Abstract: To solve the problem that it is difficult to effectively extract the mechanical fault features of high-voltage circuit breakers from the tripping vibration signals, we propose a method based on optimizing the parameters of the variational modal decomposition (VMD) and extracting the energy entropy (EE) as the fault eigenvalues. Firstly, the parameter combination K, α of the VMD algorithm is globally optimized using the Sticky Mushroom Optimization Algorithm (SMA) with the minimum envelope entropy as the fitness function. Secondly, the parameters of the VMD algorithm are set to decompose the signal according to the optimization results to obtain K intrinsic mode function (IMF). Then, the energy entropy of each modal component is calculated, the modal components that are more relevant to the original signal are filtered with the help of the correlation coefficient, and the feature vectors are constructed and randomly divided into the training set and test set. Finally, the training set is input into the support vector machine (SVM) to train the classification model, and the trained model is used to classify the samples in the test set. Moreover, experimental validation is carried out on a 12 kV high-voltage circuit breaker. The classification results show that the state recognition accuracy of the proposed SMA-VMD-EE model is 95%, which is improved compared with the accuracy of VMD-EE, PSO-VMD-EE, and SMA-VMD-SE models, and the effectiveness and feasibility of the proposed method are verified.