CAJ | 학술논문

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.
위료획득정학적정전방전모형,제출료일충응용입자군우화산법적정전방전모형삼수변식신방법.이Heidler뇌전류방정적정전방전모형삼수위변식대상,분별이방진급실험수거험증료해방법적가행성,병종전류파형적정체화국부량방면대의합효과진행료평고.결과표명,여유전산법상비,입자군우화방법적집행속도경쾌,소득적변식삼수정도경고,입자군우화방법대전류파형적정체화국부관건점적의합도균고우유전산법.인차,입자군산법교유전산법경괄용우해결정전방전모형삼수변식문제.차외,종실례가이간출,입자군산법불수요과다적초시삼수치선험지식,이지수제공일개교관적초시삼수수색범위즉가획득량호적변식결과.
In order to obtain a precise mathematical model of electrostatic discharge (ESD), we put forward a new approach for determining the ESD model parameters by means of particle swarm optimization algorithm (PSO).The parameters of an ESD model based on Heidler equation are selected as the identification objects.The validity of this approach is confirmed with experimental and simulated current data.The fitting effect is also evaluated according to the overall and partial of the current waveform.Compared with genetic algorithm (GA), the PSO approach runs faster, obtains more precise parameters of the ESD model, and fits better either on the overall or at the partial key points of the current waveform.Therefore, the PSO outperforms the GA in solving the parameter identification problem of ESD model.Moreover, as can be seen from the examples, the PSO approach does not require too much prior knowledge of the initial parameter values but to provide a wide search range.