Researches On Characteristics Of Two Kinds Of Sensors For Automobile Engine

Mass air flow sensor and universal exhaust gas oxygen sensor are two kinds of important sensors for electron control units of automobile engine. They provide feed-forward control and feed-back control signals for the close-loop air-fuel-ratio control system, respectively. It is necessary to research characteristics of these two kinds of sensors in-depth, in order to ensure the ECU to obtain the accurate information quickly within the whole speed variation range.Static and dynamic experiments of MAF sensor are performed, and dynamic nonlinear modeling methods based on block oriented models are studied for describing the characteristics of MAF sensor. The model of static non-linear part in the block oriented model is built based on the static experimental data. The input signal of dynamic experiment is constructed according to both the dynamic experimental data and actual situation of experimental equipment. The model of the dynamic linear part in the block oriented model is established by adopting particle swarm optimization algorithm combining with function link artificial neural network and system identification method, respectively.Two kinds of dynamic nonlinear correction methods based on the block oriented model are studied aiming at the sensors based on oriented model with the nonlinear static characteristic of saturation in first quadrant. The results show that different strategies should be utilized when damp ratios of sensors are different. In order to improve dynamic performance of MAF sensor, a correction system based on Wiener model has been designed for MAF sensor of Hammerstein model, and then another one based on Hammerstein model has been designed for sensor of Wiener model. The simulation results indicate that two kinds of correction systems can improve dynamic performance of MAF sensor and reflect the actual amplitude of measured signal. Two kinds of correction systems are implemented by means of a dSPACE real time simulation system. Online dynamic nonlinear correction experiments are conducted, and the experimental results are studied.The temperature of UEGO sensor is obtained indirectly by measuring the cell internal resistant under alternating current. The subsection heat control method is employed to make UEGO sensor operate at an ideal temperature according to the difference between the voltage of the reference resistance and the voltage of Nernst cell resistance. A robust PID algorithm is used to control the value and direction of the pumping cell voltage and exert a feedback control action on the UEGO sensor. This means that the pumping cell voltage has the role of draining oxygen ions from the diffusion chamber to the external environment (or vice versa), and makes the Nernst cell voltage maintain at a specified value. The conversion results of the pumping cell current can provide the information about the oxygen content or air-to-fuel ratio of the engine exhaust gases. UEGO controller is implemented by means of dSPACE real time simulation system. The experiments of UEGO controller are conducted on a carburetor engine bench. The experimental results show that the UEGO controller has good robustness and running performances. Its accuracy is high whenλvalue is in steady statues, and its response speed is fast whenλvalue changes.