Research On Anti-surge Control Of Centrifugal Compression

As a clean and high quality energy,natural gas is increasingly favored by the governments.However,natural gas resources in our country are mainly distributed in the Midwest,away from the eastern consumer market.In order to support long distance gas transmission,it is necessary to set up compressor stations on the natural gas transportation pipeline to make up for the loss of pressure caused by the gas flow.The main pressurized equipment in the compressor station is centrifugal compressor.Surge is the intrinsic characteristics of centrifugal compressor,which can bring severe damage to the compressor.Surge also can limit operating range of the compressor,and then limit compressor’s ability to adapt to gas transmission volume with great changes throughout the year.Therefore,it is significant to study how to prevent the centrifugal compressor from being damaged by surge,and make compressor has a wider operating range at the same time.This thesis mainly studies centrifugal compressor PCL803 in Kongquehe compressor station.The internal cause and external cause of the surge is analyzed under the premise of the working principle of the centrifugal compressor and regulating methods is studied.Based on Least Squares Support Vector Machine,centrifugal compressor’s modeling method is proposed to solve some problems in mechanism modeling.These problems are caused by the complexity of the gas flow in the compressor,including the model is too complex and practicality is not high.Two nonlinear models are established according to control requirements.One is pressure ratio’s two inputs and one output model,the other is mass flow’s two inputs and one output model.The simulation results show that the centrifugal compressor’s LS-SVM model has advantages of good generalization ability,simple structure,convenient modeling and high precision.Nonlinear predictive control is proposed to solve the problem of narrow operating range in the traditional PID anti-surge control.Based on rolling optimization strategy of predictive control,the reference trajectory can be followed by the operating point as smoothly as possible,which can avoid surge phenomenon caused by the overshoot of the controller as the operating point changes suddenly.In addition,whether operating point will cross the surge control line can be judged in advance,so the control signal can be send out ahead of time,and the influence of actuator’s adjusting time can be reduced.On the control scheme,setpoint predictive control is adopted to control pressure ratio,single interval predictive control is adopted to control mass flow,and adaptive quantum particle swarm optimization algorithm is adopted to realize the rolling optimization of nonlinear predictive control.Compared to the traditional anti-surge control when the distance between surge line and control line is reduced from 10%to 5%,the predictive anti-surge control can regulate outputs more smoothly and make operating range of centrifugal compressor wider.As a result,the predictive control is a kind of suitable method for anti-surge control of centrifugal compressor.