Research On Intelligent Control Method And Its Application Of Stepless Gas Flow Regulating System For Reciprocating Compressors

Reciprocating compressors are widely used energy-consuming equipment in petroleum,chemical,natural gas,metallurgical and other industrial fields.In order to meet the peak flow,the designed exhaust gas flow of the compressor is often higher than the actual required gas flow in the production process.Therefore,the compressor needs to be retrofitted with energy-saving gas flow regulation in order to avoid unnecessary waste of electrical energy.Based on the method of active opening the inlet valve in part of the stroke,the stepless gas flow regulating method is an effective compressor energy-saving regulation technology.However,this technology has been monopolized by foreign companies.In the process of developing domestic products of stepless flow regulation system,there are still many key issues that need to be solved.The existing regulation method is mainly applicable to low and medium speed compressors,when the speed is too high,the exxsting system may face problems such as loss of key signal,actuator response is not timely;In addition,the high speed reciprocating compressor gas regulation system is a multi-input multi-output nonlinear coupling system with time delay,so the traditional PID control method cannot meet the accuracy,speed and other regulatory needs,there is an urgent need to develop a more intelligent,more applicable regulation system.Another issue of concern is the research on the self-healing method of reciprocating compressor with air flow regulation.After long-term operation of the system,some components may have the risk of performance degradation or even failure,resulting in regulation failure,especially the performance of the electro-hydraulic actuator as the core component of the system is the most prominent,so the research on the self-healing regulation method of the system is very necessary.This thesis establishes a load-flow control mechanism model for stepless flow regulation from the principle of exhaust flow regulation of a reciprocating compressor with the method of active opening the inlet valve in part of the stroke The control method research is carried out for the problems of speed signal fluctuation and multi-cycle regulation of high speed compressors.Next,an intelligent control method for nonlinear,strongly coupled and time delay problems is established using tools such as neural networks,feedback linearization and Pade approximation,which effectively improves the regulation accuracy and stability of the system.In addition,this paper also develops a self-healing regulation method for the problem of partial actuator failure of the system,which provides a guarantee for long-term stable operation of the system.Finally,based on the above control method and self-healing regulation method,an integrated system scheme of reciprocating compressor gas flow regulation and online condition monitoring is designed,which has been verified on the experimental bench for its feasibility and effectiveness.It lays the foundation for the future practical application of this system in the industrial field.The major studies in this thesis are summarized as follows:Firstly,a theoretical analysis of the gas flow regulation principle of active opening the inlet valve in part of the stroke is carried out.The thermodynamic characteristics model of compressor cylinder,dynamic control volume model of working chamber,and dynamic discharge pressure model of buffer tank under gas regulating condition are established.Based on the sliding average filtering algorithm,an adaptive regulation scheme under unstable speed is proposed.The calculation model of the relationship between working load and actuator action angle under variable working conditions is proposed.Secondly,a multi-input multi-output strongly coupled system for stepless air flow regulation of high speed reciprocating compressors is analyzed.The correlation of each pairing ring was analyzed by computing the RGA matrix of the system,and the optimal pairing is obtained,whereby the control and controlled variables of each control channel are determined.The inverse decoupling control scheme of the high-dimensional system is investigated.Compared with the traditional diagonal decoupling scheme,the structure of inverse decoupling is simpler and suitable for higher-order complex process objects,according to which an internal-mode PID control method based on the Pade approximation for time-delay systems is proposed.Thirdly,to address the difficult problems of MIMO control systems such as multi-stage reciprocating compressor gas regulation with strong nonlinearity,coupling and parameter uncertainty in the regulation process,two intelligent control schemes are proposed with the evaluation indexes of regulation rapidity,low overshoot,anti-interference capability and robustness.The first scheme proposes an improved IMC-PID controller with BP neural network,which uses the nonlinear approximation capability of the neural network to adaptively adjust the unique parameters in the controller to compensate for the nonlinearity of the gas regulation system.The second scheme proposes an adaptive fuzzy sliding mode control method for a class of MIMO nonlinear systems with unknown parameters.Fourth,a study of the self-healing technology approach to the reciprocating compressor air flow regulation system is carried out.The influence of key parameters of the system on the regulation effect,including the actuator ejection and withdrawal displacement,and the solenoid valve performance degradation,is analyzed.Then,the self-healing control scheme for actuator failure or performance degradation is discussed,and the corresponding self-healing scheme is established for different types of failures to ensure the efficient and stable operation of the regulating system.Finally,based on the basic regulation principle of stepless air flow regulation of high speed reciprocating compressor,an integrated system of stepless air flow regulation and online condition monitoring is designed.A testbed prototype was built to experimentally validate the intelligent control method proposed in this paper.The effectiveness of the proposed compressor air volume regulation method is illustrated by the practical application cases of the two systems in industrial sites.