自适应差分进化算法在电力系统无功优化中的应用

An Adaptive Differential Evolution Algorithm and Its Application in Reactive Power Optimization of Power System

电力系统的无功优化是一个复杂的组合非线性优化问题。它通过调节机端电压、变压器分接头、并联电容器来减小网损并且保持良好的电压水平。文中提出了一种自适应差分进化(self-adaptive differential evolution,SADE)方法。在SADE算法中,3个控制参数(包括变异参数F、交叉参数CR、种群数量NP)和变异策略都是根据以往进化经验逐渐自适应的。IEEE30节点系统算例验证了文中所提算法比粒子群算法和标准差分进化算法的网损显著减小,同时在限制范围内保证了良好的电压波形,且计算精度高,有很强的鲁棒性。

Reactive power optimization of power system is a complex combined nonlinear optimization problem. Network loss reduction can be implemented by reactive power optimization including adjusting generator terminal voltage, changing transformer taps and connecting shunt capacitors while the network voltage can be kept at a satisfied level. In this paper an adaptive differential evolution （ADE） algorithm is proposed, in which three control parameters, i.e., mutation parameter F, cross parameter CR and numbers of population Np, as well as mutation strategy increasingly become adaptive according former evolution experiences. Simulation results of IEEE 30-bus system show that by use of the proposed algorithm the network loss reduction is more evident than those by particle swarm optimization （PSO） algorithm or standard differential evolution algorithm, meanwhile satisfied voltage waveform can be kept within the restrained range. Calculation results by the proposed algorithm are accurate and this algorithm possesses strong robustness.