ارائه كنترل كننده تطبيقي با استفاده از منطق فازي و الگوريتم ازدحام ذرات

چكيده انگليسي :

۶۲ Proposing an Adaptive Controller with the Fuzzy Logic and Particle Swarm Optimization Algorithm Poya Babahajyni p babahajyani@ec iut ac ir Date of Submission 2012 02 19 Department of Electrical and Computer Engineering Isfahan University of Technology Isfahan 84156 83111 Iran Degree M Sc Language FarsiSupervisors Farid Sheikholeslam sheikh@cc iut ac ir Saeed Hosseinia hoseinia@cc iut ac irAbstract In designing PI PID controllers with classical methods the systems are considered in nominalwork zone Therefore the controllers are not able to adapt themselves with the unstable conditionsof the systems Thus the performance of the controller is considerably reduced with the changes inparameters and the usage conditions of the systems One way to improve the performance of thePI PID controllers is to use a fuzzy system for tuning the parameters of these controllers in realtime Fuzzy controllers when dealing with complicated systems having high volumesof inaccuracy and uncertainty operates very well because these systems are not based onmathematical relationships and are independent of the under control system model In a sense fuzzy method is not based on a concrete model The combination of fuzzy controllers and PI PIDcontrollers means that the PI controller not only has non linear features but also the features of thefuzzy controller simultaneously In this way the fuzzy controller adjusts the parameters of the PIcontroller in real time and based on the conditions of the main system to improve the output of thesystem But the performance of this two level controller completely depends on the membershipfunction of the fuzzy system In practice the membership function is configured based onknowledge and experience and accordingly this knowledge and experience is obtained throughtrial and error and will not cover the unpredicted states So in order to overcome this problem usingan optimization algorithm for adjusting the parameters of the membership function of the fuzzysystem in real time and simultaneously with the occurring changes in the system will be a greatidea But because the performance of evolutionary algorithms especially PSO algorithm generallyis based on population particles these algorithms are used in offline mode So using thesealgorithms in real time causes some problems in calculating the cost function Modern power systems require increased intelligence and flexibility in the control andoptimization to ensure the capability of maintaining a generation load balance following seriousdisturbances This issue is becoming more significant today due to the increasing number ofMicrogrids The Microgrids mostly use renewable energies in electrical power production that arevarying naturally These changes and usual uncertainties in power systems cause the classiccontrollers to be unable to provide a proper performance over a wide range of operating conditions In response to this challenge the present thesis addresses a new online intelligent approach byusing a combination of the fuzzy logic and the particle swarm optimization techniques for optimaltuning of the most popular existing proportional integral PI based frequency controllers in the acMicrogrids systems The proposed intelligent PSO fuzzy PI control design methodology is usedfor secondary frequency control in an ac Microgrid To demonstrate the effectiveness of theproposed control schemes the result is compared with the pure fuzzy PI control method as well asclassical PI control design using Ziegler Nichols technique In the developed tuning algorithm thephysical and engineering aspects of Microgrid systems have been considered Simulation studiesare performed to illustrate the capability of the proposed intelligent control approach Keywords Fuzzy Systems Evolutionary Algorithms Frequency Control Microgrid

باباحاجياني، پويا

ارائه كنترل كننده تطبيقي با استفاده از منطق فازي و الگوريتم ازدحام ذرات

سيستم هاي فازي , الگوريتم هاي تكاملي , كنترل فركانس , مايكروگريد