Evolving Neural Networks for Bi-Level Human Decision-Making Prediction in Mineral Processing

This paper presents a novel approach for predicting bi-level human decision-making in mineral processing. This method aims to enhance the adaptability of human decision-making support systems and reduce dependence on expert knowledge. To achieve this, it leverages a combination of evolutionary computation and neural networks. The proposed method utilizes a hierarchical neural network framework optimized through evolutionary computation. This framework takes advantage of the sequential nature of decision-making within the mineral processing system. The effectiveness of the method is validated through experiments using 1,324 real-world data entries from a mineral processing plant in China. The results demonstrate the method’s capability to improve prediction accuracy.

I. Introduction

With the development of industrial automation and smart manufacturing technologies, accurately predicting decision-making behaviors in industrial processes is crucial for optimizing production workflows and enhancing efficiency [1]–[4]. Decision-making in mineral processing is essentially a hierarchical process, primarily involving the operational optimization layer’s decision on setting values for operational indicators based on comprehensive production indices and related variables, as well as the operational control layer’s decision on setting values for control systems based on operational indicators and related variables. The main challenge lies in decision-makers, due to insufficient experience, need to overcome the complexity of decisions due to the uncertainty of raw materials and the variability of production conditions. Additionally, decision-makers at the operational control layer also need to overcome the decision bias introduced by decision-makers at the operational optimization layer. Therefore, establishing a decision analysis system can help correct erroneous decisions made due to the lack of experience by decision-makers, and replace manual decision-making after long-term stable operation.