Research On Method Of Hydraulic Equipment Condition Monitoring & Fault Diagnosis Based On Soft-sensor

With the continuous improvement of automation, size and complexity of hydraulic equipment is increasing, equipment failures and the reasons for failure are becoming more complex, there is an urgent need to improve system reliability and safety of running an effective means and measures. Through the equipment condition monitoring and diagnosis technology, comprehensive use of various fault diagnosis of new technologies and new methods of hydraulic equipment operating status and fault-line real-time monitoring and diagnosis will be to increase the hydraulic system operating reliability and safety of an effective means. This requires monitoring of hydraulic equipment in the process parameters, and some parameters can not be used directly or difficult process of detection instruments or sensors to measure, and soft-sensing technology can effectively solve these problems.First of all, introduced the basic concepts and soft measurement commonly used method of modeling, using the method of modeling analysis of the hydraulic system of gear pump - the proportion relief valve model, and the use of Simulink simulation model for the follow-up study condition monitoring system Lay the foundation for through the mechanism of the hydraulic system and find the gas content of the auxiliary variables, and then through the inter-related soft measurement method to study the gas content of the measurement, on the basis of this use of RBF neural network for the gas content qualitative analysis.Second, the study of the strain detected in the application of the hydraulic system and, through experiments and used pressure sensors, on the basis of this detection technique using contingency measure hydraulic pressure system of signals, combined with modern technology on a non-linear process Signal and fluid pressure pipeline flow state relations.Finally, the SOM discusses in detail the basic neural network models and algorithms, on the basis of the output of the SOM visualization of the results, the establishment of the SOM visual network model and use SOM visual model of a hydraulic power Classification system running.