SDA-based Neural Network Approach for MWD Mud Pulse Signal Recognition

Abstract

Abstract. The low signal to noise ratio (SNR) of the detected mud pulse signal leads to difficult to recognize signal at once. The recognize accuracy is low. So a stacked denoising autoencoder recognition model was constructed. Combining with the drilling mud pulse signal, the recognition performance of the typical data set is analyzed and tested. The proposed method of recognizing mud pulse signal enhances the SNR of output signal by using signal detection method. And then we take the output detected signal as signal classification attribute. Test results also show that the proposed method is suitable for mud pulse signal recognition, and it has strong ability to extract features from samples automatically and robustness. Performance is better than pattern matching and support vector machine and other recognition methods under the same SNR.