The Study Of Genetic Aglorithm In Power Network Reactive Power Optimization

Practical mathematical model involves all kinds of reactive power regulation facilities and operational constr aints on the base of reading a lot of references are presented, After application of GA-based reactive power optimization techniques.Due to GA's incompletion,Some improvements are presented about genetic algorithm, As to the conflict of GA generate new individuals randomly, the author proposed a neighbourhood search strategy based on similitude. Neighbourhood search bring the ability to self-adaptively generate new individuals easily. Aiming at balancing search results and search speed, we proposed a search strategy to classify the indivi duals by its fitness.By individuals classfication to differentiate respective function in search process, that's the excellent indivi dual to mine the local optimal solution and others to explore the search domain to find new local optimal solution. Aiming at GA bad efficiency, uniting traditional algorithm and numerical analysis's speed up thought, we proposed a speed up search strategy by combining parent individuls information. The search efficiency is highly improv ed. To improve the convergence speed, with simulate annealing, we design a neigbourhood contractive technical. The search speed is enhanced. According to above search strategies, united GA characteris tics, we proposed a new algorithm frame SFEC (Similitude Frame of Evolu tionary Computation). A reactive power optimization application program is developed on the base of proposed mathematical model combined with tradition GA and improved GA by VC++. This program is applied to a small simulated power network in the reactive power optimization, The optimization results indicate that the SEFC and program presented in this thesis is feasible,stead and quick.