Coordination Optimization Of FACTS Controllers Using Multi-objective Modified Teaching-learning Algorithm

With the increasing comp lexity of power grid and the continuo us development ofheavy-duty power electronic techno logy, the technique of Flexib le AC Transmiss ionSystem (FACTS) has become one of the hottest advanced control methods in powersystem. Under the premise of mainta ining the transient and dyna mic stability, FACTScan greatly enhance the capacity of transmiss ion line and improve the voltage qualityof the system e fficie ntly. Besides FACTS provides an effective means to achieve thesmallest environmenta l pressures, the shortest construction period and the optima leconomic benefit. However, with the large-scale application o f FACTS controllers, apotentia l risk, the negative interaction between FACTS controllers, puts forward newrequirements on power system operation and control. Therefore, the study oncoordinatio n control technolo gy of FACTS controllers is of theory and engineeringvalue.At present, the interaction and coordination design of FACTS controllers is on theeleme ntary stage, and there still exist ma ny proble ms to be explored thoroughly. Basedon the dyna mical model of a large-scale multi-machine power system installed withmultiple FATCS devices, this paper analyses and validates undesirable impacts ofinteractio ns among FATCS controllers in the power system. On this basis, aMulti-Objective Modified Teaching-Learning Algorithm (MOMTLA) is presented tocoordinate the FACTS contro ller parameters to obtain a better comb ined performance,in whic h the static coordination of TCSC and SVC controllers is formulated as amulti-objective optimization proble m. Focusing on this issue，the dissertation, basedupon the work of previous lecturer, cons ists of three main parts as follows:In the first part, a novel Multi-Objective Modified Teaching-Learning Algorithm(MOMTLA) is proposed to solve the defic ienc y of TLA in dealing with multi-objectiveoptimizatio n proble ms. Compared with conventiona l TLA, an improved learner phasein teams is presented and comes with a locked device phase to improve the globalsearching ability and ideal convergence. Then several meta-heuristic techniques areapplied to make a satis factory multi-objective optimization method in MOMTLA.Simulation results in both test functions de monstrate the valid ity of MOMTLA.In the second part, a new coordination method of FACTS controllers based onMOMLTA is presented. The multi-objective optimization model in cons ideratio n ofSVC and TCSC is formulated to reduce the interactio ns between controllers. Then, MOMTLA is emp loyed to search optima l Pareto solution of controller parameters. Theproposed method is applied and validated in a typical IEEE multi-machine powersystem with SVC and TCSC. Simulation results demonstrate that FACTS controllersdesigned optima lly provide a better comb ined operation performa nce and significantlyimprove the system stability.