基于液压无源性的全海深电液伺服机械手自适应模糊控制策略研究

Full ocean depth hydraulic manipulator is an essential deep-sea operation equipment for scientific investigation and ocean resource exploitation in hadal zone. The hydraulic characteristics under thousands of meters deep in the ocean is unknown, which makes it difficult to carry out research on the electro-hydraulic servo system. A nonlinear mathematical model of the electro-hydraulic servo system is established according to the full ocean depth working condition. An 11000m online pressure experiment is conducted, which aims to analyze the full ocean depth hydraulic characteristics of the manipulator. Based on the experimental results, a system modeling and parameter matching method suitable for 11000m deep ocean can be structured. The hydraulic manipulator is inclined to suffer from complex hydrodynamic coupling loading during operation and the system parameters exist intense uncertainty due to the deep-sea operating circumstance. Therefore, the high-performance servo control of the manipulator has always been a thorny problem. In order to suppress the underwater loading disturbances, the adaptive fuzzy control is proposed, and the adaptive law for the fuzzy output is deduced. In order to compensate the time-varying parameters which vary significantly depending on the operation depth, a robust adaptive control based on passive theory is proposed, in which some important theories such as integral sliding mode variable structure control, extended parameter adaptive control and passive control theory are studied in this project. A compound controller combining the adaptive fuzzy control and the passive robust adaptive control is presented to achieve the high-performance control of the full ocean depth hydraulic manipulator. The research in this project can reveal the hydraulic characteristics in 11000m deep ocean and improve the control performance of full ocean depth hydraulic manipulator. The research achievements can provide theoretical support for developing the electro-hydraulic servo system of full ocean depth hydraulic manipulator.

全海深电液伺服机械手是我国进行万米深渊科学考察和资源开发必不可少的深海作业技术装备，由于万米环境下液压特性未知，难以有效开展其电液伺服系统的研究。本项目针对全海深工况建立其非线性模型，基于万米压力模拟实验，分析全海深液压特性，构建适于万米深海的系统建模和参数匹配方法。由于作业过程中存在复杂水动力耦合负载扰动，全海深工况导致系统参数具有强烈不确定性，因此，其高性能伺服控制是研究难点。本项目针对无法精确建模的水下扰动载荷，提出自适应模糊控制策略，研究模糊输出自适应调节规律；针对系统参数随作业深度变化大范围时变的问题，综合运用积分滑模变结构、扩展参数自适应和无源控制等理论，基于非线性模型，提出无源鲁棒自适应控制策略；在上述研究基础上，提出基于自适应模糊和无源鲁棒自适应控制相结合的复合控制策略。本项目研究可揭示万米深海液压特性，提高全海深液压机械手控制性能，为其电液伺服系统的研究提供基础理论支持。

“十三五”期间，我国启动包括全海深（最大深度11000米）载人潜水器、全海深无人潜水器等重大装备的研制，已全面进入万米深渊科考时代。而无论是极限海洋环境条件下的科学考察，还是海底资源的开发，都离不开水下作业工具和技术。全海深液压伺服机械手是万米深渊科学考察和资源开发的首选作业工具，是针对万米极限深渊的一种电液伺服控制作业系统，由于万米环境下液压特性发生变化，开展其电液伺服系统的液压特性和控制方法研究具有重要意义。.项目重点针对全海深超高压环境下，传动介质粘度显著增加，被控液压机械手系统动态特性发生变化，导致控制响应不一致甚至失稳等问题，建立了其全海深工况下的非线性模型，完成了陆上油液粘度模拟实验和万米压力实验，分析其全海深液压特性，进一步修正并验证了模型的准确性；通过全链路液压系统参数匹配与核心关节优化设计，实现了全海深机械手对液压介质粘度的宽适应性。针对系统参数随作业深度变化大范围时变的问题，综合运用积分滑模变结构、扩展参数自适应等理论，提出了自适应模糊控制策略，完成了液压机械手单关节全海深115MPa在线压力实验和伺服机械手整机的全海深在线压力实验，比较分析了液压机械手关节位置控制方法在不同环境压力和不同负载条件下的控制性能，在全海深115MPa压力环境、油液粘度增加约4.6倍的情况下，液压机械手的关节伸出时间由0MPa的1.78s变为115MPa时的2.32s，缩回时间由0MPa的2.31s变为115MPa时的3.13s，达到浅海响应速度的73%，基本实现了全海深机械手不同深度下的动态响应一致性，验证了控制方法的有效性。.项目共发表学术论文5篇，其中SCI期刊论文1篇，EI期刊论文1篇，国际会议论文3篇；录用EI国际会议论文1篇；在投SCI期刊论文1篇。相关研究成果为全海深液压机械手的研制提供了理论支撑，有效解决的全海深液压伺服机械手深海宽适应性与高动态响应一致性的问题，提高了液压伺服机械的控制性能。全海深液压机械手搭载于“奋斗者”号载人潜水器完成了岩石、生物抓取及沉积物取样器操作等精准作业任务，有效验证了项目的相关研究成果，成功填补了我国应用全海深液压机械手开展万米作业的空白。