Vehicle Mass And Road Slope Recognition For Heavy Vehicle With Hydraulic Transmission

Longitudinal dynamics model and longitudinal kinematics model of vehicles withhydraulic transmission are firstly built. Based on these two theoretical model, recursive leastsquares method (including the common recursive least squares method and the recursive leastsquares method with forgetting factor) and discrete Kalman filter are used respectively toestimate the real-time changing road angle when a vehicle is driving. These two algorithms isverified through a simulation platform established by Carsim and Matlab/Simulink together. Bycomparing the algorithms’ results under a single slope circumstance and a time-changing slopecircumstance respectively, their accuracy and real-time performance are verified.Subsequently, the effect of Kalman filter’s parameter setting has on the estimation results isanalyzed, and the parameters which can meet the requirements of both accuracy and real-timeperformance are settled. To solve the problem of realizing the algorithm in the lower computer,simplifying the Kalman filter after a deep analysis of it. The simplified Kalman filter algorithmunder different working conditions has been simulated by the established Carsim/Simulinkplatform, and compared with the un-simplified algorithm to verify the feasibility of the simplifiedalgorithm.Taking the limit of the computing ability of the hardware into consideration, the simplifiedKalman filter algorithm based on the principle of kinematics is chosen as a theoretical principleto estimate the slope. Installing an appropriate two-axel acceleration sensor in the vehicle andcombining the acceleration signals and velocity signals of the vehicle together, so theestimation for a real-time changing road angle and vehicle mass has been achieved. Thecommunication interface between the acceleration sensor and ECU of the transmission aredebugged, and a communication form which has stronger anti-interference ability is chosen toguarantee the accuracy and the reliability of the data transmission.At last, real vehicle test is carried out on several vehicles. Calibrating the static error betweenthe vehicle’s inertial navigator and the acceleration sensor before testing. By comparing the realvehicle test results of the estimation algorithm and the inertial navigator’s pitch, the methodmentioned before can realize precise real-time estimation of the road on which a vehicle disdriving.