Research On Dimension Reduction Of Correction Equations And Distribution Of Loss Reduction For Reactive Power Optimization

Reactive power optimization plays an important role in power systems and it's an effective measure to keep system in safety and economic operation state. Reactive power optimization is a ramification of optimization power flow (OPF), which consists of optimization of reactive power planning and operation. This paper focuses on the optimization of reactive power operation, which means to study on how to reduce network loss effectively with various reactive power regulatory measures when all the reactive power compensation devices and their adjustment ranges have been identified.The active power output of all the nodes is constant in conventional reactive power optimization methods, except the slack node. Setting multiple generator nodes to participate in reactive power optimization can get better reducing loss effect than the former measures. The active power output of all the aforementioned nodes (named generalized slack nodes in this paper) can be adjustable. To make up for the shortage of conventional optimization methods, this paper built up mathematical model of reactive power optimization with multiple generalized slack nodes, and solve it with embedded quadratic penalty function of nonlinear primary-dual interior point algorithm.It usually takes most time of the calculation process to solve the correction equations. According to the structural characteristics of correction equations derived from nonlinear primal-dual interior point algorithm, this paper obtains the minimum solution structure by selecting appropriate variables and reordering them, whose dimension is only related to the number of nodes. The reduced equations can speed up the calculation of reactive power optimization.The algorithm proposed in this paper, which is derived from power flow expressions, can calculate the incremental active transmission loss for each node in the system including the slack node. Combined with the physical meanings of incremental transmission loss, this paper figures out a method to identify the location of generalized slack nodes, that is to descend all the generator nodes according to their sizes of incremental transmission loss, and select the first m generator nodes as generalized slack nodes to participate in reactive power optimization.Finally, calculation results of different sized examples verify the validity and the effectiveness of the reactive power optimization model, the measures to reduce dimensions and the method to determine the location of generalized slack nodes proposed in this paper.