蚁群优化算法在风电偏航故障检测中的应用  

Ant colony optimization applied in the fault detection of wind yaw

作　　者：张海涛[1] 高锦宏[1] 吴国新[1] 左云波[1] 

机构地区：[1]北京信息科技大学现代测控技术教育部重点实验室,北京100192

出　　处：《可再生能源》2013年第11期48-50,55,共4页Renewable Energy Resources

基　　金：国家自然科学基金项目(51275052)

摘　　要：针对风电机组结构与工况特点,通过研究风力特性和目前的信号检测方法,提出一种新型风电偏航故障检测方法。将风向特性提炼为Elite因子并融合到蚁群智能算法之中,弥补蚁群算法全局搜索周期较长的不足,提高处理有效数据的精度,并应用BP神经网络对偏航故障进行诊断。通过转子故障实验台试验并采集数据,应用Matlab软件分析试验数据,结果表明,与传统方法相比,新方法缩短了故障检测时间,提高了检测精度。According to the characteristics of structure and working conditions of wind turbine, a type of wind yaw fault detection method is proposed after studying the characteristics of wind and current signal detection methods in the paper. The Elite factor is fused from wind feature, which is used to ant colony algorithm. The method makes up for a lack of long time global search, and the processing accuracy of effective data is improved. The BP neural network was applied to the fault di- agnosis of yaw. At last, an experiment is made in rotation machine test bench, and collected data are analyzed by Matlab. The result shows that the new method is better than traditional method with shorter time of the fault detection and more accurate.

关 键 词：风力发电 故障检测 蚁群算法 BP神经网络 Elite因子 

分 类 号：TK83[动力工程及工程热物理—流体机械及工程] TP206[自动化与计算机技术—检测技术与自动化装置]