Research On Control Strategy Of Automatic Window In Heading Face

Ventilation safety of coal mine is one of the important conditions to ensure the safe of coal mine.A good ventilation environment can provide safety guarantee for the underground workers.The most prone place of safety accidents in coal mine is the drivinging face.The local ventilation equipment of the driving face mainly include the local ventilators and the wind windows.The local ventilators are controlled by frequency conversion.Therefore,it is necessary to frequently adjust the window opening to control the air volume..The existing coal mine underground regulating wind window still adopts a manual adjustment method.This method cannot adjust the opening of the wind window in time according to the real-time measured air volume,and it is difficult to ensure the adjustment accuracy,and it is impossible to realize the uniform speed adjustment,which poses safety risks.Therefore,it is of great significance to realize the automation of the wind window and how to connect the control of the opening of the wind window with the ventilation volume and the concentration of harmful gases.This thesis takes the wind window of the driving face as the research object.the structure of the wind window is designed,the control process is studied,and the control process of the wind window is characterized by hysteresis and nonlinearity.The factors that influence the gas concentration are analyzed.When the speed of the local ventilators remains unchanged,the biggest factor that affects the gas concentration is the opening of the air window.Therefore,the opening of the air window.as the control quantity to control the gas concentration of the driving face.This thesis studies the classification principle and regression principle of support vector machine,the advantages of support vector machine in gas concentration prediction,the particle swarm optimization algorithm,and search the parameters of SVM and BP neural network.Establish the prediction model of gas concentration by SVM,PSO-SVM and PSO-BP.and MATLAB simulation is used to analyze the prediction results.The model of gas concentration with the highest fitting degree is selected as the prediction model in the automatic window predictive control system.The PSO-SVM predictive model is established for the wind window control system,and the model linearization is combined with the generalized predictive control.The simulation model is established by MATLAB.The wind window control system is simulated in the condition of disturbance and no disturbance.The simulation results show that the system can track the given signal,It also verifies that the prediction model established by PSO-SVM is applied to the generalized predictive control,which is effective in the control of gas concentration in the process of wind window control with good control performance.