Experimental Verification of Discretely Variable Compression Braking Control for Heavy Duty Vehicles: Final Report

In the first two chapters of this report, the development of discrete compression brake and transmission models is explained. In the vehicle model development, special efforts have been put in transmission shifting scheduling. Transmission up-shift and down-shift scheduling are separated in the modelling. Hysteresis during shifting is introduced to reduce chattering. A compression brake effect on transmission shifting is identified and considered in the modelling. The new transmission shifting model has been validated through experimental data. The transmission shifting model is combined with the compression brake model and the model of the longitudinal vehicle dynamics for a high-fidelity predictive simulation software tool. Power-width-modulation (PWM) actuation for brake coordination is proposed to further exploit the capacity of the compression brake and reduce the usage of service brake. Simulation results indicates that the PWM actuation strategy will have the same speed regulation performance as the direct torque split strategy, while the usage of the service brake is significantly reduced, and the compression brake can handle a down slope of -4 when tracking a speed of 56 mile per hour. Simulation results also indicated the importance of vehicle mass and road grade estimation in controller performance. The estimation problem is addressed in detail next. In the third chapter, a recursive least square scheme with multiple forgetting factors is proposed for on-line estimation of road grade and vehicle mass. The estimated mass and grade can be used to robustify many automatic controllers in conventional or automated heavy-duty vehicles. We demonstrate with measured test data from the July 26-27, 2002 test dates in San Diego, CA, that the proposed scheme estimates mass within 5% of its actual value and tracks grade with good accuracy. The experimental setup, signals, their source and their accuracy are discussed. Issues like lack of persistent excitations in certain parts of the run or difficulties of parameter tracking during gear shift are explained and suggestions to bypass these problems are made. In the final part of this work, an adaptive model predictive control scheme is designed for speed control of heavy vehicles. The controller coordinates use of compression brakes and service brakes on downhill slopes. The advantage of model predictive control (MPC) scheme over PI design is its ability to explicitly handle the constraints on service brake and compression command. MPC minimizes use of service brakes based on constrain optimization techniques. Given a perfect model, good performance was obtained in closedloop. We showed with simulation analysis that model uncertainties due to uncertainty in vehicle mass result in degraded and sometimes poor oscillatory closed-loop response. Unknown road disturbance (grade) was another limiting factor to the performance of the controller. It is shown that accurate estimate of mass is necessary for safe and comfortable operation in closed-loop. Also knowledge of the road grade can improve the results further by contributing in feedforward control. Therefore the estimation method presented in the third chapter of the report is used in parallel with the controller to update the estimates of mass and road grade. We show that adaptation allows robust speed tracking even if initial mass is 300 percent over or underestimated.