面向深海应用的海水柱塞泵复合故障诊断与性能退化方法研究

Safety is the first concern in deepsea equipment operation, which is greatly determined by the reliability of seawater hydraulic pump of deepsea equipment. This research proposal focuses on the problems introduced by composite faults and performance degradation of seawater hydraulic pump. The characteristics of performance degradation and composite faults will be analyzed; Cavitation and tribological performance will be proposed theoretically; The interaction behavior between Fault mechanism and fault features will be investigated. This project will establish the methods for extracting and fusing the multiple failure features from multiple physical fields. Real-time diagnosis methodology for composite faults will be brought backward. Considering the status transition of seawater hydraulic pump, the function based on probability and time-varying will be established. Typical techniques developed in the present research will be evaluated in the integrated platform of deepsea hydraulic system. The goal is to develop a systematic theory and methodology of diagnosis composite faults and performance degradation for seawater hydraulic pump. This project will provide effective theory foundation and technical support for the development of deepsea equipment.

深海海水柱塞泵是深海水下设备的心脏，其可靠性直接影响深海水下设备的运行安全。本项目进行深海海水柱塞泵复合故障实时诊断和性能退化的理论与方法研究。对海水柱塞泵故障的特点进行深入分析，揭示其复合摩擦磨损规律和空化气蚀机理，从理论上给出损伤机理与故障特征的映射关系。开发复合故障的实时诊断方法，获取多物理场故障特征量提取、融合策略。研究健康状态转移理论，建立海水柱塞泵健康状态模型，实现概率时变状态转移的准确描述。本项目的成果将在深海海水液压综合实验平台进行实验验证。形成一套针对深海海水柱塞泵的复合故障诊断和性能退化的理论与技术，为我国深海水下设备的发展提供有效的理论基础及关键技术支撑。

本项目深入剖析深海海水柱塞泵及系统的典型故障，揭示其空化气蚀机理、异常振动规律。在此基础上，结合光、电技术开发多物理场故障特征智能传感技术，建立了利用衍射方法水下声源探测技术和自适应杂质吸取理论与技术。研究了深海海水柱塞泵故障诊断与故障预测理论。建立了基于数据驱动的快速诊断方法和预测技术。提出相似样本和小样本结合的思想，探讨了利用权重迭代的迁移学习算法，解决小样本诊断预测难题，给出深海海水柱塞泵智能诊断体系，形成我国具有自主知识产权的深海海水柱塞泵健康管理理论和诊断方法。