Research On Intelligent Optimal Scheduling Methods For Automatic Guided Vehicle In Matrix Manufacturing Workshop

With the continuous improvement of social productivity and the level of science and technology,as well as the popularity of the Internet,consumer demand is increasingly showing a trend of personalization and diversification.In response to the rapidly changing market demand,many manufacturing enterprises transform the production mode from mass production to mass customization,and construct appropriate matrix manufacturing workshop based on matrix production concept.Automatic Guided Vehicle(AGV)scheduling problem is an important problem to be solved urgently in matrix manufacturing shop logistics system.Based on the AGV scheduling problems in matrix manufacturing workshop under the two production modes of single-species,large quantities and multi-species,small quantities,in this paper,the mixed integer linear programming models are established respectively.The prior knowledges and the structure features of the problems are investigated.The key theories and techniques of artificial colony algorithm and iterative greedy algorithm are explored.A series of innovative optimization scheduling theories and methods are proposed.The main research works are as follows:(1)For the matrix manufacturing workshop under the production mode of single-species,large quantities,this paper abstracts a single-load AGV scheduling problem with the capacity and time window,establishes a mixed integer linear programming model with the objective of minimizing the transportation cost,and proposes a discrete artificial bee colony algorithm.In the proposed algorithm,we propose a nearest-neighbor-based heuristic(NNH)to generate an initial solution with a high level of quality,five neighborhood operators to improve the diversity of generated solutions,four theorems to avoid the unfeasible solutions,one control parameter to balance the global exploration and local exploitation in employed bee and onlooker bee phases,and an insertion-based local search method to lead the algorithm to a promising region of the solution space.(2)For the matrix manufacturing workshop under the production mode of multi-species,small quantities,this paper abstracts a multi-load AGV scheduling problem with the compartment capacity,establishes a mixed integer linear programming model with the objective of minimizing the transportation cost,and proposes an iterated greedy algorithm.In the proposed algorithm,we propose an improved NNH heuristic(INNH)and an improved sweep-based heuristic(ISH)to generate a high-quality initial solution,an accelerated method to improve the efficiency of evaluating neighborhood solutions,an improved deconstruction procedure to improve the quality of the generated solutions,and a simulated annealing type of acceptance criterion to decide whether to accept the new solution.(3)Based on the above model,this paper considers the scenario of multi-load AGV simultaneous pick-up and delivery and the multi-objective characteristics,extracts a multi-objective AGV scheduling problem,establishes a multi-objective mixed integer linear programming model with the objective of minimizing the transportation cost and maximizing customer satisfaction,and proposes an effective multi-objective evolutionary algorithm.In the proposed algorithm,we propose an INNH-based heuristic(v NNH)to generate a high-quality initial solution,a local search based on an ideal-point selection to enforce the exploitation capability,a two-point crossover operator making full use of valuable information to improve the global detection ability of the algorithm,and a restart strategy to avoid the algorithm trapping into a local optimum.(4)Based on the above model,this paper further considers the release times of multi-load AGV and energy-efficient indicators,extracts a multi-objective AGV energy-efficient scheduling problem,establishes a multi-objective mixed integer linear programming model with the objective of minimizing energy consumption and number of vehicles used and maximizing customer satisfaction,and proposes an effective multi-objective greedy algorithm.In the proposed algorithm,we propose a hybrid heuristic to generate a high-quality initial solution,an energy-efficient strategy to obtain the energy-efficient solutions,a construction operator based on an ideal-point selection to reduce the computational complexity,a self-adaptive multiple neighborhood local search to find a better neighborhood solution,five neighborhood operators to improve the diversity of generated neighborhood solutions.For the proposed neighborhood operators,four theorems for avoiding infeasible solutions and two properties for judging the dominance relationships of solutions are proposed.Finally,the effectiveness of the proposed algorithm is verified by the simulation experiment of the AGV scheduling problem in the above scenarios.