Research On Dynamic Torque Sensor For Electric Power Steering System

The Electric Power Steering(EPS)system is a crucial component that directly influences the stability and reliability of a vehicle.The EPS system comprises various components,among which the torque sensor plays a significant role.The torque sensor provides real-time torque information of the steering wheel,enabling the control unit of the EPS system to provide power feedback under varying road conditions.This thesis selects the eddy current effect in electromagnetic induction as the theoretical basis of the sensor,and proposes an eddy current torque sensor to address problems such as insufficient output voltage,weak sensitivity,uneven distribution of rotor eddy current,and low spatial magnetic induction strength generated by the excitation coil.Based on the structure of the EPS steering column,the sensor employs a planar spiral coil as its excitation coil,two sets of complex three-loop toroidal coils as its receiving coils,and a nested toroidal metal sheet as its rotor.Resonant capacitance is introduced in each loop based on the principle of resonance.By utilizing the equivalent circuit model of mutual inductance coil resonant coupling in the sensor,along with the differential measurement principle,a nonlinear theoretical numerical model is derived and obtained,which describes the relationship between the output voltage and the angular displacement of the rotor.MATLAB software is employed to verify the numerical model and fit the nonlinear equation,further demonstrating the theoretical feasibility and measurement reliability of the sensor structures.To further enhance overall performance of the sensor,this study proposed optimization designs for the structural parameters of both the planar spiral excitation coil and the overall sensor structure.The optimization targets were set to be the coil quality factor and mutual inductance coupling efficiency.A multi-variable quadratic Response Surface model was then developed using both the single-factor experimental design method and the central composite experimental design method.The Grey Wolf Optimizer algorithm was employed to obtain the optimal structural parameters.Additionally,a control group experiment was conducted to verify the reliability and superiority of the optimization targets and methods.Based on the simulated output voltage characteristics of the sensor under the optimal structure,the accuracy of the theoretical analysis and the practical engineering significance of the optimization objectives were demonstrated.Furthermore,based on the finite element simulation results,a sensor signal processing circuit was designed,and the Multisim software was used to simulate,analyze and verify the signal processing circuit.Finally,detection signal acquisition and processing were implemented using supporting modules.A prototype sensor was fabricated using printed circuit boards,flexible circuit boards,and 3D printing technology.ANSYS Workbench simulation software was employed to verify the strength of the torsion bar and testing fixture,and processing and assembly were conducted.The functionality of the sensor signal processing circuit was also evaluated.The experimental results indicated that under rough assembly and positioning conditions,the sensor had a range of 0～±50N·m,a sensitivity of30.03 m V/N·m,and a linearity of 1.727%.