摘要
针对神经网络算法在太阳能电池最大功率点跟踪中收敛速度慢,结果容易陷入局部极小等缺点,提出了一种基于遗传算法优化神经网络来实现最大功率点跟踪的控制方法。该算法利用遗传算法具有全局搜索问题解的特性,通过选择、交叉、变异等遗传操作,实现了神经网络权值优化,克服了神经网络初始权值对控制效果的不利影响。实验结果表明:该算法提高了神经网络在最大功率跟踪中的收敛性与非线性逼近能力,在日照强度、环境温度变化时仍能快速、准确地跟踪到太阳能电池的最大功率点,具有较好的稳定性。
Aiming at the disadvantage of low convergence speed and local minimum of neural networks algorithm used in the solar cell maximum power point tracking（MPPT）,an effective neural network algorithm based on genetic algorithm（GA） is proposed to realize the MPPT control in this paper.Adopting the global searching characteristic of genetic algorithm and genetic operations such as selection,crossing and mutation,an optimization of weights in neural networks are achieved,and the negative effects of network＇s initial weights on control effect are solved.The simulation and experiments show that the algorithm can improve the convergence speed and nonlinear approximation ability in the MPPT control,and quickly and accurately track MPPT in the sunshine intensity,ambient temperature changing environment.
出处
《计算机技术与发展》
2010年第9期197-200,共4页
Computer Technology and Development
基金
国家自然科学基金资助项目(50575168)
陕西省自然科学基金资助项目(2009JM8006)
