| The problem of traditional reactive power optimization is static reactive power optimization under the circumstance of special load. But the load is changing in any time in the practical power system, so the traditional reactive power optimization can't meet the practical need. Reactive power optimization considering the load changing of the node in the system is also called as dynamic reactive power optimization. Dynamic reactive power optimization generally study the dispatch schedule for the capacitor bank switching, the reactor switching and on-load tap changer(OLTC) position for 24 hours in a whole day based on the load forecasted, to reduce the whole power loss whole satisfying the constraints on bus voltage limits and number of the operations.The paper study the difficulty of the dynamic reactive power optimization based on the dynamic reactive power optimization model constructed. Consider that the dynamic programming solving the dynamic reactive power optimization in theory is not fit for the practical power system, considering the limit of the operation times and the load changing the paper present a practical algorithm to solve the dynamic reactive power optimization based on the time-interval control, using the improved GA to optimize the discrete control variable and the predictor-corrector interior point method to the smooth control variable. On the basis of traditional genetic algorithm, the holding of population diversity, the GA operator are improved to fit the problem of the reactive power optimization, Predictor-corrector interior point method is also used to optimize the smooth control variable in the reactive power optimization.To demonstrate the effectiveness of the proposed approach, the IEEE-30 node system base on the FuZhou Forecasted load curve is built to simulate the dynamic reactive power optimization, It is concluded that a greater loss reduction can be achieved by the proposed method , and the optimal schedule is simple.
|
PDF Full Text Request