A multi-agent-based mould due date setting approach in stochastic production

Abstract

Due date (DD) quotation for the customer enquiry is strategically important in the mould industry, having a large influence on the delivery reliability performance. This paper develops a multi-agent system framework (MASF), which uses a multi-agent system to provide a reliable and competitive DD for the new arriving orders. Firstly, a decision-making structure for mould DD quotation is proposed, which can determine reasonable delivery times or reject the unsuitable orders. Compared with previous research, a distinguishing feature of the structure is to take into consideration the uncertainties of resources for processing task via a stochastic state evolution model (SSEM). Based on the presented structure, the MASF is then developed and described using six different agents. The algorithm agent (AA) is the key to the framework and contains two models: capacity check model and project progress evolution model. The former is used for verifying the feasibility of the fixed DD, and that the latter can compute the probability distribution and mathematical expectation of the negotiable DD. Other agents are responsible for providing necessary information to the AA. Finally, a numerical example of a real-life mould production system is studied and shown to be more suitable for a stochastic environment than related research.