Collaborative Optimization of Transfer Line Balancing and Buffer Allocation Based on Polychromatic Set

Abstract

The transfer line balancing problem (TLBP) and the buffer allocation problem (BAP) are amongst the most studied problems on transfer line system. However, the two problems are usually solved separately, although they are closely interrelated. Traditional optimization of TLBP probably leads to a deviation of production rate from the actual performance which is used as the optimization objective of the following BAP, especially when the equipment reliability differences are relatively large. In this paper, an innovative approach, considering machining accessibility and machine reliability, is presented to solve the TLBP and BAP simultaneously. A polychromatic-set-based constraint model and a collaborative optimization algorithm are proposed. A group of Boolean constraint matrixes both for machining feature sets and their optional stations sets are constructed through polychromatic set theory. A new coding mode based on priority, which always keeps all schemes feasible in each iteration and suitable for GA, PSO and other intelligent algorithms, is presented to describe all the allocation information for station, process, buffer and configuration. Optimization objectives including production rate and production cost are evaluated through the COM interface with simulation software. Comparison experiments and a real case of a diesel engine block demonstrate both the validity and efficiency.