Research On Worker-Machine Dual-Resource Constrained Job Shop Scheduling Problem Considering The Uncertainty Of Processing Time

The production workshop for avionics precision parts has obvious semi-manual and semi-automatic features,and the dual resource constraints of employees and machines need to be considered comprehensively when scheduling production.However,as an important condition of production scheduling,the working hours of process processing are difficult to directly determine due to the non-uniform shift system of operators and the large differences between the processing methods and environments of each process.In addition,in order to ensure the feasibility of the scheduling scheme,it is necessary to further consider the balance between the completion time,employee load balance,and employee employment cost during production scheduling.Aiming at the above problems,this thesis conducts research on the man-hours prediction method of avionics precision parts and the single-objective and multi-objective worker-machine dual-resource constrained job shop scheduling problem considering the uncertainty of processing time.The main research contents are as follows:Firstly,in view of the problem that the employee shift system is not uniform,the processing scenarios and processing characteristics of each process are quite different,which makes it difficult to uniformly predict the processing time of avionics precision parts,a man-hours prediction method based on process clustering is proposed.Through the analysis of the actual processing scene,10 core processes are selected as the research objects.Fuzzy clustering is carried out on these 10 procedures,and different man-hours prediction models are established according to the clustering results of the procedures.Based on the actual data,the calculation,fitting and verification of various man-hours prediction models are realized.Secondly,in order to solve the problems of long order processing time and unreasonable employee allocation caused by the difficulty of determining the man-hours of the processes and blindly matching employees for each process based on experience,the worker-machine dual-resource constrained job shop scheduling considering the uncertainty of processing time is carried out.Based on the man-hours prediction model and the flexible job shop scheduling model under the worker-machine dual resource constraints,a worker-machine dual-constraint mathematical model considering the processing time uncertainty is established.Aiming at the characteristics of this model,a single-objective memetic algorithm is designed by combining the neighborhood search strategy based on hidden time window proposed in this thesis with the improved genetic algorithm after encoding and decoding.The algorithm is compared with the traditional genetic algorithm and the actual production workshop scheduling scheme,and the excellent performance of the single-objective memetic algorithm on this problem is verified.Finally,in the process of employee allocation,it is difficult to balance the work intensity of employees while taking into account the employment cost,resulting in the loss of employee resources or high processing costs.Two optimization objectives of staff load balance and staff employment cost are added,and a multi-objective memetic algorithm is designed.This algorithm combines the critical path-based two-stage neighborhood search strategy proposed in this thesis with the NSGAII global optimization algorithm to achieve precise optimization of the global solution on all optimization objectives.The multi-objective memetic algorithm is compared with the traditional NSGAII algorithm to verify the effectiveness of the multi-objective memetic algorithm for this problem and the optimization ability of the proposed neighborhood search strategy.This thesis provides some basic research ideas for worker-machine dual-resource constrained job shop scheduling problem considering the uncertainty of processing time,and provides method support for effectively improving the optimization ability of scheduling algorithms under this problem model.