Optimization of Cold Rolling Jobs Under Time-of-Use Electricity Prices Based on Multi-Objective Genetic Algorithm and Power Prediction

Scientific and efficient job scheduling is crucial for maximizing production capacity and reducing energy consumption in the manufacturing industry. This paper focuses on scheduling cold rolling jobs on a single cold rolling mill, where each job may consist of varying numbers of operations. The optimization of job sequences aims to enhance production and conserve energy costs. To achieve this, two objective functions are designed and a multi-objective genetic algorithm is proposed to simultaneously minimize the total setup costs and electricity tariffs. The algorithm is further improved by incorporating accurate power prediction and heuristic rules based on real-world production data and experience. The proposed method is applied to actual rolling production and the results from experiments conducted at Dongyangguang UACJ Fine Aluminum Foil Co., Ltd. demonstrate that it can provide flexible solutions and save energy costs compared to the empirical job scheduling methods.