides some hierarchical mechanism to exploit, by reuse and parameterization, the regularities, symmetries, homogeneities, similarities, patterns, and modularities inherent in problem environments. An automatically defined function (ADF) is a function (subroutine, subprogram, DEFUN, procedure, or module) that is dynamically evolved during a run of genetic programming and that may be called by a calling program (or subprogram) that is concurrently being evolved. When automatically defined functions are being used, a program
2 in the population consists of a hierarchy of one (or more) reusable function-defining branches (i.e., automatically defined functions) along with a main result-producing branch. Typically, the automatically defined functions possess one or more dummy arguments (formal parameters) and are reused with different instantiations of these dummy arguments. During a run, genetic programming creates different subprograms in the function-defining branches of the overall program, different main programs in the resultproducing branch, different instantiations of the dummy arguments of the automatically defined functions (functiondefining branches), and different hierarchical references between the branches. When automatically defined functions are used, it is necessary to determine the architecture of the yet-to-beevolved programs. The specification of the architecture consists of (a) the number of function-defining branches in the overall program, (b) the number of arguments (if any) possessed by each function-defining branch, and (c) if there is more than one function-defining branch, the nature of the hierarchical references (if any) allowed between them. The architecture-altering operations (Koza 1994c, 1995a) provide one way to automate this architectural choice so that it can be made dynamically during a run of genetic programming. The six architecture-altering operations are motivated by the naturally occurring mechanisms of gene duplication and gene deletion in chromosome strings as described in Susumu Ohno's book Evolution by Gene Duplication (1970). In that book, Ohno advanced the thesis that the creation of new proteins (and hence new structures and new behaviors in living things) begins with a gene duplication. The potential usefulness of automatically defined functions and architecture-altering operations have been previously demonstrated by both proof-of-principle ("toy") problems and by non-trivial problems. For example, on the transmembrane segment identification problem (Koza and Andre 1996), the results produced by using genetic programming with automatically defined functions and the architecture-altering operations were competitive with the human-written algorithm for that problem. 3. Evolving a Lowpass Filter Using Automatically Defined Functions A filter is a one-input, one-output circuit that receives a signal as its input and passes the frequency components of the incoming signal that lie in a certain specified range (the passband) while stopping the frequency components of the signal that lie in other frequency ranges (the stopband). Consider a circuit synthesis problem in which the design goal is to design a lowpass filter using inductors and capacitors with a passband below 1,000 Hertz and a stopband above 2,000 Hz. The voltages in the filter's passband are to lie in the narrow range between 970 millivolts and 1 volt (i.e., the permissible passband ripple is 30 millivolts) and the voltages in the stopband are to lie in the narrow range between 0 volts and 1 millivolt. The circuit is to be driven from an alternating-current input signal with a 2 volt amplitude with a source resistance of 1,000 Ohms and a load resistance of 1,000 Ohms Preparatory Steps This paper assumes that the reader is familiar with the use of genetic programming in electrical circuit design described in the previous paper in this volume (Koza, Bennett, Andre, and Keane 1996) Embryonic Circuit and Program Architecture A one-input, one-output embryonic circuit with two writing heads is suitable for this problem. Thus, there are two result-producing branches in each program tree in the population. In this section of this paper, we pre-specified that the common architecture of each individual in the population would consist of four zero-argument automatically defined functions and that there would be no hierarchical references among the automatically defined functions. Consequently, the architecture of every program tree in the population throughout the entire run consists of a total of six branches joined by a LIST function, as shown in figure 1. ADF0 ADF1 LIST ADF2 ADF3 RPB0 RPB1 Figure 1 Program architecture Function and Terminal Sets The function set, Fccs, for each construction-continuing subtree is Fccs = {C, L, SERIES, PSS, FLIP, NOP, THGND, THVIA0, THVIA1, THVIA2, THVIA3, THVIA4, THVIA5, THVIA6, THVIA7, ADF0, ADF1, ADF2, ADF3}. The C, L, SERIES, PSS, FLIP, and NOP functions are defined in Koza, Bennett, Andre, and Keane (1996). The eight three-argument "via" functions (called THVIA0,..., THVIA7) and the three-argument "ground" function (called THGND) enable distant parts of a circuit to be connected together. These functions are similar to the VIA and the GND functions, except that they each create three (instead of two) modifiable wires. ADF0, ADF1, ADF2, and ADF3 are automatically defined functions. The terminal set, Tccs, for each construction-continuing subtree is Tccs = {END, CUT}. The zero-argument CUT function causes the highlighted component to be removed from the circuit. The CUT function also causes the writing head to be lost. The function set, Faps, for each arithmetic-performing subtree is Faps = {+, -}.
3 The terminal set, Taps, for each arithmetic-performing subtree is Taps = { }, where represents floating-point random constants between and Fitness Measure The evaluation of fitness for each individual circuitconstructing program tree in the population begins with its execution. This execution applies the functions in the program tree to the embryonic circuit thereby developing the embryonic circuit into a fully developed circuit. A netlist describing the circuit is then created. Each circuit is then simulated to determine its behavior using a version of the SPICE3 (Quarles et al. 1994) simulator that we modified to run as a submodule within the genetic programming system. The SPICE simulator is requested to perform an AC small signal analysis and to report the circuit's behavior for each of 101 frequency values chosen over five decades of frequency (from 1 Hz to 100,000 Hz). Each decade is divided into 20 parts (on a logarithmic scale). Fitness is measured in terms of the sum, over these 101 fitness cases, of the absolute weighted deviation between the actual value of the voltage in the frequency domain that is produced by the circuit at its output probe point and the desired voltage at that point. The smaller the fitness, the better (with a fitness of zero being best). The fitness measure is constructed so that it does not penalize ideal values; it slightly penalizes acceptable deviations; and it heavily penalizes unacceptable deviations. Specifically, the procedure for each of the 61 points in the three-decade interval from 1 Hz to 1,000 Hz (i.e., the desired passband) is as follows: If the voltage at the output probe point equals the ideal value of 1.0 volts in this interval, the deviation is 0.0. If the voltage is between 970 millivolts and 1,000 millivolts, the absolute value of the deviation from 1,000 millivolts is weighted by a factor of 1.0. If the voltage is less than 970 millivolts, the absolute value of the deviation from 1,000 millivolts is weighted by a factor of This arrangement reflects the fact that the ideal voltage in the passband is 1.0 volt, the fact that a 30 millivolt shortfall is acceptable, and the fact that a voltage below 970 