where and are polynomials with real coefficients and of degrees m and n, respectively. Assume that and have no zero on axis.

Similar documents
N2-1. The Voltage Source. V = ε ri. The Current Source

Exercise 1 (MC Question)

ISSN: [Reddy & Rao* et al., 5(12): December, 2016] Impact Factor: 4.116

Design of FIR Filter using Filter Response Masking Technique

HYBRID FUZZY PD CONTROL OF TEMPERATURE OF COLD STORAGE WITH PLC

Figure Geometry for Computing the Antenna Parameters.

4 Trigonometric and Inverse Trigonometric Functions

Parameters of spinning AM reticles

Low-Complexity Time-Domain SNR Estimation for OFDM Systems

School of Electrical and Computer Engineering, Cornell University. ECE 303: Electromagnetic Fields and Waves. Fall 2007

Sliding Mode Control for Half-Wave Zero Current Switching Quasi-Resonant Buck Converter

Sinusoidal Oscillators

Key Laboratory of Earthquake Engineering and Engineering Vibration, China Earthquake Administration, China

An Efficient Control Approach for DC-DC Buck-Boost Converter

Real-time Self Compensating AC/DC Digitally Controlled Power Supply

ONE-WAY RADAR EQUATION / RF PROPAGATION

Multiagent Reinforcement Learning Dynamic Spectrum Access in Cognitive Radios

Discrepancies Between Euclidean and Spherical Trigonometry. David Eigen

Optimized Fuzzy Controller Design to Stabilize Voltage and Frequency Amplitude in a Wind Turbine Based on Induction Generator Using Firefly Algorithm

Investigation. Name: a About how long would the threaded rod need to be if the jack is to be stored with

Analysis of the optimized low-nonlinearity lateral effect sensing detector

DESIGN AND PARAMETRIC EVALUATION OF RECTANGULAR MICROSTRIP PATCH ANTENNA FOR GSM APPLICATION

Spectrum Sharing between Public Safety and Commercial Users in 4G-LTE

Cyclic Constellation Mapping Method for PAPR Reduction in OFDM system

AUTO-TUNED MINIMUM-DEVIATION DIGITAL CONTROLLER FOR LLC RESONANT CONVERTERS

Discussion #7 Example Problem This problem illustrates how Fourier series are helpful tools for analyzing electronic circuits. Often in electronic

Voltage Control of a 12/8 Pole Switched Reluctance Generator Using Fuzzy Logic

A Transmission Scheme for Continuous ARQ Protocols over Underwater Acoustic Channels

ABSTRACTT FFT FFT-' Proc. of SPIE Vol U-1

Analysis of Occurrence of Digit 0 in Natural Numbers Less Than 10 n

Variance? which variance? R squared effect size measures in simple mediation models

10! !. 3. Find the probability that a five-card poker hand (i.e. 5 cards out of a 52-card deck) will be:

Near-field Computation and. Uncertainty Estimation using Basic. Cylindrical-Spherical Formulae

Controller Design of Discrete Systems by Order Reduction Technique Employing Differential Evolution Optimization Algorithm

IAS 2.4. Year 12 Mathematics. Contents. Trigonometric Relationships. ulake Ltd. Robert Lakeland & Carl Nugent

Control Limits of Three-Phase AC Voltage Controller under Induction Motor Load A. I. Alolah Ali M. Eltamaly R. M. Hamouda

THE UNIVERSITY OF NEW SOUTH WALES. School of Electrical Engineering & Telecommunications

An SMC based RZVDPWM Algorithm of Vector Controlled Induction Motor Drive for Better Speed Response with Reduced Acoustical Noise

OPTIMUM MEDIUM ACCESS TECHNIQUE FOR NEXT GENERATION WIRELESS SYSTEMS

Chamber Influence Estimation for Radiated Emission Testing in the Frequency Range of 1 GHz to 18 GHz

Small Loop Antenna and Duality Theorem

Design of an LLC Resonant Converter Using Genetic Algorithm

New Approach for Optimizing Control of Switched Reluctance Generator

Two Novel Measurements for the Drive-Mode Resonant Frequency of a Micromachined Vibratory Gyroscope

Design of A Circularly Polarized E-shaped Patch Antenna with Enhanced Bandwidth for 2.4 GHz WLAN Applications

INCREMENTAL REDUNDANCY (IR) SCHEMES FOR W-CDMA HS-DSCH

MACRO-DIVERSITY VERSUS MICRO-DIVERSITY SYSTEM CAPACITY WITH REALISTIC RECEIVER RFFE MODEL

Published in: International Journal of Material Forming. Document Version: Peer reviewed version

