Study On The Disturbance Performance And The Intelligent Control Technology Of The Regulatable Gas Lubricated Seal

The regulatable gas lubricated seal with intelligent control capability is an important part of the development of the high-end compressor units. Comparing the general gas seal (hydrodynamic gas and labyrinth seal etc.), the regulatable gas lubricated seal (short for R-GLS) not only has the advantages of long operational life and low leakage rate, but also owns the characteristics of regulation and intelligent. It can realize the intelligent control to guarantee the security of the host system, this study direction exceedingly meet the requirement of'made in China 2025'plan. Combining the numerical simulation and experiment, the disturbance performance of R-GLS, intelligent control program and the regulate performance of regulatable gas pressure system (short for R-GPS) are studied and discussed comprehensively in this paper, in order to promote industrial application of the R-GLS and the intelligent control method.Using the SST k-? turbulence models, the steady and unsteady numerical simulation were carried out for the flow field between the seal face of R-GLS by the FLUENT. The effect factors of the entrance effect were studied, the influence of the different structure of face and operating parameters have been obtained, and the forming reason of the entrance effect were also analyzed. According the surge condition of the compressor, the disturbance performance of the R-GLS were analyzed, the results of the study can provide the references and basis for design and the online control.Aiming at the problems of the numerical calculation method application during the control field, and enlightened by the BP neural network, a kind of intelligent calculation method, which can calculate the parameters of the R-GLS, was proposed. The Particle Swarm Optimization was also used to optimize the program; the accuracy of the program was improved finally.350 samples was obtained and used to train the intelligent control program, the results verified that the PSO-BP neural network can be used to realize the intelligent calculation method of the R-GLS. Intelligent calculation method proposed makes it possible to realize the intelligent control technology, and this method can provide the guidance information for the R-GPS.Proposed a kind of gas regulator system based buffer gas cylinder and solenoid control valve, and according the state equation of ideal gas and flow conservation the simulation system of R-GPS was established by Lab VIEW. The influence of the different structure and operating conditions on the control performance of the R-GPS were also analyzed. Referencing the central composite circumscribed design (short for CCC) experimental theory and aiming at the response time and the disturbance range, the regulatable gas pressure system was optimized by analyzing the response surface. The results provide the guidance information for designing the R-GPS and lay a foundation for realizing precision control of R-GLS.The R-GLS and R-GPS and the monitoring system for experiment investigation was designed and established, the control program was written by Lab VIEW for experiment, the closed loop control experiment of R-GPS, the hydrostatic opening experiment and the intelligent control experiment of the R-GLS have been done. By means of comparing the results of the experiment and the simulation results, the correctness of the simulation system of the R-GPS and the feasibility of the intelligent control method, which based on the BP neural network, were demonstrated.The study of the R-GLS laid a foundation for improving the control precision of R-GLS. And the correctness and the feasibility of the intelligent control method of the R-GLS were demonstrated, the research results can promote the industrial application of the R-GLS and intelligent control method.