Multi-objective optimization of job shops with automated guided vehicles: A non-dominated sorting cuckoo search algorithm

The reliability of machinery and automated guided vehicle has been one of the most important challenges to enhance production efficiency in several manufacturing systems. Reliability improvement would result in a simultaneous reduction of both production times and transportation costs of the materials, especially in automated guided vehicles. This article aims to conduct a practical multi-objective reliability optimization model for both automated guided vehicles and the machinery involved in a job-shop manufacturing system, where different machines and the storage area through some parallel automated guided vehicles handle materials, parts, and other production needs. While similar machines in each shop are limited to failures based on either an Exponential or a Weibull distribution via a constant rate, the machines in different shops fail based on different failure rates. Meanwhile, as the model does not contain any closed-form equation to measure the machine reliability in the case of Weibull failure, a simulation approach is employed to estimate the shop reliability to be further maximized using the proposed model. Besides, the automated guided vehicles are restricted to failures according to an Exponential distribution. Furthermore, choosing the best locations of the shops is proposed among some potential places. The proposed NP-Hard problem is then solved by designing a novel non-dominated sorting cuckoo search algorithm. Furthermore, a multi-objective teaching-learning-based optimization, as well as a multi-objective invasive weed optimization are designed to validate the results obtained. Ultimately, a novel AHP-TOPSIS method is carried out to rank the algorithms in terms of six performance metrics.