Research on electric control detection of battery motors in new energy vehicles

Under the background of global "double carbon" strategy and energy transformation, the new energy vehicle (NEV) industry is developing rapidly, and the performance and reliability of its core components, such as battery, motor and electronic control system, are very important for vehicle safety and service life. However, there are some problems such as hysteresis and one-sidedness in the current detection technology of three-power system, and it is difficult for traditional detection methods to deal with multi-parameter coupling failure under complex working conditions. In this paper, the detection problem of NEV three-electric system is studied, the cooperative detection technology system of three-electric system is constructed, the coupling mechanism of electric-thermal-mechanical fault is revealed through multiphysical field coupling modeling, and the multimodal feature extraction algorithm based on deep learning is developed to realize cross-system data fusion and collaborative diagnosis. In this study, the improved equivalent circuit model (ECM) and heat conduction equation are used to couple the battery system, and the electromagnetic-mechanical coupling model of the motor is established based on Maxwell stress tensor method, and the electro-thermal fatigue characteristics of the electronically controlled IGBT module are analyzed by using Coffin-Manson fatigue life prediction formula. At the same time, a multisource heterogeneous data acquisition system is constructed, which covers the key detection parameters of battery, motor and electronic control subsystem, and fault features are extracted by dynamic time warping (DTW) and wavelet packet decomposition (WPD). The collaborative diagnosis model based on deep learning introduces cross-system attention mechanism, dynamic graph neural network and transfer learning optimization technology, which significantly improves the diagnosis accuracy and early warning ability. The experimental results show that the cooperative detection method proposed in this paper achieves 93.6% and 88.9% in the accuracy of single fault identification and coupled fault identification respectively, which greatly reduces the false alarm rate compared with traditional methods, and shows high reliability in the detection accuracy of key parameters, which provides a new technical path for efficient detection and fault early warning of NEV three-power system.