Research on the Prediction Model of Engine Output Torque and Real-Time Estimation of the Road Rolling Resistance Coefficient in Tracked Vehicles

Road parameter identification is of great significance for the active safety control of tracked vehicles and the improvement of vehicle driving safety. In this study, a method for establishing a prediction model of the engine output torques in tracked vehicles based on vehicle driving data was proposed, and the road rolling resistance coefficient [Formula: see text] was further estimated using the model. First, the driving data from the tracked vehicle were collected and then screened by setting the driving conditions of the tracked vehicle. Then, the mapping relationship between the engine torque [Formula: see text] , the engine speed [Formula: see text] , and the accelerator pedal position [Formula: see text] was obtained by a genetic algorithm–backpropagation (GA–BP) neural network algorithm, and an engine output torque prediction model was established. Finally, based on the vehicle longitudinal dynamics model, the recursive least squares (RLS) algorithm was used to estimate the [Formula: see text]. The experimental results showed that when the driving state of the tracked vehicle satisfied the set driving conditions, the engine output torque prediction model could predict the engine output torque [Formula: see text] in real time based on the changes in the [Formula: see text] and [Formula: see text] , and then the RLS algorithm was used to estimate the road rolling resistance coefficient [Formula: see text]. The average coefficient of determination [Formula: see text] of the [Formula: see text] was 0.91, and the estimation accuracy of the [Formula: see text] was 98.421%. This method could adequately meet the requirements for engine output torque prediction and real-time estimation of the road rolling resistance coefficient during tracked vehicle driving.