Study on chaotic signal analysis and PSO-SVM diagnosis model for braking stability under exogenous environmental intrusions
Authors:
- 30 January 2026
- journal article
- Published by SAGE Publications in Journal of Vibration and Control
Abstract
This paper proposes a signal diagnosis model that integrates Particle Swarm Optimization (PSO) and Support Vector Machine (SVM), aiming to address the problem of strategic regulation and optimization of braking systems under complex environments. This study established a brake block–metal disc experimental platform and systematically investigated the mechanisms by which exogenous environmental intrusions (dry, rainy, particle, muddy, and icy) influence the surface morphology of friction pairs and braking stability, by integrating surface analysis techniques and vibration signals. Mathematical tools, including the number of Intrinsic Mode Function (IMF) components and autocorrelation coefficients, were introduced to quantitatively characterize the complexity and chaotic features of the signals. The results indicate that different environmental conditions significantly alter the morphological characteristics of the wear interface, and that there exists a clear negative correlation between the number of IMF components and the autocorrelation coefficient. A multidimensional diagnostic model constructed based on these mathematical features enables high-precision identification of time-varying friction signals under various environmental conditions, achieving an identification accuracy of 97% on the test set, which represents a significant improvement over traditional model. The research provides a theoretical basis for the adaptive control of braking systems under complex environments. The proposed interdisciplinary approach, combining tribological measurement and mathematical analysis, holds significant engineering value for the development of intelligent braking technologies.Keywords
Funding Information
- National Natural Science Foundation of China-NSAF (U2330202)
- National Natural Science Foundation of China (52175162)
- National Natural Science Foundation of China (51805086)
- Fujian Provincial Technological Innovation Key Research and Industrialization Projects (2023XQ005)
- Fujian Provincial Technological Innovation Key Research and Industrialization Projects (2024XQ010)
- National Independent Innovation Demonstration Platform Project of Fujian Province (2024QZFX07)
- Fujian Provincial Natural Science Foundation (2022J01073)
- Fujian Provincial Science and Technology Plan Regional Development Project (2024H4017)
- Fujian Provincial Science and Technology Plan Regional Development Project (2025H0024)
This publication has 41 references indexed in Scilit:
- SENSE: software effort estimation using novel stacking ensemble learningInnovations in Systems and Software Engineering, 2024
- Local maximum instantaneous extraction transform based on extended autocorrelation function for bearing fault diagnosisAdvanced Engineering Informatics, 2024
- Machine learning methods for developments of binding kinetic models in predicting protein‐ligand dissociation rate constantsSmart Molecules, 2023
- Full‐scale modeling of chemical experimentsSmart Molecules, 2023
- Magnetic‐optical dual functional Janus particles for the detection of metal ions assisted by machine learningSmart Molecules, 2023
- Salp Swarm Algorithm: A bio-inspired optimizer for engineering design problemsAdvances in Engineering Software, 2017
- Particle swarm optimization (PSO). A tutorialChemometrics and Intelligent Laboratory Systems, 2015
- Ant colony optimization for continuous domainsEuropean Journal of Operational Research, 2008
- A new model of the critical crack size for low temperature brittle fracture of structural steel and its experimental verificationEngineering Fracture Mechanics, 1994
- The third-body approach: A mechanical view of wearWear, 1984