Research On Sequence Planning For Human-robot Collaborative Disassembly

Disassembly is a key step to achieve product maintenance,remanufacturing and recycling.It plays an important role in alleviating the environmental resources problems caused by the end-of-life products and developing green manufacturing and sustainable manufacturing.Human-robot collaborative disassembly refers to the disassembly process that human and robot work together in collaborative workspace,combing their respective disassembly advantages to complete the disassembly task.It is effective to cover the heavy labor intensity and high labor costs in manual disassembly as well as the lack of flexibility and expandability in robotic disassembly.Disassembly sequence planning is the primary means to improve the disassembly efficiency and reduce disassembly cost and environmental impact before the disassembly task is performed.Thus the research on human-robot collaborative disassembly sequence planning has important theoretical value and practical significance.Combined with the characteristics of human and robot disassembly ability,the key issues in human-robot collaborative disassembly sequence planning are studied,and the main works are as follows:(1)The human-robot collaborative disassembly problem is analyzed and described,and the human-robot collaborative disassembly information model is built.Firstly,aiming at the insufficiency in the traditional interference matrix,the disassembly stability is considered,and an improved product disassembly information model is proposed to optimize the expression of disassembly information,as well as the judgment rules of removable parts are defined.After that,based on the analysis of the disassembly capabilities of human and robot,three types of parts are defined,and then combined with the necessary information of the disassembly process,the part disassembly information model is established.(2)The differences between the human-robot collaborative disassembly sequence and the ordinary parallel disassembly sequence are compared,and the generation strategies of human-robot collaborative disassembly sequence are studied.Sequence order and task assignment are considered simultaneously.In view of the over-idealization of the traditional time calculation model for the parallel disassembly process,the relevant influencing factors are analyzed in detail to fit the actual process.The calculation model of human-robot collaborative disassembly sequence time is given,and combined with disassembly cost and disassembly difficulty,a multi-objective optimization model of human-robot collaborative disassembly sequence planning is established.(3)This paper presents a multi-objective discrete bees algorithm to solve the human-robot collaborative disassembly sequence planning problem.According to the characteristics of bees algorithm,a selection strategy for elite scout bees based on Pareto fast non-dominated sorting is proposed.The global search ability and neighborhood search ability are adjusted dynamically during the iterative process by the labor adjustment strategy.And Lévy flight's step sizes are mapped to the selection of different neighborhood transform operators to improve the accuracy of solution and avoid local optimum.Finally,a computer case as an experimental case to verify the effectiveness and performance of the proposed algorithm.Aiming to improve disassembly efficiency,reduce disassembly cost,and optimize the disassembly process,this paper provides theoretical support for solving the task allocation and sequence planning problems of human-robot collaborative disassembly through human-robot collaborative disassembly sequence planning.