millivolts in the passband is not acceptable. The procedure for each of the 35 points between 2,000 Hz and 100,000 Hz (i.e., the desired stopband) is as follows: If the voltage is between 0 millivolts (the ideal value) and 1 millivolt, the absolute value of the deviation from 0 millivolts is weighted by a factor of 1.0. If the voltage is more than 1 millvolt, the absolute value of the deviation from 0 millivolts is weighted by a factor of This arrangement reflects the fact that the ideal voltage in the stopband is 0 millivolts, the fact that a 1 millivolt ripple above 0 millivolts is acceptable, and the fact that a voltage above 1 millivolt in the stopband is not acceptable. The deviation is deemed to be zero for each of the 5 points in the interval between 1,000 Hz and 2,000 Hz (i.e., the "don't care" band of frequencies). Hits are defined as the number (from 5 to 101) of fitness cases for which the voltage is acceptable or ideal or which lie in the "don't care" band Control Parameters The population size, M, is 640,000. The crossover percentage is 89% (producing 569,600 offspring), the reproduction percentage was 10%, and the mutation percentage was 1%. A maximum size of 300 points was established for each of the branches in each overall program. The other minor parameters were the default values in Koza 1994a (appendix D). The problem was run on a parallel computer system with a migration rate of B = 2% as described in Koza and Andre Results for the Lowpass Filter Initial Random Generation The best circuit from generation 0 has a fitness of 58.6 and scores 52 hits (out of 101). Figure 3 shows that this best-ofgeneration circuit from generation 0 consists of a single inductor and a single capacitor. The topological arrangement of these two components is that of the first rung of a classical ladder. Figure 9 shows the frequency domain behavior of this best-of-generation circuit from generation Emergence of Reuse A Two-Rung Ladder The best circuit from generation 9 (figure 4) has the tworung ladder topology. The ladder topology for a lowpass filter consists of repeated instances of various series inductors (so named because they run "in series" horizontally across the top of the figure) and repeated instances of various vertical shunt capacitors. The classical Butterworth or Chebychev filters are based on the ladder topology (Van Valkenburg 1982). Automatically defined function ADF0 supplies a group of three inductors (equivalent to one 154,400 µh inductor). Figure 2 shows this twice-called two-ported substructure developed by ADF0. Figure 2 Twice-called two-ported automatically defined function ADF0 from generation 9. Figure 10 shows the behavior in the frequency domain of the best-of-generation circuit from generation 9. Figures were made with 200 points per decade for greater detail A Thrice-Called Substructure Further Enhances Performance The best circuit from generation 16 (figure 5) has a fitness of 4.1 and scores 90 hits. This circuit has a three-rung ladder topology. When electrical engineers design
4 Butterworth or Chebychev filters, additional rungs on the ladder (in conjunction with properly chosen numerical values for the components) generally improve the level of performance of the filter (at the expense of additional power consumption, space, and cost). Figure 11 shows the behavior in the frequency domain of the best circuit from generation 16. ADF0 is used three times in creating this best circuit of generation 16. Figure 15 shows this thricecalled automatically defined function. Figure 15 Thrice-called two-ported automatically defined function ADF0 from generation 16.
5 Figure 3 The best circuit from generation 0 is a onerung ladder. Figure 5 The best circuit of generation 16 is a threerung ladder. Figure 4 The best circuit from generation 9 is a tworung ladder. Figure 6 The best circuit of generation 20 is a four-rung ladder. Figure 7 The 100% compliant best circuit of generation 31 has the Cauer (elliptic) topology. Figure 8 The 100% compliant best-of-run circuit from generation 35.
6 3.2.4 A Quadruply-Called Substructure Further Enhances Performance The best circuit from generation 20 (figure 6) has a fitness of 2.8 and scores 96 hits. It is a four-rung ladder. ADF0 (figure 16) is used four times. Figure 9 Frequency domain behavior of best circuit of generation 0. Figure 10 Frequency domain behavior of best circuit and generation Figure 11 Frequency domain behavior of best circuit from generation 16. Figure 16 Quadruply-called two-ported automatically defined function ADF0 from generation 20. Figure 12 shows the behavior in the frequency domain of the best circuit from generation 20. The enhanced performance of this circuit is the consequence of the additional repeated calls to ADF0. Note that in figures 10 12, there is ripple in the passband, but none in the stopband (i.e., these circuits exhibit a Chebychev-like response) Emergence of the Elliptic Topology Further Enhances Performance The best circuit in generation 31 (figure 7) has a fitness of and scores 101 hits (i.e., is 100% compliant). This individual has 41 points in its RPB0 and 7 points in its RPB1; it has 52, 297, 22, and 2 points in ADF0, ADF1, ADF2, and ADF3, respectively. Figure 13 shows its behavior in the frequency domain. Automatically defined function ADF0 is interesting because it constructs a three-ported substructure. This ADF0 (figure 17) is used five times in the best circuit from generation 31. Figure 12 Frequency domain behavior of best circuit from generation 20. Figure 13 Frequency domain behavior of best circuit from generation 31. Figure 14 Frequency domain behavior of best-of-run circuit from generation 35. Figure 17 Quintuply-called three-ported automatically defined function ADF0 from generation 31. This genetically evolved 100% compliant circuit is especially interesting because it has the topology of an elliptic (Cauer) filter. After all of the pairs and triplets of inductors are consolidated, it can be seen that the circuit has the equivalent of six inductors horizontally across the top of the circuit and five vertical shunts. Each shunt consists of an inductor and a capacitor (e.g., L34 and C18 appear in the first shunt). This is the topology of the elliptic invented by Cauer.
7 ADF0 8 V DEFUN NOP SERIES L LIST PSS V5 10 C THVIA3 FLIP 11 END END END PSS PSS END FLIP SERIES END NOP END PSS END FLIP END END END FLIP ADF END L END PSS END END END FLIP ADF0 ADF V2 L END END END C ADF V3 END V4 END Figure 18 Best-of-run individual from generation 35. The Cauer filter was a significant advance (both theoretically and commercially) over the Butterworth and Chebychev filters. For example, for one illustrative set of specifications, a fifth-order elliptic filter could equal the performance of 17th order Butterworth filter or an eighth order Chebychev filter. The eighth order Chebychev filter has one more component than the fifth order elliptic filter. As Van Valkenburg (1982, page 379) explains: "Cauer first used his new theory in solving a filter problem for the German telephone industry. His new design achieved specifications with one less inductor than had ever been done before. The world first learned of the Cauer method not through scholarly publication but through a patent disclosure, which eventually reached the Bell Laboratories. Legend has it that the entire Mathematics Department of Bell Laboratories spent the next two weeks at the New York Public library studying elliptic functions. Cauer had studied mathematics under Hilbert at Goettingen, and so elliptic functions and their applications were familiar to him." Thus, this run of genetic programming illustrates the rediscovery of the ladder topology used in Butterworth and Chebychev filters as well as the more complex topology used in elliptic (Cauer) filters An Even Better Circuit The best circuit in generation 35 (figure 