Research On Soft Sensor Of The Particle Size And Optimal Control For Grinding Process

The grinding process is an important process of the whole mineral industry,providing the qualified ore granularity task for subsequent sorting operations,its size directly affects plant product quality,metal recovery and economic and technical indicators.Because the grinding particle size is difficult to real-time online detection,and there is nonlinear,strong coupling and large time-delay complex characteristics between basic control loop output,it is difficult to use the detection technology and traditional control theory and methods to control the particle size within the range of technological requirements,therefore,we must adopt advanced control method to optimize the grinding particle size.First of all,according to the problem of the present instrument is that the grinding particle size cannot meet the real-time detection due to the long measurement,proposed the soft sensor method based on gaussian process,and the adaptive natural gradient is applied to the super gaussian process parameter optimization of the process,then build the model of grinding particle size soft sensor based on the adaptive natural gradient method in the gauss process.The simulation results show that the prediction accuracy of the soft sensing model is high,and it can effectively detect the grinding particle size in real time.Secondly,using the method of case-based reasoning and mathematical model,considering the current operating conditions and the size of soft sensor value,automatically update the preset value of basic control loop;In order to achieve the set value of the continuous and stable control,proposes an optimal control method of basis loop based on fuzzy model free adaptive control.The results show that the optimal control method can track the set value stably and quickly.Finally,In order to verify the actual availability of the soft sensing and the optimized control method,design the semi-physical simulation platform for grinding process,which includes virtual grinding process object,virtual instrument and implementation structure system,process control system,monitoring system and optimization system.Then the simulation experiment is carried out on the platform,and the results show that the method is feasible and effective.