基于PSO-RBF神经网络的微电网等效建模

MICRO-GRID DYNAMIC EQUIVALENT MODELING BASED ON PARTICLE SWARM OPTIMIZATION ALGORITHM AND RBF ARTIFICIAL NEURAL NETWORK

提出一种改进粒子群优化的RBF神经网络微电网动态等效模型及建模方法,利用RBF人工神经网络的非线性映射特性解决微电网系统并网接入的等效建模问题。基于微电网公共接入点(PCC)的电压、电流、功率等量测数据构建RBF神经网络等效模型,将接入点电压和电流分别作为神经网络的输入和输出,使神经网络的输入输出更具独立性。将混沌优化的全局遍历性引入粒子群优化算法中,构建基于全局最优解的变邻域混沌搜索提高粒子群算法的全局搜索能力,利用改进粒子群算法优化RBF神经网络模型参数提高模型计算精度。最后通过微电网并网仿真实验验证本文提出等效模型的准确性和建模方法的合理性。

Based on particle swarm optimization algorithm （PSO） and Radial Basis Function （RBF）neural network, a micro-grid equivalent modeling method was proposed in this paper. Nonlinear mapping characteristics of RBF neural network is used for the dynamic modeling of micro-grid. Voltage, current and power of the Point of Common Coupling （PCC） are collected for the input and output of the RBF neural network. The ergodicity of chaos algorithm is introduced to accelerate the convergence of PSO algorithm. Chaos variable neighborhood searching method operates around the global optimization of PSO which can improve the global searching ability of PSO algorithm. The improved PSO algorithm is used for the parameter optimization of RBF. Simulation results showed that the proposed modeling method is suitable and effective, the RBF neural network based dynamic equivalent model can descript the dynamic characteristics of microgrid in connected mode.