Soft Sensor Of Particle Size Of Grinding Process And Intelligent Optimization Control

Grinding process has synthetic complexities such as strong non-linearity,severe coupling,large time delay.Measuring techniques and traditional control theory are limited.For the increasing needs of grinding products and the high requirements of quality,traditional measurements and control methods can hardly meet practical demands.Accurate measurement and effiective control for particle size has become a research hot spot in mineral industry.Firstly,the improved chaotic self-adaptive particle swarm neural network soft measurement method was proposed to solve the real time problem when measuring the particle size.In this method,adaptive strategy was adopted to adjust inertia weight factor,which strengthen the search speed and ability.Variable learning factor is employed in enhancing the ability of globally optimal search.The chaotic state was introduced into the optimization variables to avoid the search process getting into local extremum.The simulation results showed that the accuracy requirement of actual technology can be satisfied based the proposed soft sensor method which can be applied to measure particle size in real time.Secondly,the overflow particle size control subsystem which related to processing quality indicators closely,were analyzed in detail.Variable ratio control was adopted to achieve automatic control of grinding process.And grinding particle size control circuit was used as main circuit of cascade control to realize cascade and ratio control for particle size.The simulation results showed that grinding particle size can be stabilized in the range of technical index requirements within the given feed and fluctuant.Finally,the hardware-in-loop grinding simulation platform was designed and researched to prove the practical applicability of soft sensor and variable ratio control.The platform was consisted of virtual instruments and actuators,the virtual optimizing control object,a supervisory system,and an optimization system.The control effect was tested by combining soft sensor and variable ratio control,which showed feasibility and effectiveness of the proposed method.