Broad Learning System Based on Binary Grey Wolf Optimization for Surface Roughness Prediction in Slot Milling

Abstract

Surface roughness is considered to be one of the most significant indicators for evaluating machining quality. Manufacturing industries generally aim to achieve high product quality while considering reducing production costs and time. Therefore, an accurate and reliable surface roughness prediction model plays a key role in the production process. Recently, a broad learning system (BLS) has attracted great attention due to its outstanding performance. In order to improve the prediction accuracy of surface roughness and reduce the influence of uncertain factors, sensors are necessarily added to monitor the product processing. Consequently, the surface roughness data obtained in the actual machining process tends to have a small sample size and a high feature dimension. In addition, due to the structural characteristics of the BLS, its hidden layer neuron nodes often carry some irrelevant or redundant features during the training process, which affects the prediction performance of the model. To overcome these problems, this article proposes a novel BLS architecture for surface roughness prediction, in which binary grey wolf optimization (BGWO) is used for the selection of neurons in the hidden layer. To verify the effectiveness of the proposed method, a slot milling surface roughness experiment was performed and compared with other methods for different evaluation indicators. The results show that the proposed method provides an effective solution for surface roughness prediction.