Available online at ScienceDirect. Procedia Engineering 100 (2015 )

The mathematical structure known as a graph has the valuable feature of

The Periodic Ambiguity Function Its Validity and Value

A Lifting Algorithm for Output-only Continuous Scan Laser Doppler Vibrometry

Spread Spectrum Codes Identification by Neural Networks

ScienceDirect. Simplified Generation of Electromagnetic Field within EMC Immunity Test Area

Analysis of a Fractal Microstrip Patch Antenna

Optimal Design of Smart Mobile Terminal Antennas for Wireless Communication and Computing Systems

Current Compensation Techniques for Lowvoltage High-performance Current Mirror Circuits

AMERICAN NATIONAL STANDARD

Analysis and Design of a 1MHz LLC Resonant Converter with Coreless Transformer Driver

Bidirectional Contactless Power Transfer System Expandable from Unidirectional System

A New Buck-Boost DC/DC Converter of High Efficiency by Soft Switching Technique

Figure 1-1 Sample Antenna Pattern

Analysis and Implementation of LLC Burst Mode for Light Load Efficiency Improvement

Journal of Applied Science and Agriculture

CHAPTER 4 STABILITY ANALYSIS OF INDUCTION GENERATORS USING STATCOM

Single-to-three phase induction motor sensorless drive system

Proposal of Circuit Breaker Type Disconnector for Surge Protective Device

Reliability Model of Power Transformer with ONAN Cooling

GAMMA SHAPED MONOPOLE PATCH ANTENNA FOR TABLET PC

ECEN326: Electronic Circuits Fall 2017

ME 4710 Motion and Control Frequency Domain Analysis

SERVO CONTROL OF RF CAVITIES UNDER BEAM LOADING

Performance Analysis of Z-Source Inverter Considering Inductor Resistance

Telescope interferometers: an alternative to classical wavefront sensors

Performance Analysis of the Parallel Optical All-pass Filter Equalizer for Chromatic Dispersion Compensation at 10 Gb/s

Higher moments method for generalized Pareto distribution in flood frequency analysis

Noise Attenuation Due to Vegetation

Performance analysis of ARQ Go-Back-N protocol in fading mobile radio channels

TECHNICAL REPORT: CVEL Maximum Radiated Emission Calculator: Power Bus EMI Algorithm. Chentian Zhu and Dr. Todd Hubing. Clemson University

Analytical Performance Evaluation of Mixed Services with Variable Data Rates for the Uplink of UMTS

Optimization of the law of variation of shunt regulator impedance for Proximity Contactless Smart Card Applications to reduce the loading effect.

Signal Extraction Technology

Real-Time Fault Diagnostics for a Permanent Magnet Synchronous Motor Drive for Aerospace Applications

Analysis, Design, and Performance Evaluation of Asymmetrical Half-Bridge Flyback Converter for Universal-Line-Voltage-Range Applications

Demosaicking using Adaptive Bilateral Filters

Impact of bilateral filter parameters on medical image noise reduction and edge preservation

WIRELESS SENSORS EMBEDDED IN CONCRETE

Distributed Spectrum Allocation via Local Bargaining

An Ultra Low Power Segmented Digital-to-Analog Converter

Optimal Eccentricity of a Low Permittivity Integrated Lens for a High-Gain Beam-Steering Antenna

The reliability of wireless backhaul mesh networks

Optimal Strategies in Jamming Resistant Uncoordinated Frequency Hopping Systems. Bingwen Zhang

Determination of The Winding Inductances Of A Two-Phase Machine.

Derangements. Brian Conrey and Tom Davis and March 23, 2000

Steve Alpern, Thomas Lidbetter, Alec Morton, and Katerina Papadaki Patrolling a pipeline

Near-field emission prediction of CESAME

Fault severity diagnosis of rolling element bearings based on kurtogram and envelope analysis

Week 5. Lecture Quiz 1. Forces of Friction, cont. Forces of Friction. Forces of Friction, final. Static Friction

