Integration of Kalman Filter with Ant Colony Optimization for Solving the Job Shop Scheduling Problem

Abstract In smart manufacturing, the job-shop scheduling problem (JSP) is a major obstacle that must be solved by the best possible sequencing of task operations. Dynamic job-shop environments require flexible scheduling systems that can adjust to changing conditions due to unpredictabilities like machine breakdowns. Traditional methods, which only provide the best answers when they are put into practice, are not adaptable enough to take into account shifting circumstances. Because of this limitation, temporal complexity has increased, highlighting the importance of sophisticated, flexible scheduling techniques in smart manufacturing. Several metaheuristic techniques, such as the well-known Ant Colony Optimization (ACO), are inspired by natural phenomena and are remarkably successful and efficient at solving extremely difficult (NP-hard) combinatorial optimization problems. This paper presents the implementation of an Ant Colony Optimization with Kalman Filter (ACO_KF) model algorithm applied to solve the JSP. ACO_KF is a combination of the recursive estimating algorithm for dynamic systems with the metaheuristic optimization algorithm inspired by ant foraging behavior to solve optimization problems. Our proposed approach aims to implement an ACO algorithm for solving a JSP and optimizing the makespan time by adjusting pheromone levels on paths. Also, the algorithm incorporates a Kalman filter to adaptively adjust pheromone levels according to recorded makespan times, to improve the convergence and efficiency of the ACO algorithm. Comparing the quality of the solutions to the most well-known outcomes from the most successful methods was necessary to evaluate the algorithm's performance on reference JSP. The solutions were obtained with remarkable efficiency and excellent quality.