8) has a fitness of (i.e., about an order of magnitude better than that of the best-of-generation individual from generation 31) and is also 100% compliant (i.e., scores 101 hits). Only ADF0 is referenced by the result-producing branches. Figure 14 shows the "brick wall" behavior in the frequency domain of this best-of-run best circuit from generation 35. Notice this circuit's two-fold symmetry involving the repetition of a total of four modular substructures. The symmetry of the best-of-run circuit from generation 35 is a consequence of its quadruply-called three-ported automatically defined function, ADF0 (figure 19). Two inductors (L52 and L34) and one capacitor (C51) form a triangle. There is an additional induction element (specifically, a series composition of three inductors, L43, L42, and L23) branching away from the triangle at the node where inductors L52 and L34 meet Figure 19 Quadruply-called three-ported automatically defined function ADF0 of best-of-run circuit from generation 35. Figure 18 presents the best-of-run individual from generation 35 as a rooted, point-labeled tree with ordered branches. Note that ADF0 is invoked at points 34, 43, 44, and 51 of the figure. Since ADF0 is the only functiondefining branch that is actually referenced by either resultproducing branch, the three unreferenced function-defining branches (ADF1, ADF2, and ADF3) are represented by the filled circles labeled 3, 4, and 5. The complete arithmeticperforming subtrees are not shown, but, instead, are abbreviated as V1 through V5. 4. Evolving a Double-Bandpass Filter Using Architecture-Altering Operations In the previous section, the user pre-specified the number of automatically defined functions as one of the preparatory steps for the run of genetic programming. In this section, the architecture-altering operations will be used to evolve the program architecture dynamically during the run. The goal is to design a double-bandpass filter using inductors and capacitors as primitive components. Specifically, the goal is to design a double-bandpass filter whose first passband starts at 1,000 Hz and ends at 2,000 Hz, whose second passband starts at 1,000,000 Hz and ends at 2,000,000 Hz, and whose passband and stopband ripple are that of an elliptic filter (as detailed below). The regularity inherent in a filter with two passbands suggests that automatically defined functions may potentially be useful for this design problem. 7
8 4.1. Preparatory Steps Embryonic Circuit and Program Architecture A one-input, one-output embryonic circuit with one writing head is suitable for this problem. Since the embroyonic circuit has one writing head, each program in the initial population of programs has one result-producing branch. Since we did not use the operation of branch creation on this problem, each program tree in the initial random population at generation 0 also has one one-argument automatically defined function. Thus, each program tree at generation 0 starts with a uniform architecture consisting of a LIST function joining ADF0 and the result-producing branch, RPB. The number of automatically defined functions in the eventual solution will be determined dynamically, during the run, using the architecture-altering operations of branch duplication and branch deletion. Thus, starting with generation 1 of each run, the population becomes architecturally diverse Function and Terminal Sets For any branch, the function set, Fccs, for each construction-continuing subtree is Fccs = {ADF0, C, L, SERIES, PSS, FLIP, NOP, THGND, THVIA0, THVIA1, THVIA2, THVIA3, THVIA4, THVIA5, THVIA6, THVIA7}. Here ADF0 is the first automatically defined function. The terminal set, Tccs-rpb, for each constructioncontinuing subtree in the result-producing branch is Tccs-rpb = {END, CUT}. The terminal set, Tccs-adf, for each constructioncontinuing subtree in any function-defining branch is Tccs-adf = {ARG0, END, CUT}. Since architecture-altering operations can create automatically defined functions, the set of potential new functions, Fpotential, is Fpotential = {ADF1, ADF2, ADF3, ADF4}. The function set, Faps, for an arithmetic-performing subtree in any branch and the terminal set, Taps, for an arithmetic-performing subtree in any branch are the same as in the previous section of this paper Fitness Measure The SPICE simulator is requested to perform an AC small signal analysis and to report the circuit's behavior for each of 176 frequency values chosen over seven decades of frequency (from 10 Hz to 100,000,000 Hertz). Each decade is divided into 25 parts (using a logarithmic scale). Fitness is measured in terms of the sum, over these 176 fitness cases, of the absolute weighted deviation between the actual value of the voltage that is produced by the circuit at the output probe point and the target value for voltage. The frequency range is divided into two passbands, three stopbands, and four "don't care" regions. The procedure for each of the 8 points in each of the two passbands (a total of 16 points) is as follows: If the voltage is between 960 millivolts and 1,000 millivolts (the ideal value), the absolute value of the deviation from 1,000 millivolts is weighted by a factor of 1.0. If the voltage is less than 960 millivolts, the absolute value of the deviation from 1,000 millivolts is weighted by a factor of The procedure for each of the 120 points in the three stopbands is as follows: If the voltage is between 0 millivolts (the ideal value) and 40 millivolts, the absolute value of the deviation from 0 millivolts is weighted by a factor of 1.0. If the voltage is more than 40 millivolts, the absolute value of the deviation from 0 millivolts is weighted by a factor of There are 10 "don't care" points just before each passband and 10 "don't care" points just after each passband. The deviation is deemed to be zero for each of the 40 points in these four "don't care" bands. Hits are defined as the number (between 40 and 176) of fitness cases for which the voltage is acceptable or ideal or that lie in one of the "don't care" bands Parameters The control parameters were the same as in the previous section, except for the following: (1) The architecture-altering operations are intended to be used sparingly on each generation. The percentage of operations on each generation after generation 5 was 87.5% one-offspring crossovers; 10% reproductions; 1% mutations; 1% branch duplications; 0% argument duplications; 0.5% branch deletions; 0% argument deletions; 0% branch creations; and 0% argument creations. Since we do not want to waste large amounts of computer time in early generations where only a few programs have any automatically functions, the percentage of operations on each generation before generation 6 was 82.0% oneoffspring crossovers; 11% reproductions; 1% mutations; 5.0% branch duplications; 0% argument duplications; 1% branch deletions; 0% argument deletions; 0.0% branch creations; and 0% argument creations. (2) The maximum number of automatically defined functions is five. (3) The number of arguments for each automatically defined function is one. (4) In randomly choosing functions during the creation of the initial random population, the C, L, SERIES, PSS, FLIP, NOP, THGND, and ADF0 functions were each assigned a relative weight of 8, while each of the eight THVIA functions were assigned a relative weight of Results for Double-Bandpass Filter Using Architecture-Altering Operations This run illustrates the progressive addition of functiondefining branches and references to them. The best circuit from generation 0 has a fitness of and scores 54 hits; however, the result-producing branch of the program tree does not reference its 45-point automatically defined function ADF0. In generation 2, the first best-of-generation individual with a referenced automatically defined function appears (with a fitness of and 58 hits).
