论文部分内容阅读
为了获得精确的静电放电模型,提出了一种应用粒子群优化算法的静电放电模型参数辨识新方法。以Heidler雷电流方程的静电放电模型参数为辨识对象,分别以仿真及实验数据验证了该方法的可行性,并从电流波形的整体和局部两方面对拟合效果进行了评估。结果表明,与遗传算法相比,粒子群优化方法的执行速度更快,所得的辨识参数精度更高,粒子群优化方法对电流波形的整体和局部关键点的拟合度均高于遗传算法。因此,粒子群算法较遗传算法更适用于解决静电放电模型参数辨识问题。此外,从实例可以看出,粒子群算法不需要过多的初始参数值先验知识,而只须提供一个较宽的初始参数搜索范围即可获得良好的辨识结果。
In order to get an accurate electrostatic discharge model, a new method of electrostatic discharge model parameter identification using particle swarm optimization algorithm is proposed. The parameters of electrostatic discharge model of Heidler’s lightning current equation are identified, and the feasibility of this method is verified by simulation and experimental data respectively. The fitting effect is evaluated from both the global and local current waveforms. The results show that Particle Swarm Optimization (PSO) performs faster and has higher accuracy than the genetic algorithm. Particle swarm optimization (PSO) is superior to genetic algorithm in the fit of global and local key points of the current waveform. Therefore, Particle Swarm Optimization is more suitable than genetic algorithm to solve the parameter identification of electrostatic discharge model. In addition, it can be seen from the examples that the PSO does not need too much priori knowledge of the initial parameter values, and good identification results can be obtained by only providing a wider search range of the initial parameters.