基于PID神经网络的太阳能光伏发电并网逆变技术  

Photovoltaic grid-connected generation system based on PID neural network

作　　者：张辉[1] 许俊泽[1] 陈光梦[1] 

机构地区：[1]复旦大学电子工程系,上海200433

出　　处：《信息与电子工程》2009年第5期447-452,共6页information and electronic engineering

摘　　要：随着能源危机的进一步加剧和光伏并网发电成本的持续降低,光伏并网发电技术的应用越来越广泛,研究并优化其逆变方式、并网算法,具有相当大的现实意义。针对电网中的负载扰动,提出了一种将PID控制规律融合进神经网络之中的新的控制策略,抑制负载扰动,改善稳态情况下并网电流波形,同时提高了系统的动态响应性能。通过理论分析和仿真结果证明了太阳能光伏发电系统拓扑结构及控制算法的正确性、可行性和高效性。Recently, the shortage of energy becomes a serious problem. Because the cost of photovohaic （PV） module decreases continuously, the grid-connected PV power generation system is used widely, it requires more attention on convert method and grid-connected algorithm. Aiming to solve the problem of load disturbance, this study presented a novel PID neural network algorithm which restrained the disturbance of the load, improved the wave of grid-connected current and transient responses. Through theoretical analysis and experimental result, it proves that the solar PV power system topology and control algorithm are feasible and efficient, the theoretical and economic significance are immeasurable.

关 键 词：光伏系统 并网算法 PID控制 神经网络 

分 类 号：TM615[电气工程—电力系统及自动化]