Research On Centrifugal Compressor Output Pressure Optimal Control Method In Natural Gas Pipeline System

Centrifugal compressor is one of the important equipment in the natural gas transmission pipeline system,which output pressure control performance has significant effects on the stability and energy saving of the pipeline system.Since the natural gas transmission pipeline system is operating under transient conditions,the selection of output pressure controller setpoint based on the stable state computation is not reasonable,which may cause huge energy consumption and wide-range fluctuation of output temperature.Additionally,with the characteristics of nonlinear and time-varying of centrifugal compressor system,the regular PID centrifugal compressor output pressure controller has poor control effects.To solve these problems,this thesis mainly finished the following innovative work.Firstly,in order to reduce energy consumption and output temperature fluctuation,an output pressure dynamic optimization method based on an improved particle swarm optimization algorithm is proposed to guide the selection of output pressure controller setpoint.An output pressure dynamic optimization model with the goal of minimizing energy consumption and reduce output temperature fluctuation is built to reduce energy consumption and output temperature fluctuation.Through designing an improved particle swarm optimization algorithm based on a self-adaptive penalty function constraint handling strategy and a parameters self-adaptive particle iteration updating strategy,the optimized centrifugal compressor output pressure is calculated out stability and reliability.Simulation results showed the validity of the method.Secondly,this thesis proposes an improved adaptive sliding back-stepping control method for centrifugal compressor output pressure under normal working conditions in the presence of system steady state tracking response unstable and some uncertain disturbance.Based on the analysis of dynamic operating characteristics of the system,a time-varying parameter model is proposed for the centrifugal compressor and the downstream pipeline integration system to improve the accuracy of dynamic control model.Meanwhile,an improved adaptive sliding back-stepping controller is designed with the characteristics of model parameters time-varying and disturbances uncertain.By designing a kind of adaptive law for the unknown parameters and the upper boundary of the uncertain disturbance,the problem of unstable steady state tracking response is avoided.Through integration sliding mode control technology,this controller has a strong disturbance resisting ability.Simulation results showed the validity of the method.At last,with problems of large dynamic response overshoot,serious fluctuation of the control input parameter and modeling of the system is difficult under special working condition,this thesis proposes an improved ELM_PID control method.In order to improve the transient response performance of the controller,the PID parameters are adjusted online through using the ELM neural network.Additionally,a control input signal soften handled technology,including a control increment penalty coefficient and a control signal weighting factor based on fuzzy theory,is proposed to reduce the wide-range fluctuating of control input signal.This strategy is allowing the controller to have a better tracking performance,a small dynamic response overshoot and a better robustness in various operating conditions.Besides,the control input signal is smoother than before.Simulation results showed the validity of the method.