A dynamic job-shop scheduling model based on deep learning

Ideally, the solution to job-shop scheduling problem (JSP) should effectively reduce cost of manpower and materials, thereby enhancing core competitiveness manufacturer. Deep learning (DL) neural networks have certain advantages in handling complex dynamic JSPs with a massive amount historical data. Therefore, this paper proposes model based on DL. Firstly, data prediction was established for scheduling, long short-term memory network (LSTM) as basis; Dropout technology adaptive moment estimation (ADAM) were introduced enhance generalization ability effect model. Next, JSP described details, three objective functions, namely, maximum makespan, total device load, key chosen optimization. Finally, multi-objective solved by improved genetic algorithm (MOGA). The effectiveness proved experimentally.