Comprehensive Reactive Coordinated Optimization Based On The Particle Swarm Algorithm

When all parameters and load are given in the power system, the reactive power optimization is the flow distribution to make the power system performance indicators or the objective function model to achieve optimal through choosing reasonable control variables and meeting all the given requirements of constraint equations. In substations of an urban power system, there are shunt capacitor sets connected to the low voltage buses. We can optimize the objective function of urban power system by adjusting transformer taps and changing the capacitor switching states.According to the specific characteristics of city high voltage power systems, we can change the tap positions of on-load transformers to adjust the voltage, and change the switching states of shunt capacitors to reduce the active power loss. For the static optimization, the optimization model is designed to minimize the active power loss. The constraints include two parts: the equality constraints formed by the flow equations; the inequality constraints formed by the shunt capacitors switching frequency, the transformer tap positions and node voltage. For a node reactive sensitivity coefficient in the actual calculations, if the sensitivity value is positive, then the system should reduce the reactive power compensation of the node. If the sensitivity value is negative, then the system should increase the group number of compensation device(that is switching a part of capacitors). Then we’ll adjust the transformer tap positions to meet the voltage quality requirement in power system. Through analyzing the sensitivity of a practical example, we know that this method is fast and the calculation results are reasonable.For comprehensive reactive power optimization, the optimization method is using intelligent optimization algorithm and penalty function method to bring all the inequality constraint equations into the original objective function, which is optimized as a penalty term. However, the reactive optimization goal is not only the minimum economic cost of the objective function throughout the day, but also to ensure that the minimum of the network loss, which also considers the adjustment number and time of the control equipment. In this paper, the optimization program of reactive optimization model is written by FORTRAN language. We will first do static reactive power optimization calculation of the system, and then the optimization results will be used as the initial value to optimize the dynamic reactive power. The example results will give a one-day program of the capacitor switching and transformer tap stalls adjustment and integrated economic cost before and after the system optimization etc. The results show that this algorithm can not only improve the reliability of the flow optimization results, reduce the number of equipment action and the cost of reactive power compensation, but also solve the online reactive power optimization problem efficiently, meet the requirements of its online operation. The convergence performance is good, and the convergence speed is fast, which proves the model is reasonable.