Abstract
This work focuses on multi-objective scheduling problems of automated manufacturing systems. Such an automated manufacturing system has limited resources and flexibility of processing routes of jobs, and hence is prone to deadlock. Its scheduling problem includes both deadlock avoidance and performance optimization. A new Pareto-based genetic algorithm is proposed to solve multi-objective scheduling problems of automated manufacturing systems. In automated manufacturing systems, scheduling not only sets up a routing for each job but also provides a feasible sequence of job operations. Possible solutions are expressed as individuals containing information of processing routes and the operation sequence of all jobs. The feasibility of individuals is checked by the Petri net model of an automated manufacturing system and its deadlock controller, and infeasible individuals are amended into feasible ones. The proposed algorithm has been tested with different instances and compared to the modified non-dominated sorting genetic algorithm II. The experiment results show the feasibility and effectiveness of the proposed algorithm.
Highlights
Scheduling is a resource allocation process over a period of time to perform a set of tasks.[1]
Based on the criteria introduced in the last section, we evaluate the performances of Pareto genetic algorithm (PGA) and compare PGA with non-dominated sorting genetic algorithm II (NSGAII) through the set of simulation examples
Two-objective optimization: we test the performance of PGA and modified NSGAII (MNSGAII) for scheduling automated manufacturing systems (AMSs) with two objectives, that is, Cmax and D
Summary
Scheduling is a resource allocation process over a period of time to perform a set of tasks.[1] Flexible flowshop and job-shop scheduling (FSP and FJS, respectively) are one of the most important problems in the area of production scheduling, and these are all wellknown non-deterministic polynomial-time (NP)-hard.[2] In the study of these scheduling problems, it is usually assumed that the capacity of buffers for storing jobs is unlimited, and when a machine has finished processing a job, the job can leave immediately, and the machine can start processing other jobs, that is to say, there is no blockage. For such an AMS, the required resource allocation or scheduling strategy can prevent the system from falling into
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
