A Control Method for Coal Slime Water Thickening Process in Coal Preparation Plants Based on BP- TSFNN

Aiming at the complex mechanism of coal slime water thickening process in coal preparation plant, which has the characteristics of multivariable and strong nonlinearity leading to the difficulty of realizing the intelligent optimization control, a model predictive control method based on back propagation neural network ( BPNN) and Takagi-Sugeno fuzzy neural network (TSFNN) cascade (BP-TSFNN-MPC) is proposed. Further, in order to realize the intelligent optimization control of the thickening process, the BP- TSFNN-MPC controller is constructed, and the gradient descent optimization algorithm is used to solve the optimization problem. In the experimental part, the effectiveness of the model and control method is verified by combining the production data of the slime water concentration process and the simulation model of the thickener. The results show that the BP- TSFNN model has high prediction accuracy for the concentration process and BP-TSFNN-MPC has good control performance.

I. Introduction

The thickening process of coal slime water in coal preparation plants refers to the improvement of suspended solids settling efficiency by adding flocculant dissolved liquid to the coal slime water in the thickener, so as to effectively improve the effect of coal slurry water treatment, and it is a key research direction for the clean utilization of coal. Due to the complex mechanism of the thickening process, most of the current coal preparation plants rely on manual experience to control, resulting in unstable performance of the thickening process. Therefore, it is necessary to study the intelligent control method of the concentration process.