Available online at www.sciencedirect.com ScienceDirect Procedia Computer Science 92 (2016 ) 36 41 2nd International Conference on Intelligent Computing, Communication & Convergence (ICCC-2016) Srikanta Patnaik, Editor in Chief Conference Organized by Interscience Institute of Management and Technology Bhubaneswar, Odisha, India Krill Herd Algorithm based Real Power Generation Reallocation for improvement of Voltage Profile Ch. Jayasree a and B.Sravan Kumar b * a PG Student, Department of EEE, GITAM University, Visakhapatnam-530045, INDIA b Assistant Professor, Department of EEE, GITAM University, Visakhapatnam-530045, INDIA Abstract Present-day Electric Power Systems are driven under much stressed circumstances when compared to the past and creating a developing need for accuracy, flexibility, and reliability in the areas of Transmission, Distribution and Electric Power Generation. In all stages of power system, voltage stability problems are increasing more and more. So, the only alternate solution for these problems is proper placement and sizing of UPFC. The paper presents the Placement and Tuning of UPFC for a multi-objective function consisting of minimization of transmission losses, load voltage deviation. Here L-index is used to place the UPFC in a specified location i.e., weakest bus, critical line and the weak area of the system. In this paper, a newly developed meta-heuristic algorithm named Krill Herd (KH) is introduced to solve multi-objective problem of optimization. Simulation is carried on IEEE 14-bus system and the results have been compared with the Genetic algorithm with and without UPFC. 2016 2014 The Authors. Published by by Elsevier Elsevier B.V. B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Selection and peer-review under responsibility of scientific committee of Missouri University of Science and Technology. Peer-review under responsibility of the Organizing Committee of ICCC 2016 * Corresponding author. Tel.: +91-9492622501 E-mail address:sravanbali@gmail.com 1877-0509 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the Organizing Committee of ICCC 2016 doi:10.1016/j.procs.2016.07.320
Ch. Jayasree and B. Sravan Kumar / Procedia Computer Science 92 ( 2016 ) 36 41 37 Keywords: Flexible AC Transmission System (FACTS), Krill Herd algorithm (KH), Optimal Power Flow, Unified Power Flow Controller (UPFC), L-index. 1. INTRODUCTION Electrical Power systems networks are extensively interconnected and are driven under much stressed circumstances. Power system instability is playing a major role in the present-day electric market scenario. Power system instability is mainly due to the deficiency of new transmission lines and over usage of existing lines. Therefore, the major factors occurring power system instability are well analyzed and presented [1,2]. Many remedial measures have been proposed and implemented to enhance power system voltage stability. Improved utilization of the existing electrical power system network with the employment of FACTS device has become mandatory [3,4]. Therefore, the only alternate solution for these problems is FACTS devices and this new concept was introduced by Narain G.Hingorani in 1988. Amongst several FACTS devices, UPFC provides greater flexibility in citing new generation and it is very efficient in improving the enhancement of power system instability. It is also flexible in solving optimization problems [5-7]. The paper presents a multi-objective optimization problem consisting of transmission losses, voltage deviation and it has been solved using Krill Herd (KH) Algorithm. Here L-index is used to place the UPFC in a specified location i.e., the generator bus with the highest value of L-index is considered as the weak bus in the entire system. Generation Reallocation of generator buses in the entire power system, with and without UPFC device to reduce voltage deviation and minimization of transmission losses is performed on an IEEE 14-bus system. 