9 The first best-of-generation individual with two automatically defined functions appears in generation 22 (with a fitness of and 110 hits); however, there are no references to either automatically defined function. In generation 29, the first best-of-generation individual with two referenced automatically defined functions appears (with a fitness of and 128 hits). The first circuit scoring a full 176 hits is produced in generation 82 (with a fitness of 0.731). A 100% complaint best-of-run individual emerged at generation 89 with four automatically defined functions. This individual has a fitness of and scores 176 hits. The result-producing branch has 296 points and the function-defining branches have 82, 71, 64, and 82 points, respectively. ADF0 is called four times; ADF1 is called once; ADF2 is ignored; and ADF3 is called twice. Figure 23 shows the best-of-run circuit from generation 89. Note that L79, C78, and L66 are superfluous. Boxes are used to indicate the parts of the overall circuit that are developed by the automatically defined functions. The letters, A, B, C, and D indicate the interface points at which the substructure specified by each automatically defined function connects to the remainder of the circuit. Figure 20 shows the three-ported substructure developed from ADF0 at each of the four places in the resultproducing branch where ADF0 is invoked. The figure shows ADF0's three ports (A, B, and C) indicating the interface points at which this substructure developed from ADF0 connects to the remainder of the circuit. Figure 20 Three-ported quadruply-called ADF0 from generation 89. Figure 21 shows the three-ported substructure developed from ADF1. ADF1 is invoked once. Figure 21 Three-ported ADF1 from generation 89. Figure 22 shows the four-ported substructure developed from ADF3 at each of the two places in the result-producing branch where ADF3 is invoked. The figure shows ADF3's four ports (A, B, C, and D) indicating the interface points at which this substructure developed from ADF3 connects to the remainder of the circuit. The insertion process for automatically defined functions (even those with no arguments) is contextsensitive. For example, in figure 23, the substructures developed by the two invocations of ADF3 are different. Specifically, one invocation employs ports A, B, and C (of figure 22) while the other employs ports A, B, and D. This difference was caused by parts of the overall program tree outside ADF3. Figure 22 Four-ported twice-called ADF3 from generation 89. Figure 24 shows the behavior in the frequency domain of the best-of-run circuit from generation 89. All 136 fitness cases that are not part of the "don't care" bands satisfy the design goals of this problem. Notice that the two small non-monotonic spikes lie in the "don't care" bands. 5. Conclusions This paper demonstrates the usefulness of automatically defined functions and the architecture-altering operations in runs of genetic programming in which electrical circuits are being designed. Acknowledgments Simon Handley made helpful comments on drafts of this paper. Bibliography Andre, David and Koza, John R Parallel genetic programming: A scalable implementation using the transputer architecture. In Angeline, Peter J. and Kinnear, Kenneth E. Jr. (editors). Advances in Genetic Programming 2. Cambridge, MA: MIT Press. Holland, John H Adaptation in Natural and Artificial System. Ann Arbor, MI: University of Michigan Press. Koza, John R Genetic Programming: On the Programming of Computers by Means of Natural Selection. Cambridge, MA: MIT Press. Koza, John R. 1994a. Genetic Programming II: Automatic Discovery of Reusable Programs. Cambridge, MA: MIT Press. Koza, John R. 1994b. Genetic Programming II Videotape: The Next Generation. Cambridge, MA: MIT Press. Koza, John R. 1994c. Architecture-Altering Operations for Evolving the Architecture of a Multi-Part Program in Genetic Programming. Stanford University Computer Science Department technical report STAN-CS-TR October 21, Koza, John R Evolving the architecture of a multipart program in genetic programming using architecture-
10 altering operations. In McDonnell, John R., Reynolds, Robert G., and Fogel, David B. (editors) Evolutionary Programming IV: Proceedings of the Fourth Annual Conference on Evolutionary Programming. Cambridge, MA: MIT Press. Pages Koza, John R. and Andre, David Classifying protein segments as transmembrane domains using architecture-altering operations in genetic programming. In Angeline, Peter J. and Kinnear, Kenneth E. Jr. (editors) Advances in Genetic Programming 2. Cambridge, MA: MIT Press. In Press. Koza, John R., Bennett III, Forrest H, Andre, David, and Keane, Martin A Automated WYWIWYG design of both the topology and component values of analog electrical circuits using genetic programming. In Koza, John R., Goldberg, David E., Fogel, David B., and Riolo, Rick L. (editors). Genetic Programming 1996: Proceedings of the First Annual Conference, July 28-31, 1996, Stanford University. Cambridge, MA: MIT Press. In this volume. Koza, John R., and Rice, James P Genetic Programming: The Movie. Cambridge, MA: MIT Press. Ohno, Susumu Evolution by Gene Duplication. New York: Springer-Verlag. Quarles, Thomas, Newton, A. R., Pederson, D. O., and Sangiovanni-Vincentelli, A SPICE 3 Version 3F5 User's Manual. Department of Electrical Engineering and Computer Science, University of California, Berkeley, California. March Van Valkenburg, M. E Analog Filter Design. Fort Worth, TX: Harcourt Brace Jovanovich. Figure 23 Best-of-run circuit from generation 89. Figure 24 Frequency domain behavior of the best-of-run circuit from generation 89.
Reuse, Parameterized Reuse, and Hierarchical Reuse of Substructures in Evolving Electrical Circuits Using Genetic Programming
Reuse, Parameterized Reuse, and Hierarchical Reuse of Substructures in Evolving Electrical Circuits Using Genetic Programming John R.Koza 1 Forrest H Bennett III 2 David Andre 3 Martin A. Keane 4 1) Computer
Automated Synthesis of Computational Circuits Using Genetic Programming
Automated Synthesis of Computational Circuits Using Genetic Programming John R. Koza 258 Gates Building Stanford, California 94305-9020 koza@cs.stanford.edu http://www-csfaculty.stanford.edu/~koza/ Frank
Toward Evolution of Electronic Animals Using Genetic Programming
Toward Evolution of Electronic Animals Using Genetic Programming John R. Koza Computer Science Dept. 258 Gates Building Stanford University Stanford, California 94305 koza@cs.stanford.edu http://www-csfaculty.stanford.edu/~koza/
AUTOMATED DESIGN OF BOTH THE TOPOLOGY AND SIZING OF ANALOG ELECTRICAL CIRCUITS USING GENETIC PROGRAMMING
