The Dynamic Adjustment Of Car's Operation Status Domain Based On Complicated Road

To determine the state of car running conditions control has been a major problem in automotive ECU design. Traditional method to determine vehicle condition is mainly based on the results of bench test data and based on that result to determine conditions of car. When the car is running, the engine operating conditions is in frequent and rapid change, traditional experimental results can not determine the vehicle operating conditions in complex real-time environment, thus, it has many limitations and poor adaptability. Therefore, the dynamic adjustment of running vehicle in complex real-time operating conditions has great importance.This paper provides analysis of engine operating conditions, based on neural network, established the dynamic adjustment model of the immune vehicle operating conditions, through online regulation of the immune factors to achieve the dynamic adjustment of steady-state conditions and simulated this method by Matlab / Simulink software.Thesis contents are listed as followings:(1) Analyzed automobile engine performance, its influencing factors and complex features of cars running on the road. One of the most important factors is the engine performance with the operating conditions change in different air-fuel ratio and ignition advance angle and other parameters.(2) Studied the determination of vehicle conditions in complex road in nonlinear conditions,. Prior to establish the model judgments, analyzed the relationship between engine condition and its power, fuel economy and emissions performance, analyzed deficiencies of traditional vehicle conditions determinations. Tested Vela 1.5AT car in the engine bench and based on this bench test data for determining acceleration and deceleration and the trend of acceleration and deceleration conditions.(3) Studied the immune neural network, vehicle operating conditions and the domain block diagram of the dynamic regulation, trained the BP neural network and built a dynamic simulation model in Matlab / Simulink. Determined the number of network input layer, output layer neuron, the training data samples were obtained from the hidden layer neuron number. Through the simulation obtained steady conditions. As the engine speed and load conditions is the same, so, this article only considered speed impact and established dynamic adjustment model and simulation.(4) Engine bench test verified the results of stability of the dynamic simulation are correct and the engine road test verified engine performance characteristics have been improved.