FAST液压促动器群故障预测与维护决策研究

With the continuous application of active surface technology for large aperture astronomical telescope, the operation and maintenance problem is proposed, which is related with macro-scale mechanical-electrical-hydraulic complicated systems in one location. FAST（Five-hundred meter Aperture Spherical Telescope）hydraulic actuator group will work on the condition of outdoor, lower speed and heavy loading all long. And it has the characteristics of close power transmission, hard to measure operating parameters and difficulty to obtain fault signals. Since FAST actuators are distributed at surface of 500m aperture spherical mountain, it is inconvenient to check on site and will get worse predicting accuracy. Consequently, it is necessary to establish an effective and scientific technology security system for operation and maintenance..According to FAST construction progress and previous academic achievements，this subject is scheduled to adopt Prognostics and Health Management (PHM) method, including three sections. Firstly, take accelerated life tests as guide, analyze historical fault features and big data from on-line monitoring platform, explore fault generating rule and estimate remaining useful life. Secondly, establish fault prediction model based on data driven, verify by experiment and then improve. Thirdly, by comprehensive consideration of each influencing factor, give the optimal maintenance strategy and provide accuracy repair activity at correct position on precise time..The deeply research of this project will improve operation and maintenance capability for macro-scale adjusting mechanism group of large aperture astronomical telescope in China, promote our country hydraulic industry developing to the direction of higher reliability level, and also extend PHM technology to the field of large-scale astronomical instrument.

伴随大口径天文望远镜主动反射面技术的不断应用，提出了局地大规模机电液复杂系统运行与维护问题。FAST液压促动器群长期野外低速重载运行，动力传递封闭，参数可测性差，故障信息难以提取。分布于500m口径球冠状山体表面，人工巡视困难，故障预测效果也不理想。因此，需要建立一套科学有效的运行与维护技术保障体系。.结合FAST建设进展与前期研究成果，拟采用故障预测与健康管理解决方案，包括：以加速寿命试验为先导，分析故障特征历史及在线监测状态大数据，探索液压促动器故障演化机制与推算剩余使用寿命；建立数据驱动的故障预测模型，试验验证并不断完善；综合考虑各影响因素，给出最佳维护策略，实现准确时间对准确部位采取准确维修活动。.本课题的深入研究，将有望提高我国大口径天文望远镜反射面大规模调节机构群的运行与维护能力，推动我国液压执行机构产业向更高可靠性层次发展，拓展故障与健康管理技术到大型天文仪器领域。

为解决大口径天文望远镜需求与其建造困难的矛盾，主动反射面技术逐渐不断得到应用，随即提出了局地大规模执行机构运行维护问题。FAST望远镜共有2225套液压促动器，长期野外低速重载运行，动力传递封闭，参数可测性差，故障信息难以提取。分布于500m口径球冠状山体表面，人工巡视困难，故障预测效果也不理想。因此，本项目提出了故障预测与健康管理解决方案。.经过为期三年研究，对液压促动器机械液压故障与电气故障进行了机理分析及大量实验数据归纳，具体开展了：.1）基于液压促动器齿轮泵泄漏特性及液控单向阀开启特性，提出以实际流量闭环描述活塞杆外伸过程并建立促动器数学模型方法。利用自行设计的试验装置，辨识了齿轮泵、液控单向阀样本上没有的2个关键特性参数。该机理分析所得数据，可作为后续促动器状态监测、故障分析与预测、维护升级改造的依据。.2）提出促动器单机加速寿命试验，经对故障泵体进行拆解分析，发现影响其寿命的主要因素是液压油污染。油液污染后对会引起齿轮衬套的剧烈磨损，导致内泄露增加，系统无法建压；污染物造成电磁阀动作迟滞、液压锁失效等故障。.3）促动器电气元件集中在密闭的电气舱内，热源与舱罩基本为点或线接触，且压紧簧片随时间增加产生塑性变形而贴紧力减小，以上三种因素导致电气元件散热困难，故障率较高。为此，设计了开门式电气舱罩，提高热源处的热传导效率。.4）依据单机故障发展规律及促动器群故障统计，结合物料（包含液压油、密封圈、齿轮泵、电磁阀、轴承、电机、开关电源等）贮藏属性及供货周期，现场维护车间动态调整元件数量及整机数量，做到最优维护条件下保证FAST望远镜健康运行。.以上研究工作为FAST望远镜促动器群的正常运行提供了技术支撑，保障了FAST望远镜的科学观测，对其它大型天文仪器执行机构的运行维护提供了借鉴，也推动了国内液压工业可靠性提升。