Optimised Wireless Network Using Smart Mobile Terminal Antenna (SMTA) System

HISTORY AND PROGRESS OF THE TOW THOMAS BI-QUADRATIC FILTER. PART I. GENERATION AND OP AMP REALIZATIONS

Transcription:

function whee is an unknown constant epesents fo the un-modeled dynamics The pape investigates the position contol of electical moto dives that can be configued as stuctue of Fig 1 This poblem is fomulated as follows Conside the basic stuctue of a moto dive system assume that 1 The fequency esponse fom voltage u to oto position θ is available 2 is an unknown constant Unde these assumptions the contol objective is to calculate the family of PID contolles that stabilize the uncetain plants whee satisfies -nom consideation on complementay sensitivity function in the pesence of un-modeled dynamics u W(s)Δ Mechanical Pat T e Electical Pat θ whee ae polynomials with eal coefficients of degees m n, espectively Assume that have no zeo on axis Let detemine the numbe of open ight halp plane (RHP) (open left half plane (LHP)) zeos poles of Also let denotes the net change in phase of as uns fom to Then (1) Define the (Huwitz) signatue of as (2) since has no pole zeo on axis, it can be witten (3) The value of can be calculated fom the fequency data of can be calculated fom (3) Let whee denote the eal imaginay pats of, espectively Assume that the eal, distinct, finite zeos of denote as such that Fig 1 Moto dive system with un-modeled dynamics + _ e PID u P(s,q) Fig 2 Feedback stuctue with uncetain plant PID Fig 2 shows the feedback system with PID contolle in which is an uncetain plant can be substituted by In the next section, a suvey on the algoithm poposed fo nominal stability in [11] its genealization to uncetain plants is pesented it is illustated that the contol objectives of this pape on moto dive system can be satisfied in a same way III PID CONTROLLER DESIGN: MODEL FREE APPROCH θ conside the modified PID contolle as whee, ae the popotional, integal deivative coefficients of PID contolle T is a positive constant Lemma 1: Let Then the closed loop stability is equal to Let whee In this section, the poblem of achieving the family of stabilizing PID contolles fo nominal stable plant poposed in [11] is eviewed Then by veifying a theoem, this appoach is genealized to plants with an uncetain paamete Simila appoach is poposed fo unstable plants which ae omitted hee because of oom limitation The feedback system with PID contolle is shown in Fig 2 Fist, some mathematical peliminaies ae intoduced Conside a eal ational function http://apcaastedu Conside define whee (4) Theoem 1: Let denote the distinct fequencies of odd multiplicities which ae solutions of Detemine stings of integes whee such that: 45

fo n-m even: [ fo n-m odd: [ Also let (5) (6) Then fo, the values of coesponding to the closed loop stability ae given by (7) whee i t s ae taken fom stings satisfy (5) o (6) s ae taken fom the solutions of (4) Next theoem shows how to calculate the admissible ange of Theoem 2: The necessay condition of existence the stabilizing PID fo LTI plants is that thee exists such that has at least R distinct oots of odd multiplicities such that { (8) The pocedue fo calculation the family of stabilizing PID contolles is summaized in the following algoithm [11] Algoithm 1 1 Detemine the elative degee of plant fom high fequency slope of bode magnitude of 2 Detemine fom (2) 3 Detemine fom (3) 4 Detemine fo fom (4) 5 Apply Theoem 2 to detemine the ange of 6 Fo, solve (4) obtain 7 Let Detemine fom (5) o (6) 8 Fo, detemine the values fom (7) 9 Change go to step 6 to obtain the whole family of stabilizing PID contolles Now the main esult of this pape will be pesented Conside an uncetain eal ational plant whee is an uncetain paamete The contol goal is to calculate the family of obust stabilizing PID contolles fo the uncetain plant The feedback stuctue is shown in Fig 2 The only equied data is the set of fequency esponses including fo In the poposed appoach of Algoithm 1, oots of function in (8), can be obtained fom (4) the ange of is calculated based on the value of Also the ange of admissible values obtain fom (7) which have the slope equal to If by monotonic vaiation of the uncetain paamete q, vaies monotonically, some substantial esults could be concluded Fist of all, the ange of admissible fo stabilizing the uncetain plant is the common ange of admissible fo two plants Second, since the slope of linea inequalities in (7) vaies monotonically with monotonic vaying the uncetain paamete, the inequalities coesponding to have the maximum minimum slopes, not necessaily espectively Finally, since vaiation of the uncetain paamete, does not lead to monotonic vaiation in the the common space of inequality (7) fo is lage than the admissible values An appoach to find the exact ange of admissible paametes is pesented in the following emak Remak 1: A simple appoach to exclude these egions is to choose some test points in those egions analyzing thei stability So monotonic vaiation of the function makes it easy to calculate the family of stabilizing PID contolles fo plants with an uncetain paamete Theoem 3: Let be a eal ational tansfe function is an uncetain paamete If the Numbe of zeos poles the numbe of RHP zeos of ae fixed fo then the family of stabilizing PID contolles fo uncetain plant is a subspace of common space between two set of stabilizing PID contolles fo if one of the below constaints satisfy { (9) { (1) fo only one value of uncetain paamete q fo Poof: If monotonic vaying the uncetain paamete leads to monotonic vaying the function w, then the family of stabilizing PID contolles fo uncetain plant is a subspace of common space between two set of stabilizing PID contolles fo Then http://apcaastedu 46

