Development And Model Prediction Of Electronic Control System For Single Cylinder Gasoline Engine

China’s automobile industry has become a pillar industry in the field of microelectronics application.However,while the automobile industry is booming,domestic enterprises still encounter great obstacles in the development of electronic control systems for automotive engines.On the one hand,not only the vehicle-specific chip is monopolized by foreign leading companies,but the supply and production costs of the chip are also difficult to control under the influence of many factors.On the other hand,many domestic enterprises can only carry out simple calibration work in the development of electronic control system,making that the product development is restricted to others.In this paper,a 125 m L small displacement single cylinder engine was taken as the control object,and the software and hardware design were carried out based on the domestic chip as the main controller.The underlying control model and application layer control strategy were built to meet the basic development requirements of the electronic control system.The main research contents are as follows.(1)After analyzing the single-cylinder motorcycle engine,and the circuit analysis of the electronic control system according to the functional requirements,this paper completed the hardware design of the electronic control system.In the software design,the relevant functional modules are configured and initialized in the project,and the functional interfaces of the underlying control model and the application layer control strategy are left.(2)According to the engine principle,the engine bottom control model was built to ensure the correct calculation of the engine bottom state parameters and the reliable output of the actuator control instructions.The realization of the underlying control model included the conversion of sensor data,the calculation of speed,the judgment of compression top dead center,the output of ignition command and injection command.(3)Combined with the relevant data of electronic control system development,the control objectives were formulated for different working conditions,and the corresponding application layer control strategy was built for each working condition of the engine.The control strategy model mainly included air-fuel ratio control and ignition control under starting condition,idle condition and transient condition,and realized the calculation function of fuel injection quantity and ignition advance angle algorithm.(4)The underlying function code in the software design is combined with the code generated by the model built on the Simulink platform to complete the software program development of the electronic control system.Using the output signal of the simulated engine,the engine signal acquisition module,the state parameter calculation module and the injection and ignition output effect in the control strategy model of the underlying control model are tested to verify the real-time and accuracy of the underlying control model and the application layer control strategy.Finally,the developed electronic control system is tested on the bench and real vehicle to complete the verification of the electronic control system.(5)Aiming at the problem that the narrow-domain oxygen sensor cannot obtain the specific air-fuel ratio value in the actual development process,the artificial intelligence method was used to combine the advantages of long-term and short-term memory neural network and attention mechanism,and the LA-LSTM air-fuel ratio prediction model was built to predict the actual value of air-fuel ratio.Finally,the model was further compared with LSTM and BP models to prove the superiority and effectiveness of the LA-LSTM model in the air-fuel ratio prediction task.