Reactive Power Optimization Of Oilfield Distribution Network Based On Improved Genetic Algorithm

The load specificity and improper reactive power compensation appeared in the distribution network of Daqing Oilfield leads to low power factor, low power quality and serious power loss on distributing line,the direct result of which is the energy waste.The reactive power optimization of a radial distribution network can obviously reduce the operating power losses and enhance voltage qualification rate,and then the annual operating costs will be decreased as well as quality of power energy can be improved. Therefore,it is very important that reseaching on the reactive power optimization problem of oilfield distribution network from the Daqing Oilfield distribution network operating conditions in order to improve the distribution network operation economics and reduce oil production costs,which can protect the sustainable development of Daqing Oilfield, and create a century oilfield for the national oil strategic security.Power flow method of a distribution system is the base and tool of reactive power optimization, whose computational efficiency and convergence speed will impact directly on the results of optimization.This paper systematically studys the common power flow mothod of a distribution network, and derive the forward/backward sweep method according to oilfield distribution network lines long, with many points, closed-loop structure,open-loop operation.This method is suitable for oilfield distribution network power flow calculation and has the merits of simpleness, high computational efficiency, memory save and good convergence.The thesis has conducted a study on the reactive power compensation principle, compensation method,compensation mode,and made a compensation configuration of oilfield distribution network reactive power compensation.In order to determine the locations of the reactive power compensation candidate points and improve the convergence speed of optimization,it uses sensitivity analysis.The established mathematical model which is suitable for realistic situation of a distribution network is crucial to execute the reactive power optimization.In this paper, considering the voltage quality and various constraints,the objective function is set based on minimize the line loss, punished functions is the node voltage of system which is step beyond the boundary. This paper carry out reactive power optimization calculation based on above mathematical model by Genetic Algorithm.To overcome the poorly computational efficiency of Simple Genetic Algorothm,we also makes an improvement of it from the facts of coding,fitness function,select operator,crossover operator,mutation operator and evolution end condition,to enhance its convergence speed and ability of global optimization.According to The above-mentioned mathematical model and algorithm, a computer application program of oilfield distribution reactive power optimization is developed based on delphi platform,and then we use it to calculate a 6.6kV pump recovery well distribution line of a station of Daqing Oilfield. All the results given in the paper proof that the improved genetic algorithm not only greatly improve operational efficiency, but also ensure the accuracy of the operation,and its performance is obviously better than Simple Genetic Algorothm. So the improved algorithm is reasonable and effective, it has some practical value to use it on the application of oilfield distribution reactive power optimization.