The Reactive Power Optimization Of Distribution Network Containing DFIG-based Wind Supply

As a significant way to supply electric power, wind power generation plays an important role in power system. Comparing to traditional power generation, wind power has the ability of absorbing local electric power and generate electric power effectively and flexibly. Wind power generation could also take the advantage of decreasing environment pollution. Double-fed induction generator owns a small capacity of frequency converter as well as independent control of active and reactive power output. Thus double-fed induction generator is regarded as popular type among megawatt generators. In conditions of wind farms connected in distributed network, the power flow, network loss, bus voltage will be affected. The grid structure and operation states are changed due to instability characteristic of wind power output. Adjusting capacity of reactive power compensator can perform optimal reactive power control of distributed network, decreasing network loss and improving electric power quality. Thus the economic operation of distributed network can be guaranteed. It is significant for double-fed induction generator to perform secure and economic operation of distributed network based on optimal configuration and reasonable control strategies.The paper aims to research the optimal of reactive power and configuration in distributed wind generation which utilize double-fed induction generators.For the reason of instability and randomness of wind power generation and load, traditional certain power flow methods are no longer acceptable to calculate power flow in distributed networks which contain wind power farms. In the paper, operation characteristic, operation mode, reactive power adjusting and reactive power output limitation of double-fed wind power generators are analyzed. Moreover, randomness output of double-fed wind farms which are based on scene strategies and random load fluctuation in small scale are combined to build random power flow calculation model. The novel randomness power flow methods are proposed based on double-fed wind farms in the paper. Therefore, reactive power complement, load fluctuation and random wind power output are concerned in performance testing. The results show that the proposed calculating method can improve accuracy of power flow because it takes randomness of wind power output and load fluctuation into consideration. The proposed method can satisfy the reactive power optimization in distributed network containing wind power generations.Not only the effect of randomness of wind power output should be concerned in optimizing reactive power, but also the effect of load changing to power flow plays an very important role. For comprehensively consideration of wind power output and load fluctuation, objectives including minimum network loss, best static stability and minimum voltage deviation are proposed in the paper. The model of wind power farms containing double-fed induction generators are built according to those objectives. An improved particle swarm optimization method is designed which is suitable to be utilized both in continuous and discrete variable. Capacity of reactive power compensator and continuous wind power output could be able to be optimized simultaneously. Optimization of reactive power can be realized in double-fed wind farms which are simulated in IEEE 33-bus system. The results show that the demand of optimizing reactive power has been satisfied nicely.In optimal configuration of distributed network, the system is usually regarded as a static network. In realistic, the load curve is not considered in the schemes. This is the reason that the electric power demand in different load rate could be hardly satisfy, which causes wasting or lacking capacity of compensators. In the paper, factors such as optimal control of reactive power, load curve, change of wind power output are comprehensively considered in realizing dynamic optimization configuration of wind power farms. Moreover, optimization of reactive power configuration and control are combined to realize accurate configuration of dynamic distributed network as well as optimal control strategies. Thus reactive power optimization configuration becomes more specific and feasible. Experimental results show that the performance of dynamic reactive power optimization configuration schemes is better, which can satisfy the different electric power supply under different load rate. The distributed network, meanwhile, owns better economy of power supply.