Intelligent Fault Diagnose System Of Proportional Direction Valve Based On Artificial Immune Algorithm

Electro-hydraulic proportional system is an important component of the control field, has the characteristics of large power, fast response, and high precision, which has been widely used in the field of industrial production. Electro-hydraulic proportional valve is the key component of Electro-hydraulic Proportional system, and has relatively high failure rate in the system; its service condition and performance will have a direct impact on the whole system performance. Therefore, a practical study of intelligent fault diagnosis method is very important, to predict and control the fault of electro-hydraulic proportional valve may occur, correct and rapid diagnose to the fault has happened, and provide a viable method of exclusion faults.This paper studies in depth the key technology of the electro-hydraulic proportional valve, on the base of collecting broadly the relate literature and works, combined with the theories of modern fault diagnosis,recognizing the fault model,artificial intelligence and computer science, which relies on a new kind electric-hydraulic static character test station .1. Study carefully on the main fault mechanism of electrical feedback proportional valve, such as the: abrasion, cavitation erosion, spool chucking, elevated temperature.2. Use the laboratory of proportional valve experiment platform, and artificial simulate the fault, including the pressure curve of normal, and proportional electromagnet failure, one side of the spool chucking, t one side of the centering spring fracture, one side of the centering spring with debris, spool wear serious. Then extract the data of pressure experimental curve and normalize the data, preparing for fault diagnosis.3. Study the combination method of artificial immune algorithm and the artificial neural network, and set up a suitable immune neural network for electric feedback proportional valve Fault Pattern Identification model. The model's input layer neurons number is the sample's dimension, the output layer unit number is the number of failures had been simulated, the initial number of hidden units is one, which can be calculated based on the changes of the network, so that can keep the network working best.4. Use the Object-oriented, visual programming language Visual C++, developed the immune neural network fault diagnosis system for the electro-hydraulic proportional valve. The system is based on parameterize design ideas, with commonality; can be used in a variety kind of proportional valve fault diagnosis.