基于改进的LDW粒子群算法的风-火电力系统联合优化调度策略

Combined optimization dispatch strategy of wind-thermal power system based on improved LDW particle swarm algorithm

风-火电力系统联合优化调度是一个极其复杂的NP问题,不易求解。改进粒子群算法,并将其应用于风-火电力系统联合优化调度,提出了一种改进的惯性权重线性递减的粒子群算法。针对粒子群算法容易局部收敛的缺陷。首先,本文在惯性权重线性递减(LDW)的基础上,加入常数扰动,使惯性权重大幅增大,以便于跳出局部搜索,进行全局搜索,从而防止局部收敛;其次,为尽可能的避免粒子群算法出现粒子高度聚集在最优粒子的周围的情况,使得粒子趋于相同以致于大大损失粒子群的多样性,一定概率的自适应的改变惯性权重并混入随机个体,以便于更好的保持种群多样性。最后,在Matlab2010a GUI平台下采用几种不同的粒子群算法进行仿真试验。仿真结果表明,在相同条件下改进的粒子群算法能够寻到更精确的解。

Combined optimization dispatch of wind-thermal power system is an extremely complex NP problem, which is difficult to solve. In order to solve the problem in a relatively short time, this paper improve on the particle swarm optimization （PSO） algorithm and applies it to optimal dispatch of wind-thermal power system. An improved PSO algorithm with linearly decreasing inertia weight （LDW） is proposed, and several measures to correct the partial convergence problem embedded in PSO algorism. Firstly, based on the LDW algorithm, constant perturbation is added to increase LDW, so as to jump out of the local search for a global search, thus preventing loeal convergence. Secondly, in order to avoid a loss of diversity in the particle group caused by particles gathering up around the optimal particles and assimilating, we propose a self-adapting algorithm that maintains population diversity through changing the weights and adding random individuals. Finally, results from several different PSO algorithm simulations conducted on Matlab2010a GUI platform show that eeteris paribus, our improved PSO algorithm finds more accurate solutions comparing to all existing algorithms.