 This paper proposes an improved artificial rabbit optimization algorithm, HERO, to optimize the parameters of a support vector machine, SVM, in order to achieve accurate identification of vibrations from hydraulic units. The VMD method was used to extract features from the vibration signals, and these were then fed into the IRO-SVM model to classify and identify the vibration states of hydraulic units. Compared to other algorithms such as ROSVM, ASO, SVM, PSOSVM, and WSVM, the IRO-SVM model achieved a higher average identification accuracy of 97.78 percent, which was 3.34 percent higher than the closest ROSVM model. This shows that the IRO-SVM model has higher identification accuracy and better stability, making it suitable for use in the identification of vibrations from hydraulic units. This article was authored by Qin Jiaqiao, Li Yingwang, Wei Guazhao, and others. We are article.tv, links in the description below.