Study On Intake System Control Algorithm Of Gasoline Engine

Some nonlinear factors exist in the electronic throttle control process, such as frictiontorque’s non-linear and return spring’s non-linear. Moreover, the traditional control methodsof intake system uses PID control algorithm, the control accuracy is low, response time islong, and the effect is poor. So traditional control system can’t achieve good control of thethrottle angle as the car has different torque requirements.Therefore, In order to solve this problem, this article takes naturally aspirated four-cylindergasoline engine as the research object, uses a torque-based engine control strategy; and takestorque demand as a central idea, for the purpose of accurate control of the mass of intake flowair, studies the intake control algorithms.Based on the mean value engine model theory, the intake system is divided into foursubsystems, and the mathematical model of each subsystem is modeled. The nonlinear factorvolumetric efficiency in the process of modeling is coped with the form of Map. Taking theelectronic throttle’s nonlinear factors into account and these factors are modeled in meanvalue model. Finally, the mean value engine simulation model is established in the Simulinksoftware, and the accuracy of the model has been verified on the en-DYNA simulation model.Open-loop control scheme and dual-loop control scheme of intake system are introducedin this article, the advantages and disadvantages of the two schemes are analyzed. Accordingto these two schemes, for the purpose of accurate control of the mass of intake flow air,single-loop control scheme has been designed, and the difficulty of the scheme is the solvingof the mass of flow air, a method has been proposed to solve this problem which is usingMAP to cope with the mass of flow air. Finally the mass of flow air is got by en-DYNA test.The conversion between engine torque and throttle angle is achieved via an inverseaerodynamics method.Four electronic throttle controllers are designed according PID control algorithm, fuzzy PID control algorithm, neural network PID control algorithm and fuzzy neural networkcontrol algorithm. The performance of the four controllers is analyzed by offline simulationtest on the mean value engine model; analysis and comparison of four controllers are achieved.Finally, the accuracy of the controllers has been verified on the en-DYNA engine simulationtest.