Without loss of geneality, conside the case of monotonic inceasing in Inceasing the uncetain paamete leads to inceasing in if the following inequality holds: Assumption 1: Let (12) note that the symbol > could be invesed; but fo simplicity it is assumed as > Finally, conside the eal ational symbol of (12) So (11) can be tansfomed to (13) by change of vaiable as: the inequality (13) can be changed to: Assumption 2: Let (14) (15) (16) simila to Assumption 1, the eal ational symbol of (15) (16) will be consideed finally Fom (15) (16), the inequality (14) leads to: since, then that is one of the necessay conditions imply that inceasing the uncetain paamete leads to inceasing in Assumptions 1 2 imply anothe necessay condition as Remak 2: Bode diagams coesponding to could be ecognized fom the fequency esponses set of uncetain plant In fact if (9) o (1) holds, then fom monotonic vaiation of the function, it could be deduced that the uppemost lowemost plots of ae coesponding to, not necessaily espectively The following Coollay shows that the poblem can be solved easie when the uncetain paamete is in the feedfowad path of contol loop Coollay 1 Let P( s, q ) be a eal ational function http://apcaastedu q{ q, q } fo an uncetain LTI plant min max is an uncetain paamete that appeas in the feedfowad path Then the family of stabilizing PID contolles fo uncetain plant is a subset of common set between two set of stabilizing PID paametes fo P( s, qmin ) max P( s, q ) if one of the below sets satisfy ( P P) i : if g( ) is ascendant ( P P) i : if g( ) is decendant (17) fo fo only one value of uncetain paamete q poof: It can be witten P( j, q) qp ( j) qp( j) qp( j) qpi( j) g( ) 2 q ( P ( j) P( j)) Then i i dg 1 P 2 Pi 2 2 dq q P Pi So g( ) is monotonic if (17) satisfies The pocedue fo calculating the family of stabilizing PID contolles is summaized in the below algoithm Algoithm 2 1 Detemine fo fom plots using Remak 2 2 Using Algoithm 1, calculate the family of stabilizing PID contolles fo two plants Detemine the common space of PID paametes between two calculated family The family of stabilizing PID contolles fo uncetain plant is a subspace of this common space 3 Calculate the exact family of stabilizing PID contolles using the appoach poposed in Remak 1 Now it can be shown that the poblem of satisfying some pefomance specifications fo uncetain plant could be tansfomed to poblem of obust stabilizing the uncetain plant with additional vitual uncetain paamete Many pefomance attainment poblems fo uncetain plant can be cast as the simultaneously stabilization of the uncetain plant the family of eal complex plants [11] Fo example an -nom achievment on the complementay sensitivity function, that is, is equivalent to simultaneously stabilizing the uncetain plant as 47

