Optimizing And Control Of Double-in And Double-out Steel Ball Mill Based On Intelligent Method

Double-in and double-out steel ball mill is widely used in domestic and international power plant because of its high reliability and better coal type adaptability. However, its primary disadvantage is high specific consumption, usually account for 15 percent to 20 percent of station service system. And then coal level in steel ball mill is very difficult to be monitored and controlled. So how to reduce energy sources consuming of steel ball mill and increase control performance is a very important topic all the time.The working principle, characteristic and development status of double-in and double-out steel ball mill are introduced summarily.A method to monitor the coal bunker level using the acoustic signal picked up at specific location near the cylinder is presented, based on the sound radiation mechanism. The method of modulus maximum value is used to deal with the acoustic signal. The feature which can reflect coal bunker level variation is extracted from the signal. The result is that interferential noises are removed through using this method. The useful signal in the noise of mill is reserved and reflects effectively variation status of coal level in the mill. The difficult problem of measurement of coal level has been overcome.The work characteristics of steel ball mill are analyzed. The control scheme of coal level is established. The control object is a multivariable coupled system. The theory of PID control based on neural network is applied in coal level control system.Comparison between the compound control and the PID control is made in simulation concerning the robustness and ability of anti-disturb. Simulation demonstrates that the arithmetic of PID control based on neural network is better than the PID method. The arithmetic of PID control based on neural network can obtain small speed overshoot and strong anti-interference feature. It is shown by simulation that the system has the adaptive ability toparameter change of the object by suing this method. And has short transition time. This method solves the problem of controller deviation.