Distribution System Of Reactive Power Optimization Based On Improved Artificial Fish Algorithm

Optimization of reactive power is an important element of the safe and economic operation of the power system. In this thesis, a research has been made on the reactive power distribution network optimization based on the improved artificial fish swarm algorithm, and achieved satisfactory results.The artificial fish swarm algorithm,which is combined with the mutation operator and simulated annealing algorithm is researched. We have improved the mutation operator and presented a method to tentative determining of mutation probability. This method solves the stagnation of variation behavior and the circumstance of falling into a optimal local solution in original algorithm, avoids the damage of structure and convergence, and improves the accuracy of original algorithm.The economic objective function and security objective function is combined to improve the established optimization models of reactive power distribution network. In economic objective function currency form which represents investment and income is taken as the evaluation index instead of minimum net loss. And in security objective function, as best as a voltage level of security objective function, taking into account the distribution network security indicators.The artificial fish algorithm which combined the impoved mutation operator and simulated annealing algorithm is applied to the reactive power optimization. We have presented the solution process and implement it with C++ program developed.The high-voltage and low-voltage distribution network in a city is taken as an example to calculate The result shows that the improved algorithm for reactive power distribution network optimization can obtain a higher precision solution and holds the merit of fast convergence with a comprehensive search of solution space to achieve the optimal solution. And the improved model is more valuable in practical application than the original one.