2. KRILL HERD ALGORITHM (KH) To solve multi-objective and complex engineering problems of optimization in power systems, Gandomi introduced a newly developed nature-inspired meta-heuristic algorithm namely KRILL HERD (KH) algorithm. Figure 1 shows the Flow chart of Krill Herd Algorithm. The distance between each individual krill and the location of food considering the density of the highest krill in the swarm is the major functionality of the krill movement [8].In this mechanism while searching for highest density of the krill and location of food, all krill individuals step towards the finest possible solution in the search space. By prolonging the algorithm to an n-dimensional space, the generalized fitness function of the KH algorithm (for krill individual) is certified below: = + + (1) The Algorithm for Krill Herd is as follows STEP1: Primarily define the size of the population (s) and iteration ( ). STEP2: Randomly generate the population, where j = 1, 2, 3...S krill individuals. Set the parameters for the following: (maximum iteration number) STEP3: Enumerate the fitness function such that evaluate all krill individuals based on its current position. STEP4: Calculate the motion by considering the three factors which are mentioned below: i) Based on position of other krill individuals. ii) Foraging motion. iii) Physical diffusion. STEP5: If the condition of the optimization problem is not satisfied, then go to step3. STEP6: Enhance the new positions of the individual krill s in the population respectively,
38 Ch. Jayasree and B. Sravan Kumar / Procedia Computer Science 92 ( 2016 ) 36 41 STEP7: If the termination criterion is not satisfied, then go to step3 and repeat the procedure duly. STEP8: If the termination criterion is met, then find the finest possible solution in the search space. Start Primarily compute the size of the population, maximum Iteration and data structures Initialization of parameters Enumerate the fitness function value Calculate the motion for the following three actions respectively i) Induced Motion = + ii) Foraging Motion = + iii) Physical Diffusion = δ NO If all constraints satisfied? YES Update the new position of the krill Is stop criterion reached? YES NO Best solution found YES Stop
Ch. Jayasree and B. Sravan Kumar / Procedia Computer Science 92 ( 2016 ) 36 41 39 3. PROBLEM FORMULATION 3.1. Objective Function Figure 1. Flow chart for Krill Herd Algorithm The objective is to obtain the best possible outcome of UPFC device by minimizing the below mentioned objective function. Therefore, the objective function can be formulated as: Min F = Min (W1* TL + W 2 * VD) (2) 3. 2 Transmission Loses The main objective is to reduce the total transmission losses in the transmission lines respectively. 3.3. Voltage Deviation To attain a standard voltage profile, it is necessary that the voltage deviation should be minimum at all buses. The voltage deviation (VD) can be formulated as: = (4) 3.4. L-index: Based on the equations of the power flow model, Kessel et al [9] developed a voltage stability index model. To determine the distance between the actual position of the system and the desired state, L-index is quantitatively used. The stability of the system characterized by L-index is given by: (5) The limits of L-index lies between the range 0(close to no load) and 1(close to voltage collapse) 4. RESULTS AND DISCUSSION An IEEE 14-bus system consists of (i) Five generator buses (bus numbers: 1,2,3,6 and 8). Out of these buses, bus number 1 is considered as the slack bus and the remaining 2, 3, 6 and 8 buses are considered as generator buses. (ii) Nine load buses (bus numbers: 4, 5, 7,9,10,11,12,13 and 14). (iii)twenty interconnected transmission lines. Here Krill Herd Algorithm based on optimal power flow functionality which is applied for the UPFC device on an IEEE 14-bus respectively. Using MATLAB, an optimal power flow program is written using the Krill Herd Algorithm with UPFC. Basically, the input parameters generated to the KH algorithm is shown in the Table 1. Therefore, to establish the effectiveness of the Krill Herd Algorithm, the obtained results are compared with the Genetic Algorithm respectively. Whereas, the input parameters generated to the Genetic Algorithm is shown in Table 2. Table 1 Parameters of Krill Herd Algorithm S.No Parameters Value 1 Number of krill s(nk) 20 2 Number of runs(nr) 10 3 Number of iterations 50 4 Foraging speed ( 0.02 5 0.005 6 0.01 Table 2 Specification of input parameters of Genetic Algorithm S.No Parameters Quantity 1 Size of the Population 20 2 Maximum no. of Generations 50 3 Crossover Fraction 0.8
