粒子群算法论文：改进粒子群算法在火电厂主汽温控制系统中的应用.docVIP

粒子群算法论文：改进粒子群算法在火电厂主汽温控制系统中的应用 【中文摘要】在现代火力发电厂中,主蒸汽温度的稳定对机组的安全经济运行有重大的影响,所以对主蒸汽温度的控制要求很高,要对其进行有效合理的控制。PID控制在电厂主汽温控制中仍占有主导地位,但是主汽温对象具有大惯性、大延迟、非线性和时变特性,采用常规的PID控制存在着不足。粒子群优化算法(PSO)是一种启发式群体智能优化进化算法,通过模拟鸟类的觅食行为,将寻找具体问题的最优解转化为粒子在特定空间内的迭代搜索过程,本文将其应用于主汽温控制系统PID参数优化中。本文首先介绍了PID控制器的基本原理、参数整定方法和控制系统性能评价指标;接着介绍了粒子群算法的原理及常见的改进方法,针对粒子群算法存在早熟收敛和后期收敛速度慢,易陷入局部最优的缺点,引入了具有量子行为的粒子群算法,并借鉴遗传算法中的变异机制,以自适应变异概率对个体进行变异,来增加种群的多样性,由此产生一种改进的PSO算法,并通过对几个标准测试函数的优化,证明了改进PSO算法的收敛速度快、精度高;然后将改进的PSO算法应用到PID控制器的参数优化整定过程中,仿真结果验证了该算法的有效性;最后针对主汽温串级控制系统,采用改的PSO算法对其PID参数进行优化,仿真结果表明基于改进的PSO算法的主汽温控制系统获得了很好的调节品质和较强的鲁棒性。 【英文摘要】In modern thermal power plant,stability of main steam temperature is very important to security and economy of unit operation.Therefore, It is very strict to the control of main steam temperamre , and it is important to get a reasonable and effective control in the themal power plant. PID controller is still the most primary control strategy in the control of main steam temperamre of power plant. But main steam temperature is a complex object with the features of large inertia, large delays,non-linear and varity time, it has some shortcomings to control the main stream by using conventional PID controller. Particle Swarm Optimization(PSO)is a kind of group intelligent heuristic evolutional algorithm, it transfers the process of seeking the best solution for specific problems to the process which particles scout the best position in a certain area by simulating the foraging behavior of birds. So this thesis uses PSO to optimize PID controller抯 parameters of the control of main steam temperamre.This thesis first introduces the principles of PID controller, parameter tuning methods and control system performance evaluation; then introduces PSO algorithm principle and the mainly involved method improving the PSO algorithm. PSO has some defects such as following:premature convergence, slow convergence rate in the optimization process later, and easlly turning into local optimization. In respo