Grey Wolf Optimization-Based Deep Neural Network for a Flexible Job Shop Scheduling Problem

Deep neural networks (DNNs) are a robust and versatile machine learning technique that offers a wide range of applications across various domains. They constitute effective and approximate models that can replace realistic models. However, designing, training, and refining the model poses a significant challenge to achieve the desired outcome. To address this challenge, metaheuristics are employed for DNN structure optimization to improve their accuracy and efficiency. The present work proposes a Grey Wolf Optimization (GWO) technique to find an optimal or near-optimal DNN configuration to approximate a realistic FJSSP by considering makespan as a scheduling objective function. Different DNN hyper-parameters are adjusted according to the accuracy of the solutions, including the number of neurons in each hidden layer, activation functions for both hidden and output layers, and learning rate and momentum. For analyzing and evaluating DNN performances, an investigation is carried out considering the minimization of the maximum value of Relative Errors (REs). These latter are calculated between the DNN-predicted makespan values and their real corresponding values of the data validation set. The results obtained during the application of DNN to a case study representing a Flexible Job Shop Scheduling Problem (FJSSP) demonstrate the effectiveness and high precision of the model. The case study results confirm that the model can effectively handle FJSSP.