飞轮储能系统多PI控制器参数优化

Parameter optimization of FESS PI controllers

对飞轮储能系统(FESS)的有功和无功PI控制器参数优化进行了研究,提出了应用加入模拟退火思想的改进粒子群优化(AIPSO)算法优化FESS有功和无功多PI控制器参数。该算法在混沌初始化、迭代中加入混沌扰动和自适应调整惯性权重系数的改进粒子群优化算法基础上,引入模拟退火思想,既能限制位置更新,又能跳出局部最优解,具有更高的效率。以有功功率和电压偏差的ITAE指标的和最小为目标函数,应用AIPSO优化FESS的多PI控制器参数,并以FESS接入4机系统为例,通过非线性仿真验证了优化结果的有效性。

AIPSO（Annealing-added Improved PSO） algorithm is proposed to optimize the parameters of active and reactive power PI controllers of FESS（Flywheel Energy Storage System）,which,based on the improved PSO（adding the chaos disturbance and adaptive inertia weight factor into its chaotic initialization and iteration）,adopts the concept of simulated annealing to raise its efficiency by limiting the location updating and avoiding the local optimal solution. The parameters of muhi-PI controller in FESS are optimized by AIPSO,which makes the sum of deviation ITAEs of active power and voltage lowest as its objective function. The nonlinear simulation for a four-machine power system with a FESS verifies its effectiveness.