40 Ch. Jayasree and B. Sravan Kumar / Procedia Computer Science 92 ( 2016 ) 36 41 4 Migration Fraction 0.2 5 Migration Interval 20 Table 3 L-index for various lines of IEEE 14bus system S.No Bus no L-Index 1 14 0.209 2 13 0.101 3 10 0.0875 4 12 0.0839 5 9 0.0827 6 11 0.0746 7 4 0.0525 8 5 0.0346 9 7 0.0099 Table 4 shows the results for with and without UPFC device for an IEEE 14-bus power system considering the total transmission losses, total real power generation, voltage deviation and optimal objective function values. Considering the proposed Krill Herd Algorithm based optimal power flow solution with and without UPFC, it has been ascertained that the total real power generated is reduced from 267.3 MW without UPFC to 262.9 MW and transmission loss is reduced from 8.35 MW, without UPFC to 3.95 MW with UPFC. By comparing the results of KH with GA. It has been observed that by using GA, the active power generation is reduced from 269.5 MW to 268.1MW and transmission loss is reduced from 9.02MW without UPFC to 7.63MW with UPFC. Table 4. Power flows without UPFC and with UPFC placed between bus no 13 and bus no 14 for 14 bus system Power Flow Solution Total real power generation (MW) Voltage Deviation (p.u.) Total Real Power loss (MW) GA-OPF Without UPFC 269.5 0.902 10.113 5.5082 With UPFC 268.1 0.763 8.767 4.7651 Objective Function Value (p.u.) KH-OPF Without UPFC 267.3 0.836 8.356 4.596 With UPFC 262.9 0.606 3.958 2.282 Table 5 UPFC parameters using KH UPFC placed between bus number 13 and 14 Series converter voltage in p.u 0. 1396704 Series converter angle (degree) -129.6329 Shunt converter voltage in p.u 1.0413505 Shunt converter angle (degree) -12.51194 Table 6 Comparison of Real Power Generation of Generator Busses in Various Methods PV bus NO Generation limits NR GA-OPF KH-OPF KH-OPF with Min Max Method With UPFC With UPFC Without UPFC UPFC 1 10 200 191.90 188.3 115.854 94.018 2 10 50 20.0 20.0 50 47.054 3 10 50 20.0 20.00 18.893 37.054 6 10 50 20.0 17.54 42.872 38.052 8 10 100 20.0 22.32 39.736 46.119
Ch. Jayasree and B. Sravan Kumar / Procedia Computer Science 92 ( 2016 ) 36 41 41 In Table 6, the total real power generation of each single generator (PG1, PG2, PG3, PG6, and PG8) of the system has been compared. Therefore in the reduction of real power generation Krill Herd Algorithm based optimal power flow is most sufficient and effective in use. Generation reallocation has been carried out in an optimal way which results in minimization of transmission loss and voltage deviation in the system, based on the proposed Krill Herd Algorithm. Fig.2 shows the comparison of Voltage profile with and without UPFC. Voltage Magnitude in p.u 1.2 1.1 1 0.9 0.8 0.7 NR Method(Base case) KH-OPF without UPFC KH-OPF with UPFC 0 5 10 15 Bus Number Fig.2.Comparision of the voltage profile with and without UPFC 5. CONCLUSION In this paper, Krill Herd Algorithm is introduced and applied to determine the rating of the FACTS device named UPFC. Therefore, this device satisfies the multi-objective function with equal weight age to minimization of voltage deviation, transmission loses in the power system respectively. L-index is used to identify the weakest bus, critical line in the entire system for optimal location of UPFC. By using simulation of standard IEEE 14-bus, the proposed method has been verified for without placing of UPFC and with placing of UPFC. The results show that by placing of UPFC the transmission loses are reduced. It is also observed that KH is effective optimization method to solve generation reallocation problem as compared to GA. REFERENCES [1] IEEE/CIGRE Joint Task Force on Stability Terms and Definitions, Definition and Classification of Power System Stability, IEEE Transactions on Power Systems, Vol. 5, No. 2, May 2004, pp. 1387 1401. [2] S. C. Savulescu, Real-time Stability in Power Systems, Springer, 2006. [3] N. G. Hingorani and L. Gyugyi, Understanding FACTS: Concepts and Technology of Flexible AC Transmission System, IEEE Press, 2000. [4] Selvarasu, R.,Kalavathi M.S. UPFC placement: A new self adaptive firefly algorithm 2013,Sustainable Energy and Intelligent systems(seiscon 2013),IET Chennai, Fourth International Conference, Pages 204-209. [5] Ghahremani E Optimal placement of multiple-type FACTS devices to maximize power system loadability using a generic graphical user interface 2013 Power systems, IEEE transactions vol-28, Pages 764-768. [6] Sapna Khanchi1, Vijay Kumar Garg, Unified Power Flow Controller (FACTS Device): A Review, International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 4, Jul-Aug 2013, pp.1430-1435. [7] P Pradosh.Kumar. Adhvaryyu, Pranab Kumar Chattopadhyay & Aniruddha Bhattacharjya Application of Bio-Inspired Krill Herd Algorithm to Combined Heat and Power Economic Dispatch. 2014 IEEE innovative smart grid technologies- Asia. [8] Gobind Preet Singh, Abhay Singh, Comparative Study of Krill Herd, Firefly and Cuckoo Search Algorithms for Unimodal and Multimodal Optimization I.J. Intelligent Systems and Applications, 2014, 03, 35-49.