Research On Management Algorithm And Optimal Sensor Placement Of Coal-Rock Interface Recognition

This paper has mainly studied the optimal sensor placement and management algorithm of coal-rock interface recognition based on finite element dynamic analysis. Finite element dynamics analysis of cutting part of shearer have been made to study its vibration characteristics, including the transient response from the perspective of time-domain and the harmonic response from the perspective of frequency–domain, it also analyses the conflict signal of cutting states based on weight evidence combination, and have achieved good results.The development of coal-rock interface recognition is overviewed, the latest technology and methods is summed up. For the current existing problems, the method of coal-rock interface recognition based on multi-sensor management was proposed. Sensor management can take full advantage of multi-sensor respective or joint operation, greater access to comprehensive information on the object being detected by sensor management system for screening and assessment,these information are accurately and reliably backed to the data fusion system to complete coal-rock interface recognition automatically.The principle of coal-rock interface recognition is to pick up the response signals of cutting force of shearer by multi-sensor and process these data to achieve state of cutting coal or rock. So the method of extracting state signal of cutting coal or rock is determined, and five types of sensor are installed. Meanwhile, the physical testing system of shearer is established based on the methodsIn view of optimization of the sensor's measuring point position and selection of sensor's performance closely related to the dynamic characteristics of cutting part of physical simulation system, the physical simulation system for finite element dynamic analysis has been made in this paper. Transient dynamic analysis for the cutting part of shearer by finite element software ANSYS has been made, the law of vibration displacement is studied from the perspective of time response. The vibration points of sensors are optimized to ensure maximum amplitude output signal. Meanwhile, dynamic harmonic response has been made, the vibration response characteristics of cutting part are analyzed from the perspective of frequency response. Study not only validates the maximum vibration displacement region of nodes, while the frequency range of the selected vibration sensors was defined. Then it satisfies the undistorted condition of the measurement, also avoids the abnormal output affected by resonance.For the conflict signal of coal-rock interface recognition, weight evidence theory has been applied to process these conflict signals. With the results processed by D-S evidence theory is compared, modified algorithm can solve the impact of the conflict signal. The application of sensor management prevents the impact of data fusion affected by sensor failure, and improves the performance of data fusion system, as well as optimization of resource, so both effectiveness and accuracy of coal-rock interface recognition was greater improved.