Research On AFR Accuracy Control For Small SI Engine Based On Model-Based Design Method

The small SI engines have a lot of advantages, such as compact structure, reliable working and convenient operating. They are widely applied in the industry and farming field. However, it is hard for these small SI engines to meet the increasingly strict emission rules due to its rough controlling. The CO,HC and NOX emission can be significantly reduced and meet the emission rules by using ECU precise controlled AFR as well as three-way catalytic converter. In this paper, an AFR accurate control system without oxygen sensor is designed according to the working characteristic of small SI engines. The Model-based design method is borrowed and used in the research. It is of great importance for the independence research and development on AFR accurate control of small SI engine and environment protection.The development process is divided into two parts: the function design and the software design. The function design process is as follows. Firstly, the variant subsystem plant model of non-road SI engines are built in the environment of Simulink, including intake air gas system model, combustion model, engine speed model, AFR model and oxygen sensor model. The model structure parameters are calculated accurately by generic algorithm and test data. And then, design the AFR accurate control algorithm model based on the engine model. The AFR control algorithm model is made up of three modules; AFR calculation module, mode schedule module and PI controller module. By collecting engine speed signal, load signal, intake pressure signal, intake temperature signal and engine body temperature signal, the control algorithm calculate the AFR number real-time according to the engine current conditions and the working range of AFR closed-loop control. Then, current oxygen sensor voltage number is got by oxygen sensor model and input to the PI controller. Finally, the PI controller module will output the current condition AFR fuel correction. For the software design process, the first step is to perform the unit test and integrated test of controller model in detail. After that, Simulink model code generation parameters are configured to automatically generate the C code of AFR controller model. Then, under the CodeWarrior IDE compiling environment of Freescale processor chip, the interface functions of AFR control features need to be creating again to integrate the generated code into the existing ECU software architecture. At last, the compiler compiles the C code into the binary file that can be executed by SCM(single chip microcomputer).AFR closed-loop control algorithm without oxygen sensor will be verified by engine test bench experiment. By using the common calibration software, controller parameters can be calibrated and optimized. Experiment results show that AFR closed-loop control algorithm without oxygen sensor might work normally in the ECU software and control AFR accurately in the range between 14.31 and 15.01.