Electric Two-Wheeler Multi-Dimensional Data Remote Monitoring System Based on CAN Bus

In practical applications, electric two-wheelers face challenges such as insufficient positioning accuracy and limited active safety. To address these issues, this paper proposes a CAN bus-based multi-dimensional data remote monitoring system utilizing an STM32F407 main controller integrated with a CAN transceiver and EC20 LTE 4G module. This integrated solution enables comprehensive data acquisition, processing, and long-distance transmission of critical parameters. The system implements a novel noise suppression algorithm combining Ensemble Empirical Mode Decomposition (EEMD) with optimized wavelet thresholding to enhance GPS positioning accuracy, while multi-dimensional operational data is transmitted to the OneNET platform. This architecture facilitates dynamic real-time monitoring of 32 driving parameters through a web interface and provides 30-day local storage via SD card backup, significantly improving active safety capabilities. Experimental validation confirms stable system operation with a 0.32 Hz refresh rate and approximately 300 ms latency. Most notably, in complex urban environments, the proposed algorithm elevates GPS signal-to-noise ratio (SNR) to 28.6856 dB, outperforming traditional denoising methods including soft threshold, hard threshold, and EMD-wavelet hybrid approaches, as well as existing commercial systems.