Optimizing Job Shop Scheduling with Dynamic Arrivals using the Deep SOMs Approach

Modern manufacturing companies are more focused on serving the diverse demands of consumers in a timely manner, thanks to the exponential rise of globalization and information technology. Job shop manufacturing is characterized by a focus on small batches and a propensity toward producing a variety of products, which is why it has been an important production method for many industrial manufacturing enterprises. Data preprocessing, model training, and feature selection are the three stages that make up this suggested method. Step one is preparing the data to make it more accurate and reliable; step two is changing the data’s structure to make it more usable by machine learning algorithms. The concept of feature selection is based on the premise that features connected to individual fitness and its contribution to individuals are important. The use of a Deep SOM persisted throughout training. This novel method outperforms SOM and ELM with an average accuracy of 92.63%.