Multi-variety, Small Batch And Order Production System Optimization Study

The optimization problems of multi-variety, small batch and order production system for medium and small-sized enterprises in China are studied in this paper. It includes five parts.In the first part, introducing this paper study background and meaning ,discussing the domestic and foreign study actuality and summarizing theory and method of multi-variety, small batch and order production system.In the second part, using the synchronizer to simulate the flow of multi-variety, small batch and order production system, analyzing this system's flow bottleneck in its foundation.In the third part, the job-shop scheduling problem of multi-variety, small batch and order production system is studied. First, using Time Place Petri-Net to build up job-shop scheduling module, and then describing the module with algebra, discussing that the Job-Shop scheduling problem solution space is Non-polynomial complete problems so that getting the optimal solution is impossible by analytical method, so using the Depth-First rule and the Shortest Processing Time rule to get the solution, which is a Shortest Processing Time - Depth-First heuristic scheduling algorithm.In the fourth part, the delivery date forecast problem of multi-variety, small batch and order production system is studied. Because of the characteristics of BP neural networks such as self-adapt, self-study and altitudinal fault tolerance, especially having the ability of simulation for the nonlinear, discreteness dynamic system, so building up the production finishing date forecast model of multi-variety, small batch and order production system by ameliorated BP neural networks to forecast the order date, which can solve the order date problem of being difficult to ascertain.In the fifth part, production batch problem of multi-variety, small batch and order production system is studied, mainly including the optimizations of put-into-production batch and order batch. In the optimization of put-into-production batch batch, studying the issue of decision-making of planning put-into-production batch in multi-variety, small batch and order production system and analyzing the effect of output excess and output insufficiency on production cost. Due to different requirements for technique, procedure and material in client's orders and the uncertainty of production process, so that the rates of qualified products fluctuate stochastically. Taking the minimum expectation of losses as the criterion of decision-Making and supposing the qualified-products-rates submitted to normal distribution, developing a decision-making model of put-into-production batch. The results of numeric simulation and sensitivity analysis prove that the model is effective.In the optimization of order batch, Using Activity-Based Costing method to analyze the cost multi-variety, small batch and order production system's structure, giving eight math's formula to compute the different kinds' activity cost in its foundation, and then building up the profit optimize model, in which the most profit is the object function and the customer's order quantities are decision-variable. At last, applying genetic algorithm for imitation by some enterprise's part data, the results show that enterprise can increase its profit at the condition of adjusting the customer's order quantities.