AUTOMATED TOPOLOGY AND SIZING OF ANALOG CIRCUITS AUTOMATED DESIGN OF BOTH THE TOPOLOGY AND SIZING OF ANALOG ELECTRICAL CIRCUITS USING GENETIC PROGRAMMING JOHN R. KOZA, FORREST H BENNETT III, DAVID ANDRE
Genetic Programming: Biologically Inspired Computation that Creatively Solves Non-Trivial Problems
Version 1 Submitted December 11, 1998 for Evolution as Computation Workshop (EAC) at DIMACS to be held in Princeton, New Jersey on January 11 12 (Monday Tuesday), 1999. Genetic Programming: Biologically
Four Problems for which a Computer Program Evolved by Genetic Programming is Competitive with Human Performance
Four Problems for which a Computer Program Evolved by Genetic Programming is Competitive with Human Performance John R. Koza Computer Science Dept. 258 Gates Building Stanford University Stanford, California
Evolution of a Time-Optimal Fly-To Controller Circuit using Genetic Programming
Evolution of a Time-Optimal Fly-To Controller Circuit using Genetic Programming John R. Koza Computer Science Dept. 258 Gates Building Stanford University Stanford, California 94305-9020 koza@cs.stanford.edu
Automatic Synthesis of the Topology and Sizing for Analog Electrical Circuits using Genetic Programming
1 Version 2 - Submitted December 31, 1998 for EUROGEN workshop in Jyvdskyld, Finland on May 30 June 3, 1999. 9,987 words 28 pages of text plus 17 figures. Automatic Synthesis of the Topology and Sizing
J. R. Koza Computer Science Dept., Stanford University, Stanford, CA
AUTOMATIC CREATION OF COMPUTER PROGRAMS FOR DESIGNING ELECTRICAL CIRCUITS USING GENETIC PROGRAMMING J. R. Koza Computer Science Dept., Stanford University, Stanford, CA 94305 E-mail: koza@cs.stanford.edu
Fourteen Instances where Genetic Programming has Produced Results that are Competitive with Results Produced by Humans
Genetic Programming Fourteen Instances where Genetic Programming has Produced Results that are Competitive with Results Produced by Humans JOHN R. KOZA Stanford University Stanford, California 94305 koza@genetic-programming.com
AUTOMATIC SYNTHESIS USING GENETIC PROGRAMMING OF BOTH THE TOPOLOGY AND SIZING FOR FIVE POST-2000 PATENTED ANALOG AND MIXED ANALOG-DIGITAL CIRCUITS
AUTOMATIC SYNTHESIS USING GENETIC PROGRAMMING OF BOTH THE TOPOLOGY AND SIZING FOR FIVE POST-2000 PATENTED ANALOG AND MIXED ANALOG-DIGITAL CIRCUITS Matthew J. Streeter Genetic Programming Inc. Mountain
Use of Genetic Programming for Automatic Synthesis of Post-2000 Patented Analog Electrical Circuits and Patentable Controllers
Use of Genetic Programming for Automatic Synthesis of Post-2000 Patented Analog Electrical Circuits and Patentable Controllers Matthew J. Streeter 1, Martin A. Keane 2, & John R. Koza 3 1 Genetic Programming
Use of Time-Domain Simulations in Automatic Synthesis of Computational Circuits Using Genetic Programming
Use of -Domain Simulations in Automatic Synthesis of Computational Circuits Using Genetic Programming William Mydlowec Genetic Programming Inc. Los Altos, California myd@cs.stanford.edu John R. Koza Stanford
Routine High-Return Human-Competitive Machine Learning
Routine High-Return Human-Competitive Machine Learning John R. Koza Stanford University koza@stanford.edu Matthew J. Streeter Genetic Programming Inc. matt@genetic-programming.com Martin A. Keane Econometrics
Version 3 June 25, 1996 for Handbook of Evolutionary Computation. Future Work and Practical Applications of Genetic Programming
1 Version 3 June 25, 1996 for Handbook of Evolutionary Computation. Future Work and Practical Applications of Genetic Programming John R. Koza Computer Science Department Stanford University 258 Gates
Submitted November 19, 1989 to 2nd Conference Economics and Artificial Intelligence, July 2-6, 1990, Paris
1 Submitted November 19, 1989 to 2nd Conference Economics and Artificial Intelligence, July 2-6, 1990, Paris DISCOVERING AN ECONOMETRIC MODEL BY. GENETIC BREEDING OF A POPULATION OF MATHEMATICAL FUNCTIONS
Routine Human-Competitive Machine Intelligence by Means of Genetic Programming
Routine Human-Competitive Machine Intelligence by Means of Genetic Programming John R. Koza *a, Matthew J. Streeter b, Martin A. Keane c a Stanford University, Stanford, CA, USA 94305 b Genetic Programming
Version 2 Submitted August 18, 1997 for Encyclopedia of Computer Science and Technology. Genetic Programming
Version 2 Submitted August 18, 1997 for Encyclopedia of Computer Science and Technology to be edited by Allen Kent and James G. Williams. 7,734 words. 1 1. Introduction Genetic Programming John R. Koza
Automatic Synthesis of Both the Topology and Numerical Parameters for Complex Structures Using Genetic Programming
Version 4 Submitted ---, 2001 for Engineering Design Synthesis: Understanding, Approaches and Tools, edited by: Amaresh Chakrabarti. Automatic Synthesis of Both the Topology and Numerical Parameters for
Human-competitive Applications of Genetic Programming
Human-competitive Applications of Genetic Programming John R. Koza Stanford Medical Informatics, Department of Medicine, School of Medicine, Department of Electrical Engineering, School of Engineering,
Producing human-competitive results is a primary reason why the AI and machine
What s AI Done for Me Lately? Genetic Programming s Human-Competitive Results John R. Koza, Stanford University Martin A. Keane, Econometrics Inc. Matthew J. Streeter, Genetic Programming Inc. Producing
TOWARD AUTOMATED DESIGN OF INDUSTRIAL-STRENGTH ANALOG CIRCUITS BY MEANS OF GENETIC PROGRAMMING
Chapter 8 TOWARD AUTOMATED DESIGN OF INDUSTRIAL-STRENGTH ANALOG CIRCUITS BY MEANS OF GENETIC PROGRAMMING John R. Koza 1, Lee W. Jones 2, Martin A. Keane 3, Matthew J. Streeter 4 and Sameer H. Al-Sakran
CYCLIC GENETIC ALGORITHMS FOR EVOLVING MULTI-LOOP CONTROL PROGRAMS
CYCLIC GENETIC ALGORITHMS FOR EVOLVING MULTI-LOOP CONTROL PROGRAMS GARY B. PARKER, CONNECTICUT COLLEGE, USA, parker@conncoll.edu IVO I. PARASHKEVOV, CONNECTICUT COLLEGE, USA, iipar@conncoll.edu H. JOSEPH
Reactive Planning with Evolutionary Computation
Reactive Planning with Evolutionary Computation Chaiwat Jassadapakorn and Prabhas Chongstitvatana Intelligent System Laboratory, Department of Computer Engineering Chulalongkorn University, Bangkok 10330,
LANDSCAPE SMOOTHING OF NUMERICAL PERMUTATION SPACES IN GENETIC ALGORITHMS
LANDSCAPE SMOOTHING OF NUMERICAL PERMUTATION SPACES IN GENETIC ALGORITHMS ABSTRACT The recent popularity of genetic algorithms (GA s) and their application to a wide range of problems is a result of their
Evolution of a Subsumption Architecture that Performs a Wall Following Task. for an Autonomous Mobile Robot via Genetic Programming. John R.
