Reactive Power Optimization Based On Hierarchical Clustering And Ant Colony Optimization Algorithm

In power system, reactive power optimization can effectively reduce network losses, improve the quality of voltage, and make the system security, economic and high-quanlity operation. This paper do some research on reactive power planning of distribution and dynamic reactive power optimization of grid system, the main works are as follows:①According to reactive power optimization problem,a reactive optimization model that make network active power loss to be minimumed as the objective function.The optimal network loss differental rate rule is introduced in the model to assure the compensation buses automatically which is reasonable in economica in the process of optimization.②A new hierarchical clustering algorithm is proposed to analyse the buses sensitivity and to select the key buses of compensation.According to the disadvantage of hierarchical clustering, the new aggregation and separation rule are proposed to eliminate the clusters which do not meet the requirement or need to make sure the number of clusters by people. The algorithm is not affected by random or human disturbance.③On the basis of analyzing the basic principles of ant colony optimization algorithm(ACOA) and TSP problem solved by ACOA,make some mend to it.A new state transition probabilily rule that put the sensitivity of the candidate buses as the visibility factor is brought forward and the ant colony optimization search strategy is improved.④In order to extend the life of capacitor and OLTC ,this paper unification considered the distribution changes of system's load rate,the conception is proposed, which in the time section discrete control equipment can't be adjust,that make sure the movement number of times of equipment meet the request.⑤Cluster ant colony optimization algorithm is proposed ,which the cluster algorithm and ant colony optimization algorithm are united,to guarantees the algorithm optimization capability。The concept of way multiple is also be proposed to make sure the algorithm can find the best solutionA Matlab program based on the optimization algorithms are implemented, IEEE standard systems are calculated and analyzed.The results show the feasibility, accuracy and effectiveness of the method and the progam in this paper.