(18) whee is a vitual uncetain paamete is a weight selected by designe Let the family of stabilizing PID paametes fo uncetain plant can be calculated fom Algoithm 1 Let in (18) be a eal ational function If coesponding to (18), vaies monotonically by monotonic vaying the vitual uncetain paamete, fo any any, then the family of stabilizing PID contolles fo uncetain plant that satisfy the -nom specification on the complementay sensitivity function, is a subspace of PID contolles obtained fom simultaneously stabilization of two plants family of obust PID contolles fo plants faced to loss of effectiveness can be calculated by the appoach poposed in the pevious section Fo this special case, the poblem of contolle synthesis could be hled easily using Coollay 1 Example 3: Robustness in plants with a dominant uncetain paamete Some plants that ae faced to paamete vaiations can be appoximated by plants with a dominant uncetain paamete Obviously the poposed appoach can be applied in this case to calculate the family of obust stabilizing PID contolle [ ] (19) [ ] (2) the exact family can be obtained using the appoach poposed in Remak 1 IV A REVIEW ON THE APPLICATIONS OF THE PROPOSED APPROACH In this section, some applications of the poposed method ae pesented In fact, it is illustated that some contol objectives can be cast as obust stabilizing of a plant with an uncetain paamete Fo example: Example 1: Pefomance achievement Many pefomance achievement poblems fo uncetain plant P( s, q) can be cast as the simultaneously stabilization of the uncetain plant the family of eal complex plants Some of these pefomance achievement poblems fo nominal plant ae listed in [11]; fo example, the poblem of H -nom achievement on the complementay sensitivity function is equivalent to simultaneously stabilizing the plant P(s) the family of eal plants C P ( s, q, ) : q{ q, q }, {,2 } min max the exact family can be obtained using some test points The othe specifications such as H - nom achievement on the sensitivity function phase magin can be satisfied by the same appoach Example 2: Robustness against loss of effectiveness in actuato Fig 3 Shows the feedback stuctue of a plant with PID faced to loss of effectiveness in actuato whee L is the paamete coesponding to loss of effectiveness belongs to (,1] Obviously this stuctue can be cast as a plant with an uncetain paamete, ie, P( j, L) with PID whee L q is the uncetain paamete Thee is simila case when loss of effectiveness happens in sensos Thus the Fig 3 The feedback stuctue of plant faced to loss of actuato effectiveness with PID Magnitude (db) Phase (deg) 15 5-5 54 45 36 27 18 9 1-2 1 1 2 Fequency (ad/sec) Fig 4 The fequency esponse of moto dive system V SIMULATION ON INDUCTION MOTOR The model studied in this section is an induction moto dive system intoduced in [9, 12] that has the simila stuctue to Fig 1 in [14] its nominal fequency esponse is shown in Fig 4 The fequency esponse of such systems can be obtained by vitual sine sweeping [13] The coesponding plot of is shown in Fig 5 Since (fom (8)), the admissible ange of is Calculating the values fo T=1 is esulted to the following inequalities { Bode Diagam http://apcaastedu 48

k p :=g() k p :=g() k p -2-4 -6 5 4 3 2 1 2 3 4 5 6 7 8 9 =715 g()=2 1 2 3 4 5 6 7 8 Fig 5 The plot of function g with the line kp=2 Fig 6 The whole family of stabilizing PID contolles fo nominal model of induction moto dive system Theefoe the family of obust PID contolles fo the induction moto dive with uncetainties can be calculated by stabilizing the following two plants { =181 g()=2 The ange of obust stabilizing PID paametes fo induction moto faced to un-modeled dynamics is illustated in Fig 7-a The admissible ange of fo all admissible is shown in Fig 7-b k p 35 3 25 2 15 6 4 2 6-5 5 k d 5 1 5 5 4 4 3 3 2 2 (a) -5 =812 g()=2-2 k d 5 2 k d 15 2 3 4 5 6 Fig 7 (a) Robust stabilizing PID paametes fo induction moto dive faced to un-modeled dynamics; (b) the admissible ange of (,k d) (b) Admissible ange of (,k d ) fo all admissible k p http://apcaastedu 14 12 8 Admissible ange 6 fo pefomance attainment 4 2 Test points: k p =5 Admissible ange fo nominal stability Admissible ange fo unmodeled dynamics -2-1 1 2 3 4 5 6 7 k d Fig 8 The ange of admissible (,k d)values that satisfy diffeent contol objectives fo k p=5 Now conside the poblem of satisfying -nom on the complementay sensitivity function The admissible ange of values that satisfy this pefomance citeion fo is shown in Fig 8 togethe with egions coesponding to nominal stability stability in the pesence of un-modeled dynamics The whole ange of values can be obtained similaly The step esponses of closed loop system with contolles coesponding to test points maked in Fig 8 ae shown in Fig 9 Applying the PID contolle that satisfies pefomance consideation is esulted to deceasing the oveshoot Also, the contol signals coesponding to selected contolles ae plotted in Fig 1 It can be seen that the contol signals dift to zeo afte the easonable times Fo bette tacking of the poposed appoach in this pape, the eades can efe to the many academic examples implemented in [15] VI CONCLUSION In this pape, a obust contol appoach pesented based on a geneal model fo diffeent types of electical moto dives The poblem of obustness against un-modeled dynamics in moto dives is tansfomed to the poblem of stabilizing a plant with an uncetain paamete It is shown that knowing the fequency esponses of moto dive system coesponding to maximum minimum values of uncetainty is sufficient to calculate the family of obust stabilizing PID contolles In fact, thee is no need to plant mathematical model Also it is illustated that the poblem of -nom achievement on the complementay sensitivity function can be solved by the same appoach Though the pape, it is assumed that the fequency esponse of plant is available Such an assumption is often valid in many pactical applications Also, this is an assumption that has aleady been used seveal times in othe papes dealing with contolle synthesis using fequency domain data [5,7,11,13,15-17] 49