Study On Bidding-based Decentralized Production Planning And Control

The production planning and control is the important function of business management, which is used for solving problems in the management of material flowing and the product flowing and use for guiding every enterprise's production and operating activities. The traditional production planning and control exists some problems, such as a static set of the parameter–due date in advance ;the deviation of production planning and control phases; lack of considering capacity fully; weak in low level and so on. With the developing of market, the demand for customized products with high quality, low cost and short life cycle is increasing, traditional production planning and control can't satisfy the market's request for products. So making production planning and control scientific plays an vital role in reducing production costs, shortening the sum of earliness and tardiness penalties, decreasing make-span time,improving the productivity of manufacturing firm.This paper generalized the concept of production planning and control, which occupies an important position in modern manufacturing system. The paper studied in the decentralization of production planning and control issue by simulating the bidding behavior in market. The paper proposed a bidding-based decentralized production planning and control model,pointed out its superiority,discussed its characteristic and the implementation step.The paper focused on four key issues of the bidding-based production planning and control -- the realization of multi-agent system functions, the construction of bidding evaluation functions, the proposing and embedding task allocation algorithm, the control issue under the decentralization. We constructed a distributed multi-agent system, set the permission of each agent, proposed a fuzzy evaluation model dealing with make-span,finish date,processing cost,processing precision of tasks ,while proposed a set covering-based task team allocating heuristic search algorithm.Last but not least important, the paper employed Tianjin Exploratory Machinery Company to witness the effectiveness of task planning steps of this distributed multi-agent system, compared this result with that of traditional task allocating mechanism, and proved that the bidding-based decentralized production planning and control can reduce the conflicts of capacity in production system, shorten the make-span time of each product ,which can optimize the entire production system and obtain the near optimal task planning and scheduling, then guarantee profit maximization as well as robustness of plan execution.