Abstract: According to the reverse osmosis membrane fault problems in reverse osmosis water desalination system, a fault diagnosis method based on support vector machine (SVM) was introduced for fault diagnoses. To solve the problem of parameter optimization in SVM, an improved chaos particle swarm algorithm was proposed. The introduction of Chaos theory to particle swarm optimization algorithm may not only enhance the diversity of the population and particle global search ability, but also improve the convergence speed and accuracy of the particle swarm algorithm. The optimized SVM model was applied to the fault diagnosis of reverse osmosis water desalination system. The simulation results show that the improved SVM classifier can effectively diagnose the reverse osmosis membrane fault diagnosis and achieve a higher diagnostic accuracy and efficiency.

Key words: reverse osmosis water desalination, support vector machine, chaos particle swarm optimization, fault diagnosis