基于改进S变换的煤岩界面超声反射信号处理

煤岩识别技术是实现煤矿综采工作面智能化和无人化的关键技术之一。准确分解超声波检测煤岩界面的混叠严重的回波信号方法是研究的重点和难点。分析不同窗函数调节参数对时频分辨率的影响，提出一种改进的S变换，建立非对称高斯调制模型的多重回波混叠信号，采用改进S变换方法对超声波检测煤岩界面的仿真与实验反射回波信号进行处理，采用时间幅值包络曲线方法分离混叠回波。研究结果表明，改进S变换方法具有更好的时频分辨率及时频聚集性，可有效分离煤岩界面的混叠反射回波，获得较准确地超声波煤岩界面反射回波的到达时间，从而，为煤岩界面的识别提供技术支撑。

Processing of ultrasonic reflection signal from coal-rock interface using modified S-transform

Coal rock interface recognition is one of the key technologies for the intelligent and unmanned production in fully mechanized coal mining face.The separation of overlapping ultrasonic reflection signal from coal-rock interface is highly essential.After analyzing the impact of existing window functions,a modified S-transform (MST) is proposed.Multiple overlapping echoes,gained by asymmetric Gaussian chirplet model,are processed with the proposed MST.The arrival times of overlapping echoes can be obtained accurately using the envelope of time-frequency distribution.Simulation results show that the MST performs better time-frequency resolution and time-frequency concentration than the standard S-transform does.Subsequently,the proposed MST algorithm is applied to separate ultrasonic reflection wave and shows correct arrival time information.Therefore,the proposed method has been applied on coal seam thickness measurement to show its effectiveness and provides a basis for the coal rock interface automatic recognition.