Reseach Of Detection Equipment Of Ball Mill's Audio Features Based On DSP

Cement grinding is the final process of cement production,and also the most power consumption process, the ball mill has been difficult to solve optimal control problems because of its multi-variable, nonlinear, parameter variability and model uncertainty. To achieve optimal control of the mill is the effective way of production efficiency improvement and energy consumption reducement. how detect the mill's current load condition accurately is the the key of the optimal control of the mill.This article analyzes the current status of cement grinding load detection reseach by domestic and foreign researchers, compared to the use of current testing, import and export of differential pressure testing, pressure testing, vibration testing, power detection testing. An new type of detection methods for detecting the mill load is proposed after extract and summarize the diversity of detecting method, this method is used for extract the characteristics of the spectrum of audio signal in real-time detection of ball mill, and make the foundation of achieving optimal control of mill.The audio features of the mill in different conditions has the visible distinct. Grinding sound trend of several typical working conditions is collected after a large number of experiments,Basic featuers of mill sounds in typical working conditions is also summerized . We can realize the real-time detection of mill loading by extracting audio featuers after give denoising, smoothing filtering, signal processing,which we can get to reflect the real loading.This issue utilize the DSP processors using its high precision, fast speed charicterise to capture the audio signal. The audio sensor is used to capture the audio signal,through the A / D conversion, filtering,amplification to make the signal suitable for the DSP controller. The design based on the TMS320VC5502 chip, full use of its high performance, low power consumption, and suitable for audio detection features, its peripheral circuits(Audio detection circuit, power circuit, clock circuit, reset circuit, and JTAG circuit) and software is designed in this paper. FFT algorithm is used to caculate the energy of different frequency band. The useful audio information is extracted that can really mirror the loading conditions within the mill and the real-time audio information can also displayed during mill works, and it also has the ability to conduct real-time communication with the DCS system, besides, it can also as a separate measuring instrument of mill load conditions.Mill load state is reflected timely and accurately transferred to optimal controller,the controller can make the mill working in the optimal condition by adjusting the amount of feed and hour output to avoid abnormal condition. This equipment has the advantage of simple, reliable, easy to maintain which can help us to the realization of automatic control systems.Datong, Shanxi Qifengshan cement factory grinding # 2 is the background of this paper,We analysed the real-time load of ball mill using the electricity ear signals ,the current of the the main motors, combined with experience of workers for testing.which can achieve the goal.The DSP instrumentation has good application prospect, and can also provide the important experience of the design of detection equipment used in power plants, chemical factory and other interprise where use grinding system.