水产养殖水下作业机器人关键技术研究进展

传统水产养殖的水下作业任务主要依靠人工完成,劳动强度大,危险性高,水产养殖水下作业面临严峻的人工危机。随着技术进步和制造成本的降低,将水下机器人应用于水产养殖作业有着巨大的需求空间。机器人水下捡拾作业强耦合、多干扰、时变、多体、多环节、作业目标规格形状不一,且作业的精度和速度要求高,机器人水下精准作业是困扰水产养殖界多年的公认难题。本文在对水下机器人-机械手系统(underwater vehicle-manipulator system,UVMS)进行了系统分析的基础上,从弱光照、强耦合、非结构化海洋环境下UVMS的精准捕捞作业目标识别、时变多体捡拾作业条件下水下机器人-机械手系统动力学模型、多扰动条件下目标动物位置识别与定位模型、多约束条件下浅海养殖精准捡拾作业的优化控制等4个方面,对UVMS捕捞作业所涉及的关键技术的地位、国内外研究现状展开分析与讨论,并对所述关键技术的研究前景进行了展望,以期为水产养殖水下作业机器人软件开发提供理论依据和综合性参考。

Research progress on key technologies of underwater operation robot for aquaculture

Abstract: Underwater operations in the traditional aquaculture are mainly carried out by manual labor, and it is labor-intensive and dangerous, at the same time the aquaculture industry is facing a severe labor crisis. With the advancement of technologies and the reduction of manufacturing costs, there is a huge demand for the application of underwater robots for aquaculture operations. The underwater vehicle-manipulator system (UVMS) is a new type of underwater operational robots. In recent years, the research of UVMS has gradually attracted the attention of scholars at home and abroad, and the application of UVMS to aquaculture fishing operations has a broad prospect. UVMS is a multibody dynamics system with dynamic coupling between the robot body and the manipulator. The modeling and control of the UVMS in the floating situation is very complicated, and its coupling control is not only affected by the conservation of momentum, but also affected by hydrodynamics. In addition, the aquaculture underwater operation environment is complex, the light is dim, the ocean current is time-varying, the target animals (sea cucumber, dead fish) are different in size, and the precision and speed requirements of the fishing operation are high, and the rapid identification of target animals and precision fishing operations are recognized challenges for the aquaculture community for many years. The application of UMVS to aquaculture fishing operations requires addressing several key technical issues, such as real-time target object recognition in complex environment, dynamics analysis and modeling of the UVMS as a multibody object, dynamic estimation of the underwater vehicle position relative to a target object, and constrained optimal guidance and control of UVMS under uncertainties. In order to provide a theoretical basis and comprehensive reference for the software development of aquaculture underwater robots, this paper focuses on the analysis and discussion of the role as well as the domestic and foreign research status of the aforementioned key technologies involved in UVMS fishing operations, and looks into the research prospects of the key technologies. In the research of key technologies for aquaculture underwater robots, the following suggestions should be noted: 1) Using multi-sensor information fusion technology to study the fast and accurate recognition algorithm of target animals under complex environment and disturbance conditions, and improving the target recognition speed and accuracy of aquaculture underwater operation robots are the inevitable direction of future research; 2) The development of integrated navigation systems with high reliability, high integration and comprehensive compensation and correction functions represents the development trend of aquaculture underwater robot navigation technology; 3) Under the premise of ensuring the stability of motion control of aquaculture underwater robots, improving the adaptability and fault tolerance of the control system, and continuously improving the feasibility of the intelligent system in practical applications are the focus of future research work.