July 22, 1992 version. Evolution of a Subsumption Architecture that Performs a Wall Following Task for an Autonomous Mobile Robot via Genetic Programming John R. Koza Computer Science Department Stanford
Automatic Creation of Human-Competitive Programs and Controllers by Means of Genetic Programming
Ž. Genetic Programming and Evolvable Machines, 1, 121 164 2000 2000 Kluwer Academic Publishers. Manufactured in The Netherlands. Automatic Creation of Human-Competitive Programs and Controllers by Means
The Behavior Evolving Model and Application of Virtual Robots
The Behavior Evolving Model and Application of Virtual Robots Suchul Hwang Kyungdal Cho V. Scott Gordon Inha Tech. College Inha Tech College CSUS, Sacramento 253 Yonghyundong Namku 253 Yonghyundong Namku
Evolution of Sensor Suites for Complex Environments
Evolution of Sensor Suites for Complex Environments Annie S. Wu, Ayse S. Yilmaz, and John C. Sciortino, Jr. Abstract We present a genetic algorithm (GA) based decision tool for the design and configuration
Genetic Algorithms with Heuristic Knight s Tour Problem
Genetic Algorithms with Heuristic Knight s Tour Problem Jafar Al-Gharaibeh Computer Department University of Idaho Moscow, Idaho, USA Zakariya Qawagneh Computer Department Jordan University for Science
Application Note 4. Analog Audio Passive Crossover
Application Note 4 App Note Application Note 4 Highlights Importing Transducer Response Data Importing Transducer Impedance Data Conjugate Impedance Compensation Circuit Optimization n Design Objective
Automatic Synthesis of a Wire Antenna Using Genetic Programming
Automatic Synthesis of a Wire Antenna Using Genetic Programming William Comisky Genetic Programming Inc. Los Altos, California bcomisky@pobox.com Jessen Yu Genetic Programming Inc. Los Altos, California
DETERMINING AN OPTIMAL SOLUTION
DETERMINING AN OPTIMAL SOLUTION TO A THREE DIMENSIONAL PACKING PROBLEM USING GENETIC ALGORITHMS DONALD YING STANFORD UNIVERSITY dying@leland.stanford.edu ABSTRACT This paper determines the plausibility
Evolving Digital Logic Circuits on Xilinx 6000 Family FPGAs
Evolving Digital Logic Circuits on Xilinx 6000 Family FPGAs T. C. Fogarty 1, J. F. Miller 1, P. Thomson 1 1 Department of Computer Studies Napier University, 219 Colinton Road, Edinburgh t.fogarty@dcs.napier.ac.uk
Genetic Programming Approach to Benelearn 99: II
Genetic Programming Approach to Benelearn 99: II W.B. Langdon 1 Centrum voor Wiskunde en Informatica, Kruislaan 413, NL-1098 SJ, Amsterdam bill@cwi.nl http://www.cwi.nl/ bill Tel: +31 20 592 4093, Fax:
Meta-Heuristic Approach for Supporting Design-for- Disassembly towards Efficient Material Utilization
Meta-Heuristic Approach for Supporting Design-for- Disassembly towards Efficient Material Utilization Yoshiaki Shimizu *, Kyohei Tsuji and Masayuki Nomura Production Systems Engineering Toyohashi University
APPENDIX A to VOLUME A1 TIMS FILTER RESPONSES
APPENDIX A to VOLUME A1 TIMS FILTER RESPONSES A2 TABLE OF CONTENTS... 5 Filter Specifications... 7 3 khz LPF (within the HEADPHONE AMPLIFIER)... 8 TUNEABLE LPF... 9 BASEBAND CHANNEL FILTERS - #2 Butterworth
Evolution of a Controller with a Free Variable using Genetic Programming
Evolution of a Controller with a Free Variable using Genetic Programming John R. Koza Stanford University, Stanford, California koza@stanford.edu Jessen Yu Genetic Programming Inc., Los Altos, California
Performance Analysis of FIR Filter Design Using Reconfigurable Mac Unit
Volume 4 Issue 4 December 2016 ISSN: 2320-9984 (Online) International Journal of Modern Engineering & Management Research Website: www.ijmemr.org Performance Analysis of FIR Filter Design Using Reconfigurable
GENETICALLY DERIVED FILTER CIRCUITS USING PREFERRED VALUE COMPONENTS
GENETICALLY DERIVED FILTER CIRCUITS USING PREFERRED VALUE COMPONENTS D.H. Horrocks and Y.M.A. Khalifa Introduction In the realisation of discrete-component analogue electronic circuits it is common practice,
Genetic Programming: Turing s Third Way to Achieve Machine Intelligence
Version 2 - Submitted ---, 1999 for EUROGEN workshop in Jyvdskyld, Finland on May 30 June 3, 1999. Genetic Programming: Turing s Third Way to Achieve Machine Intelligence J. R. KOZA 1, F. H BENNETT 2 III,
Developing Frogger Player Intelligence Using NEAT and a Score Driven Fitness Function
Developing Frogger Player Intelligence Using NEAT and a Score Driven Fitness Function Davis Ancona and Jake Weiner Abstract In this report, we examine the plausibility of implementing a NEAT-based solution
Short Running Title: Genetic Modeling
Short Running Title: 1 Genetic Modeling Send communications to: John R. KOZA Computer Science Department, Stanford University, Stanford, CA 94305 USA EMAIL: Koza@Sunburn.Stanford.Edu PHONE: 415-941-0336.
CHAPTER 5 PERFORMANCE EVALUATION OF SYMMETRIC H- BRIDGE MLI FED THREE PHASE INDUCTION MOTOR
85 CHAPTER 5 PERFORMANCE EVALUATION OF SYMMETRIC H- BRIDGE MLI FED THREE PHASE INDUCTION MOTOR 5.1 INTRODUCTION The topological structure of multilevel inverter must have lower switching frequency for
Using the isppac 80 Programmable Lowpass Filter IC
Using the isppac Programmable Lowpass Filter IC Introduction This application note describes the isppac, an In- System Programmable (ISP ) Analog Circuit from Lattice Semiconductor, and the filters that
Constrained and Unconstrained evolution of LCR low-pass filters with oscillating length representation
2006 IEEE Congress on Evolutionary Computation Sheraton Vancouver Wall Centre Hotel, Vancouver, BC, Canada July 16-21, 2006 Constrained and Unconstrained evolution of LCR low-pass filters with oscillating
Multi-level and Multi-Objective Design Optimisation of a MEMS Bandpass Filter
Multi-level and Multi-Objective Design Optimisation of a MEMS Bandpass Filter Michael Farnsworth a,, Ashutosh Tiwari a, Meiling Zhu b a School of Aerospace, Transport Systems and Manufacturing, Cranfield
Evolutions of communication
Evolutions of communication Alex Bell, Andrew Pace, and Raul Santos May 12, 2009 Abstract In this paper a experiment is presented in which two simulated robots evolved a form of communication to allow
Memetic Crossover for Genetic Programming: Evolution Through Imitation
Memetic Crossover for Genetic Programming: Evolution Through Imitation Brent E. Eskridge and Dean F. Hougen University of Oklahoma, Norman OK 7319, USA {eskridge,hougen}@ou.edu, http://air.cs.ou.edu/ Abstract.
An Evolutionary Approach to the Synthesis of Combinational Circuits
An Evolutionary Approach to the Synthesis of Combinational Circuits Cecília Reis Institute of Engineering of Porto Polytechnic Institute of Porto Rua Dr. António Bernardino de Almeida, 4200-072 Porto Portugal
Learning Behaviors for Environment Modeling by Genetic Algorithm
Learning Behaviors for Environment Modeling by Genetic Algorithm Seiji Yamada Department of Computational Intelligence and Systems Science Interdisciplinary Graduate School of Science and Engineering Tokyo
A Genetic Algorithm-Based Controller for Decentralized Multi-Agent Robotic Systems
A Genetic Algorithm-Based Controller for Decentralized Multi-Agent Robotic Systems Arvin Agah Bio-Robotics Division Mechanical Engineering Laboratory, AIST-MITI 1-2 Namiki, Tsukuba 305, JAPAN agah@melcy.mel.go.jp
Evolution is an immensely powerful
Inventions By John R. Koza, Martin A. Keane and Matthew J. Streeter Evolution is an immensely powerful creative process. From the intricate biochemistry of individual cells to the elaborate structure of
Genetic Programming of Autonomous Agents. Senior Project Proposal. Scott O'Dell. Advisors: Dr. Joel Schipper and Dr. Arnold Patton
Genetic Programming of Autonomous Agents Senior Project Proposal Scott O'Dell Advisors: Dr. Joel Schipper and Dr. Arnold Patton December 9, 2010 GPAA 1 Introduction to Genetic Programming Genetic programming
GENETIC PROGRAMMING. In artificial intelligence, genetic programming (GP) is an evolutionary algorithmbased
GENETIC PROGRAMMING Definition In artificial intelligence, genetic programming (GP) is an evolutionary algorithmbased methodology inspired by biological evolution to find computer programs that perform
The Genetic Algorithm
The Genetic Algorithm The Genetic Algorithm, (GA) is finding increasing applications in electromagnetics including antenna design. In this lesson we will learn about some of these techniques so you are
A Short SPICE Tutorial
A Short SPICE Tutorial Kenneth H. Carpenter Department of Electrical and Computer Engineering Kanas State University September 15, 2003 - November 10, 2004 1 Introduction SPICE is an acronym for Simulation
CHAPTER 8: EXTENDED TETRACHORD CLASSIFICATION
CHAPTER 8: EXTENDED TETRACHORD CLASSIFICATION Chapter 7 introduced the notion of strange circles: using various circles of musical intervals as equivalence classes to which input pitch-classes are assigned.
