Recognition and prediction of ground vibration signal based on machine learning algorithm

A Data Compression Method for Wellbore Stability Monitoring Based on Deep Autoencoder

The compression method for wellbore trajectory data is crucial for monitoring wellbore stability. However, classical methods like methods based on Huffman coding, compressed sensing, and Differential Pulse Code Modulation (DPCM) suffer from low real-time performance, low compression ratios, and large errors between the reconstructed data and the source data. To address these issues, a new compression method is proposed, leveraging a deep autoencoder for the first time to significantly improve the compression ratio. Additionally, the method reduces error by compressing and transmitting residual data from the feature extraction process using quantization coding and Huffman coding. Furthermore, a mean filter based on the optimal standard deviation threshold is applied to further minimize error. Experimental results show that the proposed method achieves an average compression ratio of 4.05 for inclination and azimuth data; compared to the DPCM method, it is improved by 118.54%. Meanwhile, the average mean square error of the proposed method is 76.88, which is decreased by 82.46% when compared to the DPCM method. Ablation studies confirm the effectiveness of the proposed improvements. These findings highlight the efficacy of the proposed method in enhancing wellbore stability monitoring performance.

Self-Powered Long-Life Microsystem for Vibration Sensing and Target Recognition

Microsystems play an important role in the Internet of Things (IoT). In many unattended IoT applications, microsystems with small size, lightweight, and long life are urgently needed to achieve covert, large-scale, and long-term distribution for target detection and recognition. This paper presents for the first time a low-power, long-life microsystem that integrates self-power supply, event wake-up, continuous vibration sensing, and target recognition. The microsystem is mainly used for unattended long-term target perception and recognition. A composite energy source of solar energy and battery is designed to achieve self-powering. The microsystem's sensing module, circuit module, signal processing module, and transceiver module are optimized to further realize the small size and low-power consumption. A low-computational recognition algorithm based on support vector machine learning is designed and ported into the microsystem. Taking the pedestrian, wheeled vehicle, and tracked vehicle as targets, the proposed microsystem of 15 cm³ and 35 g successfully realizes target recognitions both indoors and outdoors with an accuracy rate of over 84% and 65%, respectively. Self-powering of the microsystem is up to 22.7 mW under the midday sunlight, and 11 min self-powering can maintain 24 h operation of the microsystem in sleep mode.