Currently, motors have been widely applied in various fields such as industrial production and aerospace, serving as indispensable power equipment. The health status of motors is directly related to the safety and reliability of equipment, affecting the efficiency and cost of production and operations. Consequently, the demand for motor health assessment is increasingly growing. However, traditional research on motor health assessment largely relies on a single signal, presenting considerable limitations. Additionally, different types of signals exhibit significant variability under various operational conditions. Therefore, this paper conducts research on motor health assessment technology based on the fusion of electrical and vibration signal information. It proposes signal feature extraction based on EWT-Hilbert and joint information entropy, as well as an improved motor health assessment method based on the Gaussian Mixture Model, achieving high precision and robustness in motor health assessment.