Optimization Study, Based On The Overall Sensitivity Of Reactive Power Planning

Reactive power planning was an important aspect in the power system planning. It was related with the safe stable operation and the economical running of the entire system. Reactive power planning optimization included two aspects:to determine reactive power compensation nodes and the compensation capacity of every node. The effect of reactive power compensation was close related to the installation location, so this article studied the method of choice candidate nodes that were combined with sensitivity guide line and grid divisions.Because it was requested to stratification and zoning in the voltage/reactive power optimization control. Reactive power compensation should be on the spot balance. When it chose the reactive power configuration node, grid was zoned first, and then selected parts of load node in every district for reactive power compensation. So it also avoided the shortcoming that was easily becoming adjacent nodes according sensitivity to choice.Reactive power compensation had close relationship with network loss and node voltage, which respectively reflected the economic and secure value of reactive power compensation. On the basis of power grid zoning, according to the load flow equations, this article derived the loss sensitivity and the system voltage sensitivity that were caused by the reactive power change of load nodes. According to the two sensitivities, this article defined the comprehensive sensitivity. Then, it chose the load nodes with great comprehensive sensitivity as candidate nodes at every district. So it achieved the aim that to install reactive power supplies on the candidate nodes were effective to reduce the loss and improve the level of nodes voltage. At the same time the reactive power supply configurations were hierarchical and partitions, as well as the selected nodes were not adjacent nodes. It also avoided the reactive power long-distance flow in the grid.Different objective function had different compensation project. This paper established the objective function of minimum expenditure at one year. It applied the method of "Power Division-Comprehensive sensitivity" to elect the candidate nodes of reactive power compensation as the initial solution space. So it narrowed the scope of the search and improves the speed of optimization. Then this paper used genetic algorithms to solve the objective function. The result was the best reactive power compensation nodes and the capacity of every compensation node. The article also established the mathematical model and the solution process of reactive power planning that considered the different years in planning period, different requirements of reactive power compensation and multi-variety cases of load. It also gave the mathematical model of testing accident.Finally, As IEEE39 bus system for example, this paper used the proposed method to make reactive power optimization planning. The result showed that the method was accuracy and validity.