Research On Automatic Control System Of Ash Content In Dense Medium Separation Process

In recent years,the intelligent transformation of coal preparation plants is in full swing.The heavy medium separation process is an extremely important link in the coal washing and processing process.It consists of automatic flow distribution,automatic water replenishment,automatic medium addition,barrel position balance,and density.The realization of the automation of each link such as intelligent setting has made the automatic control of heavy medium sorting have been greatly developed.In the actual production process,due to the change and fluctuation of raw coal quality and the measurement error of the online ash analyzer itself,the adjustment speed and accuracy of the ash closed-loop control link(intelligent density setting)still need to be improved.According to the investigation and analysis of the actual production situation of the coal preparation plant,the main problems of the closed-loop control of ash in the dense-medium separation system are that when the on-line ash analyzer detects raw coal and clean coal,due to unstable coal flow,fluctuations in coal quality,and different thickness of coal seams,etc.As a result,the online ash analyzer has large errors in the detection data,serious data loss,and interference from other factors in the correlation between the ash value of clean coal and raw coal and the set value of the sorting density,which makes the correlation poor and the comprehensiveness is not enough.In view of the above problems,this paper proposes an automatic control system for ash content in the heavy-medium separation process.First,analyze and study the heavy-medium separation process.Starting from the measurement data conditions of the online ash analyzer,the coal washing process is rectified,and then the ash content data,etc.Industrial experimental data is used for data preprocessing—interpolation analysis,and the principal component analysis method in Multivariable process statistical control is used to process and analyze the ash value data,belt transportation volume data,production data,etc.,and use the T~2 statistics graph in MSPC to The data is monitored in real time and analyzed whether it is controlled.When controlled,the difference between the actual value of the ash content of the clean coal and the target value of the ash content is less than or equal to 0.5%.When out of control,the ash content of the clean coal exceeds the qualified range of the ash content target value,the density prediction model constructed by the fuzzy neural network is used to predict the density,adjust the set value of the sorting density,and then adjust the actual density value to the minute value through the density automatic control system.Choose stable fluctuations near the density setting value.The thesis first introduces the automatic diversion,automatic water replenishment,automatic medium addition,barrel balance and other links in the heavy-medium sorting process,as well as the heavy-medium sorting process,and analyzes in detail the fluctuations in the quality of clean coal products in the heavy-medium sorting process.The main influencing factors,and a variety of factors are processed and analyzed through MSPC.From the hardware aspect,improve the measurement error of the online ash analyzer itself;from the software aspect,improve the accuracy of the ash data through interpolation analysis.Finally,a density prediction model is established based on the fuzzy neural network to improve the accuracy of the density setting value prediction.These three parts constitute an intelligent control system for ash back control in the heavy-medium sorting process.In the formulated automatic control plan for ash in the heavy-medium sorting process,through industrial field tests and collecting real-time data for analysis,the parameters of the interpolation analysis method,the principal component of the MSPC,the boundary range of the T~2 statistic map,and the fuzzy neural network are obtained.The parameters of the algorithm.The simulations are performed through MATLAB.The interpolation analysis method can represent more than 90%of the original data,and the principal component analysis method of MSPC can represent more than 80%of the total data.The F distribution boundary in the T~2 statistic graph can clearly distinguish whether the statistical sample is Controlled,fuzzy neural network density prediction model,the error between the predicted value of the density setting and the actual density setting value is small,and the prediction effect is good.This article uses the 1769 series PLC produced by AB as the control bottom layer,the upper computer software uses Kingview as the management layer,and the software platform uses MATLAB,SQL database and OPC server as the backend.The control bottom layer and the management layer adopt 4-20m A current signal and Modbus/TCP and Modbus 485communication protocol to realize data transmission,and the management layer and background adopt OPC communication protocol to realize data transmission.This system was put into trial operation in Zhongxing Coal Preparation Plant of Fenxi Mining Group.The operation effect of the trial operation stage was good.It can better solve the problems of coal quality fluctuation,detection data error,control time lag,etc.in the ash closed-loop control stage,and the stability and stability of ash The qualification rate has been greatly improved.