The Research Of Indirect Tire Pressure Monitoring Algorithm Based On Frequency Domain Feature Analysis

With the rapid development of automobile industry, people not only enjoy great convenience from automobile, but also pay more attention to the driving safety of automobile. Especially in the case of more complex traffic environment and higher speed, the tire as the key component of vehicle driving system, the occurrence of its faults will cause negative impact on the driving safety of automobile. Therefore, the research of how to reduce the possibility of tire faults has great significance in the field of driving safety of the automobile.Tire Pressure Monitoring System(TPMS) is the third largest security system after the anti-lock braking system and airbag. As the automobile is running, TPMS can automatically monitor tire pressure in real time and give an alarm about deflated pressure, to improve the driving safety as much as possible. Presently, TPMS is usually divided into two types, one is the direct tire pressure monitoring system, and the other is the indirect tire pressure monitoring system. Generally, the former directly measures tire pressure to realize monitoring tire pressure through sensors. Comparatively, the latter analyzes the wheel speed of existing ABS or ESP system to achieve the function of tire pressure monitoring and alarming. Indirect TPMS has advantage of no need of hardware investment and convenient installation, which is beneficial for its application and even standard equipment. Hence, Indirect TPMS has broad market prospect and profound research significance.Indirect Tire Pressure Monitoring Algorithm based on the analysis of frequency character is put forward in this thesis. Through spectral estimation of wheel speed signal, the corresponding relationship between the tire pressure and the circumferential resonance frequency of the tire is established to achieve the purpose of tire pressure monitoring. Firstly, through understanding the structure and the working principle of wheel speed sensor, the source of the error of the sensor is confirmed and the model of identifying the error of the toothed ring is established. Then, errors of tooth ring are estimated based on the measured wheel speed using Recursive Least Square(RLS) algorithm and eliminated from the wheel speed signal. For event based sampling of the wheel speed signal, segmented interpolation method is used to complete the wheel speed signal reconstruction. Secondly, Fast Fourier Transform is used for spectrum estimation of wheel speed signal, and the adverse effects of the drivetrain torsional vibration and the unstable vehicle state on the spectrum estimation are analyzed. What’s more, selection criteria of frequency character are proposed. Thirdly, preparing for establishing the scheme of the calibration of resonance frequency and the strategy of deflated pressure monitoring, standardization, averageness and centroid method are used to pick up resonance frequency of tire, and then the protype of tire pressure monitoring algorithm is accomplished. Finally, the proposed algorithm based on MATLAB/Simulink simulation platform is verified by test data, the simulation results show that the resonance frequency can be estimated and verify tire pressure monitoring algorithm based on the analysis of frequency character.