FLOWX Engineer 86-21-54150349
At present, the country has put the instrument industry into the list of "11th five-year" plan and the first one catching the focus is the development of the industrial automation instrument. Since pneumatic actuators inventor invented the actuator, it has been as a focus on the development of high technology products and has been widely used in various industries. And electric actuators system modeling research is the key theoretical basis of a research and development of high performance index of the intelligent positioner precise positioning control algorithm, the control valve optimization design theory, the respect such as actuator fault diagnosis method, which is also an important theory problem of the development of electric actuators technology.
There is often a big deviation between the mechanism of the model established for them and the actual situation because of the complex internal structure of the pneumatic valve positioner, so actuator inventor has put forward a step response identification approach based on partial data. First, use step response identification methods, such as the characteristics of the area of law, rising curve method and the least-squares method, to identify the step response data of pneumatic position system, and as a result, the larger deviation between transfer function model and the actual step response curve can be seen. Based on the detailed analysis of actuator pneumatic positioning system working characteristics, actuator inventor greatly improve the accuracy of the model identification by using the step response identification approach based on partial data.
On the basis of establishing the actuator pneumatic positioning system model, actuator inventor makes analysis and diagnosis of the common faults of valve positioner and actuator. First of all, actuator inventor finds the possible fault components in the valve positioner and actuator, and then analyzes the different failures, making the fault simulation by using BP neural network in Matlab/Simulink platform so as to realize the fault diagnosis of actuator pneumatic positioning system, and to verify the correctness of the BP neural network algorithm that is used.
Actuator inventor makes modeling experiment of actuator pneumatic positioning system, and the Matlab/Simulink platform of the actuator pneumatic positioning system is used to simulate and diagnose the common breakdown. All these jobs will lay a certain foundation for the next step of the research, but there are many problems remaining to be further researched and this paper still needs further research from the following aspects:
(1) Actuator inventor puts forward a step response identification approach based on partial data of actuator pneumatic positioning system step response data analysis.
(2) During the diagnosis process of common faults in various parts of actuator, actuator inventor uses the seven electric actuators system parameters to do training on the network, choosing Levenberg - Marquardt algorithm for network training and learning. But there are certain errors before we get precise results, which is subject to further analysis and research to reduce experimental error.
(3) actuator inventor has only carried on the analysis and modeling the several common faults of each part of actuators but has not made further in-depth analysis and research on other faults of actuators, so the research content about actuator fault is slightly deficient.