In this paper, a tri-objective optimization problem related to sequencing in multi-model reconfigurable assembly systems (RAS) is addressed. RASs are also designed around product families similar to the more generic reconfigurable manufacturing systems (RMSs). A multi-model assembly line needs reconfiguration after assembling each batch of the products from the family. Further, the changeover of assembly line from one product type to another is also of varying degree of complexity and reconfiguration cost. Two novel indices for commonality and reconfiguration complexity are proposed here to incorporate this feature of RAS. Optimizing the reconfiguration time/cost is taken as the first objective while fulfilling the product due dates and the work load balance among the assembly stations as the second and third objective functions, respectively. This optimization problem is solved using a Multi-objective Self-Organizing Migrating Algorithm (MOSOMA) metaheuristic to find the non-dominated Pareto optimal solutions representing optimum sequencing of products in the family. An approach for identifying the best solution from the multiple Pareto optimal solutions using aggregate fitness is also proposed. A numerical illustration along with computational procedure is presented on a hypothetical RAS model.

Optimal sequence planning for multi-model reconfigurable assembly systems

Reconfigurable assembly system (RAS) is an innovative manufacturing paradigm which is an extension of reconfigurable manufacturing system (RMS) having the advantage of both dedicated serial lines and flexible assembly systems to address the challenges of dynamic market conditions. The primary goal of RAS is to improve the responsiveness of assembly system to fluctuating market demand by enabling rapid and cost-effective reconfiguration. In this paper, a model for optimizing the schedule plan with three objective functions related to RAS namely reconfiguration cost/time, delivery schedule, and work content balance is proposed. Finding the optimum sequence of a product in the product family while optimizing the three given objectives while meeting the customer demand are the main focus of this optimization framework.

Multi-objective Scheduling Model for Reconfigurable Assembly Systems

Reconfigurable Assembly System (RAS) is designed at the outset for rapid change in structure, in order to quickly respond to the needs of market in terms of product variety and quantity. This paper is focus on the architecture of RAS. We propose a hybrid architecture and negotiation protocol for the dynamic scheduling of assembling systems. The architecture is based on multi-agent systems. This architecture combines the advantage of both hierarchical structure and distributed structure. A global management coordinates the whole system and every agent is highly autonomous. Otherwise, we introduce the agent specification. The well-known Contract Net Protocol has been adapted as the coordination model. The purpose of this protocol is to dynamically assign operation to the resources of the assembly system to realize system reconfiguration.

Xuemei Huang

Papers

Study on the Architecture of Reconfigurable Assembly System