Understanding Coevolution
Understanding Coevolution Theory and Analysis of Coevolutionary Algorithms R. Paul Wiegand Kenneth A. De Jong paul@tesseract.org kdejong@.gmu.edu ECLab Department of Computer Science George Mason University
A Divide-and-Conquer Approach to Evolvable Hardware
A Divide-and-Conquer Approach to Evolvable Hardware Jim Torresen Department of Informatics, University of Oslo, PO Box 1080 Blindern N-0316 Oslo, Norway E-mail: jimtoer@idi.ntnu.no Abstract. Evolvable
A Case Study of GP and GAs in the Design of a Control System
A Case Study of GP and GAs in the Design of a Control System Andrea Soltoggio Department of Computer and Information Science Norwegian University of Science and Technology N-749, Trondheim, Norway soltoggi@stud.ntnu.no
A Note on General Adaptation in Populations of Painting Robots
A Note on General Adaptation in Populations of Painting Robots Dan Ashlock Mathematics Department Iowa State University, Ames, Iowa 511 danwell@iastate.edu Elizabeth Blankenship Computer Science Department
Filters and Tuned Amplifiers
CHAPTER 6 Filters and Tuned Amplifiers Introduction 55 6. Filter Transmission, Types, and Specification 56 6. The Filter Transfer Function 60 6.7 Second-Order Active Filters Based on the Two-Integrator-Loop
System on a Chip. Prof. Dr. Michael Kraft
System on a Chip Prof. Dr. Michael Kraft Lecture 4: Filters Filters General Theory Continuous Time Filters Background Filters are used to separate signals in the frequency domain, e.g. remove noise, tune
Arcade Game Maker Product Line Requirements Model
Arcade Game Maker Product Line Requirements Model ArcadeGame Team July 2003 Table of Contents Overview 2 1.1 Identification 2 1.2 Document Map 2 1.3 Concepts 3 1.4 Reusable Components 3 1.5 Readership
AC CURRENTS, VOLTAGES, FILTERS, and RESONANCE
July 22, 2008 AC Currents, Voltages, Filters, Resonance 1 Name Date Partners AC CURRENTS, VOLTAGES, FILTERS, and RESONANCE V(volts) t(s) OBJECTIVES To understand the meanings of amplitude, frequency, phase,
AUTOMATED DESIGN OF COMPLEX STRUCTURES USING DARWINIAN EVOLUTION AND GENETIC PROGRAMMING EE 380 FEBRUARY 18, 2009
1 AUTOMATED DESIGN OF COMPLEX STRUCTURES USING DARWINIAN EVOLUTION AND GENETIC PROGRAMMING EE 380 FEBRUARY 18, 2009 John R. Koza Consulting Professor Department of Electrical Engineering Stanford University
Analog Design-filters
Analog Design-filters Introduction and Motivation Filters are networks that process signals in a frequency-dependent manner. The basic concept of a filter can be explained by examining the frequency dependent
Analog Lowpass Filter Specifications
Analog Lowpass Filter Specifications Typical magnitude response analog lowpass filter may be given as indicated below H a ( j of an Copyright 005, S. K. Mitra Analog Lowpass Filter Specifications In the
Cooperative Behavior Acquisition in A Multiple Mobile Robot Environment by Co-evolution
Cooperative Behavior Acquisition in A Multiple Mobile Robot Environment by Co-evolution Eiji Uchibe, Masateru Nakamura, Minoru Asada Dept. of Adaptive Machine Systems, Graduate School of Eng., Osaka University,
Vesselin K. Vassilev South Bank University London Dominic Job Napier University Edinburgh Julian F. Miller The University of Birmingham Birmingham
Towards the Automatic Design of More Efficient Digital Circuits Vesselin K. Vassilev South Bank University London Dominic Job Napier University Edinburgh Julian F. Miller The University of Birmingham Birmingham
A Review on Genetic Algorithm and Its Applications
2017 IJSRST Volume 3 Issue 8 Print ISSN: 2395-6011 Online ISSN: 2395-602X Themed Section: Science and Technology A Review on Genetic Algorithm and Its Applications Anju Bala Research Scholar, Department
A FRAMEWORK FOR PERFORMING V&V WITHIN REUSE-BASED SOFTWARE ENGINEERING
A FRAMEWORK FOR PERFORMING V&V WITHIN REUSE-BASED SOFTWARE ENGINEERING Edward A. Addy eaddy@wvu.edu NASA/WVU Software Research Laboratory ABSTRACT Verification and validation (V&V) is performed during
A Graph Grammar Based Approach to Automated Multi-Objective Analog Circuit Design
A Graph Grammar Based Approach to Automated Multi-Objective Analog Circuit Design Angan Das and Ranga Vemuri Department of Electrical and Computer Engineering, University of Cincinnati, Cincinnati, OH
Introduction to oscilloscope. and time dependent circuits
Physics 9 Intro to oscilloscope, v.1.0 p. 1 NAME: SECTION DAY/TIME: TA: LAB PARTNER: Introduction to oscilloscope and time dependent circuits Introduction In this lab, you ll learn the basics of how to
The Discrete Fourier Transform. Claudia Feregrino-Uribe, Alicia Morales-Reyes Original material: Dr. René Cumplido
The Discrete Fourier Transform Claudia Feregrino-Uribe, Alicia Morales-Reyes Original material: Dr. René Cumplido CCC-INAOE Autumn 2015 The Discrete Fourier Transform Fourier analysis is a family of mathematical
Electrical Engineer. Lab2. Dr. Lars Hansen
Electrical Engineer Lab2 Dr. Lars Hansen David Sanchez University of Texas at San Antonio May 5 th, 2009 Table of Contents Abstract... 3 1.0 Introduction and Product Description... 3 1.1 Problem Specifications...
