A Driver Model for Virtual Drivetrain Endurance Testing

Abstract

Starting from an assumed vehicle path on a given road section we derive the formulae for the calculation of an appropriate reference speed profile tabled over road arc length. Together with a speed error feedback we thus emulate what a real driver does while driving and scheduling his actions on throttle and brake pedal. The resulting driver model may be used for automatic speed control in vehicle dynamics simulation. The application addressed here is software in the loop simulation for virtual drivetrain endurance testing at Daimler AG. A prototypical Modelica implementation was made at DLR and tested with a simple longitudinal vehicle dynamics model. Finally, we discuss the experiences with the reference speed profiles made in the industrial practice.

A Scade Suite to Modelica Interface

“…The desired vehicle speed used in the controller comes from an algorithm which computes a reference speed profile for the upcoming road section. Details can be found in [2]. The simulation results from the automatically driving vehicle using the imported Scade speed controller were well matching our expectations.…”

Experimental investigation of online path planning for electric vehicles

“…The coupling between acceleration , which is indirectly minimized, and the path curvature , which is directly addressed in the optimization problem, is given by the velocity profile generation. The velocity profile generation is based on [9] and can act as a reference for motion control. The generation of the velocity profile consists of the following main steps as depicted in Fig.…”

“…From multiple tests it turned out that the proportional gain = 1 is a good choice. The start value of the iterative approach is computed as r, ( = 0) = r, −1 * + Δ c (9) in the -th time step regarding the piecewise path planning. The term Δ c determines the driven path length of the vehicle during one time step.…”

Real-time capable path planning for energy management systems in future vehicle architectures

“…Solving the system of linear equations (3), (5) and (6) and substituting the results into (1) yields to the compact path representation with the variables and . The representation is used for our new equality constraint, which results from substituting the derived dependencies from and on and into the remaining original constraint (4). Equation (7) shows the reduced equality constraint for and…”

“…After we have gathered the spatial information of the path via optimization in the preceding sub-sections it is now necessary to generate the time information along the path. The velocity profile generation is based on [4] and consists of three main steps. First, a desired velocity profile is defined considering the maximum vehicle velocity (constraint by technical limits) and the road speed limits.…”