| With the improvement and development of society, the structure of feeder has become more complex. The reactive power optimization of feeder is an important measure to increase the feeder power supply capacity and ensure feeder economic operation. Through configuring the equipment of reactive power compensation properly in the feeder, feeder voltage qualification can be improved and reactive loss can be effectively reduced. The traditional method of reactive power allocation is carried out under the supposing ideal conditions so that the allocation is unreasonable and the error is bigger. Therefore, to research the method of reactive power allocation under the precision model has not only the theoretical significance but also the practical value.Firstly, the characteristics of load are summarized. Secondly, the reactive power compensation is researched on this basis. By the classic method of reactive power compensation, the maximum capacity of the compensation and the packet capacity of the capacitor are calculated. For a feeder, the mathematical model of reactive power optimization is established. In the model, the objective is that the summation of the line loss and the invested costs of reactive power compensation is least, and the constraints of the current, the voltage and the compensatory capacity are considered.The question is solved by the ant colony algorithm. The function of fitness is solved by BX method, and the result of optimization will be directly affected by the calculating speed and the convergence. This program of trend and line loss is written by BX method, and it is proved to be Correct by an example.On the basis of this, ant colony algorithm is used to calculate optimization model. Sufficiency function compution of optimization need to use the line loss. According to the characteristic of feeder line and optimization, BX power flow calculation is used in solving line loss of sufficiency function.Finally, by the method created in this paper, the program of reactive power optimization based on ant colony algorithm is developed in the feeder. As an example to the 10kV feeder of 28 nodes and 33 nodes, the reactive power compensation is calculated. Through the analysis of the computing result, the optimization model and algorithm are proved to be effective. |