项目从海上风电系统可监测、可预测和可维护等需求出发，针对海洋复杂环境下双馈式风力发电机系统的机电多故障特征提取与定量识别难题，从机理分析入手，研究海洋环境下双馈电机轴承锈蚀及机电多故障失效机理和动态响应特性，明确多故障特征与信号征兆的映射关系，为特征提取奠定理论基础；研究广义两尺度变换和多小波提升策略，提出集成多小波构造新方法，构建高自由度、多参数调控的大差异柔性基函数，为信息特征分离提供方法；基于智能优化方法，研究集成多小波子空间信息自适应匹配准则，实现多基函数与发电机机电多故障特征的高效自适应匹配，增强特征提取能力；提出自适应性集成多小波的最优分解策略，建立无量纲测度指标与故障等级判断标准，实现机电多故障的定量诊断。通过实验研究和工程应用检验和修正项目的理论方法。研究成果为海上双馈风力发电机故障预示与运行评估提供理论和技术支持，提高发电系统服役性能，具有重要理论意义及应用价值。

According to the monitorable, predictable and maintainable requirement of offshore wind-driven doubly-fed generator, this program aims at studying the problem of multi-fault feature extraction and quantitative identification of the doubly-fed generator, starting from the research of mechanism analysis, the failure mechanism and dynamic response characteristics of bearing corrosion and the electromechanical coupled multi-fault of wind-driven doubly-fed generator in marine environment are researched to reveal the relation between multi faults and the signal characters, which lays the theoretical foundation for multi-fault feature extraction and separation; An new ensemble multiwavelet construction method based on the generalized two scale transform and multiwavelet lifting scheme theory is proposed to build a discrepant flexible multiwavelet basic function library with the large freedom and multi-parameters, which provides technical means for separation of multiple features; Based on the optimization algorithm, the adaptive ensemble multiwavelet subspace information matching criterion is proposed to study the dimensionless measure optimization index, adaptive optimal construction is obtained, which enhance the ability of multiwavelet to match the multi-fault features, and it is benefit to realize the extraction features. Furthermore, the optimal decomposition strategy of adaptive ensemble multiwavelet is proposed, and the dimensionless measure index and fault grade judgment standard are built, which can realize the quantitative diagnosis of electromechanical multi-faults potentially. Through the experimental research and engineering application of the theoretical method of testing and correction, the research achievement will provide theoretical and technical support to the fault prediction and operational safety evaluation of the offshore wind power system, which will effectively improve the service performance of the generator, which is of high valuable for both academic sense and applied meaning.