Voltage And Reactive Power Optimization Algorithm, A Comparative Study

Based on an improved genetic algorithm (IGA),a primal-dual interior point method(PDIPM),a smoothing Newton method and a branch-and-branch method, this paper uses four algorithms and their combined optimization algorithms for effectively solving the optimal reactive power flow problem(ORPF). In these algorithms, the ORPF problem is firstly converted to an integer nonlinear programming problem with continuous control variables, and then discrete control variables is locked. The initial solutions are obtained by PDIPM or SNM. The other situation is that the ORPF problem is transferred into a nonlinear programming problem with discrete control variables, which are solved by the other algorithm (for example: IGA, BBM) and continuous control variables are dealt with by PDIPM. To enhance the computational efficiency of the combined method, constraint -handling technique, improved crossover operator and mutation operator are incorporated into the traditional GA.Numerical simulation results on the IEEE 30 test system and IEEE118 test system show that each combined algorithm has its fit and unfit qualities. Then in this paper, the author obtains the conclusion from the time of convergence, the iterative times and the result of optimization.These combined methods actually used in Fujian province power grid automatic voltage control (AVC) system have proved to be very effective.