采用改进型SOS算法的光伏组件模型参数辨识

针对当前大部分光伏（photovoltaic，PV）模型参数辨识算法均存在准确性低和可靠性差等问题，提出了一种采用改进型共生生物搜索算法（symbiotic organisms search，SOS）的光伏组件模型参数辨识方法。首先，为提高标准SOS算法的寻优性能，提出了新的改进型SOS算法，记作ImSOS算法。该算法在标准SOS算法的生物种群初始化阶段采用了准反射学习机制；在互利共生搜索阶段采用了改进受益因子策略；在偏利共生搜索阶段采用了收缩随机数产生因子区间策略。其次，给出了采用ImSOS算法求解基于实验测量电流—电压（I-V）数据的光伏组件模型参数辨识问题的具体步骤及实现流程。最后，利用实际Sharp ND-R250A5光伏组件进行实验，通过与标准SOS算法以及其他七种新颖智能优化算法进行对比验证，结果表明了ImSOS算法在光伏组件模型参数辨识的有效性和优越性。可见ImSOS算法为准确可靠地辨识光伏组件模型参数提供了一种新的有效方法。

Parameter identification of photovoltaic module models using improved symbiotic organisms search algorithm

To solve the disadvantages of the most photovoltaic(PV) models parameter identification algorithms at present, which have low accuracy and poor reliability, this paper proposed an improved symbiotic organisms search(SOS) algorithm for parameter identification of PV module models. First, to enhance the performance of original SOS, this paper propesd novel improved SOS algorithm, named as ImSOS. In ImSOS, it employed a quasi-reflection-based learning(QRBL) scheme in the population initialization step of original SOS. Moreover, it used the strategy of the modifications of benefit factors in the mutualism phase of SOS. It a adopted strategy of narrowing the search range of randomly generated coefficients in the commensalism phase of SOS. And then, the procedures and flowchart of employing the ImSOS for solving the PV module models parameter identification problem based on experimental current versus voltage(I-V) data of a real PV module was detailed. Finally, the ImSOS was demonstrated on the parameter identification of different PV module models of the Sharp ND-R250A5 PV module. Experimental results and comparisons with original SOS and the other seven novel intelligent optimization algorithms implied the effectiveness and superiority of the ImSOS. Therefore, the ImSOS becomes a new effective method to accurately and reliably identify PV module models parameters.