Research On The Dynamic Reactive Power Optimization Of Oil Field Distribution Network

Currently，the exploitation of onshore oil fields in China has been in later period. Mostof remote and dispersed wells and unreasonable settled distribution lines lead to the difficultyof exploitation in the oil field and energy loss in the distribution network. So, this paper takesadvantage of reactive power optimization to reduce loss. The static reactive poweroptimization is an optimization which aims at fixed load in a particular moment. It is can’tmeet the needs of optimization because of the changing of load over time. Therefore,dynamic reactive power optimization which considers the motion numbers of equipments andchanges of loads, plays an important role in theoretical research and in practice.Firstly, this paper analyzes the present situation of oil field power grid, the influences ofreactive power, and the methods of reactive power compensation in distribution network;Secondly, this paper summarizes the optimization algorithms, improves the pheromoneupdate rules in ant colony algorithm due to the disadvantages of convergence speed and localoptimal value, and treats agents in the multi-agent system as the ants in the ant colonyalgorithm. At the same time, the improved algorithm was applied to IEEE-30node systemand distribution network with14nodes. The result verifies the accuracy of the algorithm. Setup the objective function and constraint conditions in the process of dynamic optimization,and divide the load curve into segments. The objective function is the minimum of networkloss and compensation capacity in the limit of the voltage, and the constraints includes themaximum numbers of compensation equipments and so on. In addition, the paper improvesthe segmentation technology on the basis of further researching the advantages anddisadvantages of the existing load curve segmentation methods. Normalize the reactive loadin the whole day and compare the difference of average and original data with threshold valuefor segmentation again. The results show that the improved method is superior to the existingsegmentation methods on the computing scale.Finally, apply the improved segmentation method and improved multi-agent ant colonyalgorithm which are put forward in this paper to dynamic reactive power optimization indistribution network. The experimental results verify the effectiveness of improved algorithmand the feasibility of improved segmentation method.