中文摘要：

研究提出了一种基于声发射源特征识别的风电机组碰摩故障检测方法。同时为了对风电机组实时远程监控并实现分布式网络化管理,设计了一种基于ARM嵌入式系统的风电机组振动监测系统。系统有24通道的模拟信号采集电路,并借助于FPGA对周围电路进行逻辑控制和数据的实时采样;FPGA与ARM通信应用EDMA技术,提高了数据传输速率,可满足高速率采样的数据传输要求;此外,上位机与目标板之间的数据通信采用TCP/IP协议。通过实验观察上位机输出结果,验证了数据的实时性和准确性,达到了对风电机运行的状态信息监测和故障诊断的要求。

英文摘要：

A method of fault detection for wind turbine based on acoustic emission source feature recognition is proposed. To realize real-time monitoring and distributed network management, a wind turbines vibration monitoring system based on ARM embedded system is designed. The system includes 24 channels of analog signal acquisition circuit,the FPGA is applied to realize the logic control of the peripheral circuit including real-time sampling of data;to meet data transmission requirement for high speed sampling, the EDMA technology is adopted to access the data communication between FPGA and ARM, which extremely improve the data transmission rate;besides, TCP/IP protocol is used in the data communication between the object board and the upper computer. Tests results verify the real-time and accuracy of data, and indicate that it reaches the requirements of the state monitoring and fault diagno- sis of the wind turbines.