Reactive Power Optimization Incorporating Unified Power Flow Controller

The safety and economic operation problem has increased with power systemdevelopment. Reactive power optimization (RPO) is one of the effective means toguarantee the power system operate economically and safely, so has been widely used andresearch. Flexible AC transmission system (FACTS), especially unified power flowcontroller (UPFC), could be used to control the voltage, phase shift and impedance insuccession, to achieve the purpose of precise control active and reactive power. RPO is atypical nonlinear mixed-integer programming problem with multi-vitiate andmulti-constraints, when consider UPFC later, this problem becomes more complicated. Sothe research of reactive power optimization problem incorporating UPFC has greatsignificance in theory and practice.Based on the currently research on domestic and international literatures of reactivepower optimization, this paper made a research into the application of genetic algorithmwith RPO problem. The main works are as follows:①For the shortage of common genetic algorithm, the combination method of DNAFuzzy Genetic Algorithm is presented to handle RPO problem. The proposed methodused DNA bases to code the optimized variables and introduces some genetic operation,such as the bases replace (mutation) and inversion, to avoid the problem of binary codinggenetic algorithm has low accuracy of convergence. The crossover, mutation andinversion rates are adjusted with online fuzzy control according to the evolution situation,to improve the convergence speed. This method is implemented and applied into IEEEtest system and a practical network. The calculation results indicate that DNA-FGA canimprove the searching efficiency and optimal performance.②For effective use the regulation ability of UPFC to achieve the power systemreactive power optimization, the multi-objective reactive power optimization model withUPFC is proposed. In order to simplify the calculation program, based on power injectionmodel, the UPFC can embed reactive power optimization model with the case that do notmodify the node admittance matrix. In addition to reflect the influence of UPFC onvoltage stability, the objective functions also consider maximum voltage stability marginbesides minimum active power loss. The control variables contain UPFC controlparameters and the constraints include operating constraints of UPFC, except foroperating constraints of AC system. ③the conventional multi-objective optimization was based on weighted method, sohas the difficulties in unity of the dimensional and weights of the setting, thus for solvingthe RPO problem with UPFC, a fast and elitist non-dominated sorting geneticalgorithm-II (NSGA-II) is proposed. In this method, the DNA coding mode and the DNAgenetic operation were introduced in order to verify the adaptability of the DNA codemethod. By improving the fast non-dominated sorting approach, NSGA-II algorithm candeal with the divergence solution of power flow calculation effectively and maintenancethe convergence of algorithm. After obtaining the Pareto optimal solutions, the analytichierarchy process (AHP) was adopted to set each objective function’s weight and rankthese solutions to determine the best solution of reactive power optimization. This methodis implemented and applied into IEEE test system. The calculation results indicate that thepresent model and algorithms are effective and feasible.