Enhancing Embodied Evolution with Punctuated Anytime Learning
Enhancing Embodied Evolution with Punctuated Anytime Learning Gary B. Parker, Member IEEE, and Gregory E. Fedynyshyn Abstract This paper discusses a new implementation of embodied evolution that uses the
Lab 1: Basic RL and RC DC Circuits
Name- Surname: ID: Department: Lab 1: Basic RL and RC DC Circuits Objective In this exercise, the DC steady state response of simple RL and RC circuits is examined. The transient behavior of RC circuits
Evolution, Re-evolution, and Prototype of an X-Band Antenna for NASA s Space Technology 5 Mission
Evolution, Re-evolution, and Prototype of an X-Band Antenna for NASA s Space Technology 5 Mission Jason D. Lohn 1, Gregory S. Hornby 2, and Derek S. Linden 3 1 Computational Sciences Division NASA Ames
An Optimized Performance Amplifier
Electrical and Electronic Engineering 217, 7(3): 85-89 DOI: 1.5923/j.eee.21773.3 An Optimized Performance Amplifier Amir Ashtari Gargari *, Neginsadat Tabatabaei, Ghazal Mirzaei School of Electrical and
Implicit Fitness Functions for Evolving a Drawing Robot
Implicit Fitness Functions for Evolving a Drawing Robot Jon Bird, Phil Husbands, Martin Perris, Bill Bigge and Paul Brown Centre for Computational Neuroscience and Robotics University of Sussex, Brighton,
A Numerical Approach to Understanding Oscillator Neural Networks
A Numerical Approach to Understanding Oscillator Neural Networks Natalie Klein Mentored by Jon Wilkins Networks of coupled oscillators are a form of dynamical network originally inspired by various biological
Fall 2003 BMI 226 / CS 426 Notes P- 1 PARALLELIZATION
Fall 2003 BMI 226 / CS 426 Notes P- 1 PARALLELIZATION Fall 2003 BMI 226 / CS 426 Notes P- 2 COMPUTER TIME IS THE MOTHER'S MILK OF MACHINE INTELLIGENCE Fall 2003 BMI 226 / CS 426 Notes P- 3 THE GOOD GNU
New Genetic Operators to Facilitate Understanding of Evolved Transistor Circuits
New Genetic Operators to Facilitate Understanding of Evolved Transistor Circuits Martin Trefzer, Jörg Langeheine, Johannes Schemmel, Karlheinz Meier University of Heidelberg Kirchhoff-Institute for Physics
Total Harmonic Distortion Minimization of Multilevel Converters Using Genetic Algorithms
Applied Mathematics, 013, 4, 103-107 http://dx.doi.org/10.436/am.013.47139 Published Online July 013 (http://www.scirp.org/journal/am) Total Harmonic Distortion Minimization of Multilevel Converters Using
AUTOMATED INVENTION BY MEANS OF GENETIC PROGRAMMING AAAI-2004 TUTORIAL SAN JOSE SUNDAY JULY 25, AM
1 AUTOMATED INVENTION BY MEANS OF GENETIC PROGRAMMING AAAI-2004 TUTORIAL SAN JOSE SUNDAY JULY 25, 2004 9AM John R. Koza Stanford University koza@stanford.edu http://smi-web.stanford.edu/people/koza/ http://www.genetic-programming.org
Fault Location Using Sparse Wide Area Measurements
319 Study Committee B5 Colloquium October 19-24, 2009 Jeju Island, Korea Fault Location Using Sparse Wide Area Measurements KEZUNOVIC, M., DUTTA, P. (Texas A & M University, USA) Summary Transmission line
Achieving Desirable Gameplay Objectives by Niched Evolution of Game Parameters
Achieving Desirable Gameplay Objectives by Niched Evolution of Game Parameters Scott Watson, Andrew Vardy, Wolfgang Banzhaf Department of Computer Science Memorial University of Newfoundland St John s.
2. Simulated Based Evolutionary Heuristic Methodology
XXVII SIM - South Symposium on Microelectronics 1 Simulation-Based Evolutionary Heuristic to Sizing Analog Integrated Circuits Lucas Compassi Severo, Alessandro Girardi {lucassevero, alessandro.girardi}@unipampa.edu.br
EE320L Electronics I. Laboratory. Laboratory Exercise #6. Current-Voltage Characteristics of Electronic Devices. Angsuman Roy
EE320L Electronics I Laboratory Laboratory Exercise #6 Current-Voltage Characteristics of Electronic Devices By Angsuman Roy Department of Electrical and Computer Engineering University of Nevada, Las
Department of Electrical & Computer Engineering Technology. EET 3086C Circuit Analysis Laboratory Experiments. Masood Ejaz
Department of Electrical & Computer Engineering Technology EET 3086C Circuit Analysis Laboratory Experiments Masood Ejaz Experiment # 1 DC Measurements of a Resistive Circuit and Proof of Thevenin Theorem
INTERACTIVE DYNAMIC PRODUCTION BY GENETIC ALGORITHMS
INTERACTIVE DYNAMIC PRODUCTION BY GENETIC ALGORITHMS M.Baioletti, A.Milani, V.Poggioni and S.Suriani Mathematics and Computer Science Department University of Perugia Via Vanvitelli 1, 06123 Perugia, Italy
Graduate Texts in Mathematics. Editorial Board. F. W. Gehring P. R. Halmos Managing Editor. c. C. Moore
Graduate Texts in Mathematics 49 Editorial Board F. W. Gehring P. R. Halmos Managing Editor c. C. Moore K. W. Gruenberg A.J. Weir Linear Geometry 2nd Edition Springer Science+Business Media, LLC K. W.
Analog-aware Schematic Synthesis
12 Analog-aware Schematic Synthesis Yuping Wu Institute of Microelectronics, Chinese Academy of Sciences, China 1. Introduction An analog circuit has great requirements of constraints on circuit and layout
Intrinsic Evolution of Analog Circuits on a Programmable Analog Multiplexer Array
Intrinsic Evolution of Analog Circuits on a Programmable Analog Multiplexer Array José Franco M. Amaral 1, Jorge Luís M. Amaral 1, Cristina C. Santini 2, Marco A.C. Pacheco 2, Ricardo Tanscheit 2, and
Population Adaptation for Genetic Algorithm-based Cognitive Radios
Population Adaptation for Genetic Algorithm-based Cognitive Radios Timothy R. Newman, Rakesh Rajbanshi, Alexander M. Wyglinski, Joseph B. Evans, and Gary J. Minden Information Technology and Telecommunications
Electric Circuit Theory
Electric Circuit Theory Nam Ki Min nkmin@korea.ac.kr 010-9419-2320 Chapter 15 Active Filter Circuits Nam Ki Min nkmin@korea.ac.kr 010-9419-2320 Contents and Objectives 3 Chapter Contents 15.1 First-Order
DISCRETE-TIME CHANNELIZERS FOR AERONAUTICAL TELEMETRY: PART II VARIABLE BANDWIDTH
DISCRETE-TIME CHANNELIZERS FOR AERONAUTICAL TELEMETRY: PART II VARIABLE BANDWIDTH Brian Swenson, Michael Rice Brigham Young University Provo, Utah, USA ABSTRACT A discrete-time channelizer capable of variable
444 Index. F Fermi potential, 146 FGMOS transistor, 20 23, 57, 83, 84, 98, 205, 208, 213, 215, 216, 241, 242, 251, 280, 311, 318, 332, 354, 407
Index A Accuracy active resistor structures, 46, 323, 328, 329, 341, 344, 360 computational circuits, 171 differential amplifiers, 30, 31 exponential circuits, 285, 291, 292 multifunctional structures,