Structure of the PowerTrain 2.0 library
In the sense of object-oriented modelling (and as consequence of our various model-ling activities), it is essential to have libraries available which comprise components and assembled models. The main fields of our automotive modelling activities are vehicle dynamics, power trains (both conventional and alternative) and vehicle controls.
One important aspect in development of different libraries for automotive applica-tions is interoperability. As a common effort of multiple (not only DLR) developers of Modelica automotive libraries, a general architecture for the modelling of vehicles was created with the scope to enable inter-operability of developed libraries. The result-ing Modelica library called VehicleInterfaces was released by partners including DLR.
The institute’s expertise in the field of powertrain modelling started in 2000. The knowledge and experience was gathered steadily in a commercial Modelica library called PowerTrain. It is sold to automotive manufacturers and suppliers worldwide. Further development of the library led to its second version which was completed in 2007. This version was significantly enhanced and completely redesigned to be able to interoperate with other automotive libraries on the base of the VehicleInterfaces library.
The PowerTrain library provides sophisticated models of manual transmissions, drivelines, drivers and other components. The range of examples was extended significantly including tests for both particular component models as well as complete vehicle architectures (e.g. for power consumption calculation and performance tests or analysis of shift strategies)
Model of a simple vehicle dynamics control (based on single track model)
Simulation-based evaluation and model-based controller design for future mechatronic chassis require multidisciplinary modelling of both involved vehicle dynamics and its control systems.
The VehicleControls library (currently at the beginning stage of development) will provide basic models of common vehicle dynamics control and driver assistant systems, such as anti-lock braking system (ABS), anti-slip regulation (ASR), electronic stability control (ESP), etc. The scope is to model the function principles rather than to mirror detailed (proprietary) systems.
Future integrated control strategies will as well be provided that are based on non-linear inverse vehicle models. Consequently, the necessary vehicle models will be provided as well. The vehicle model complexities vary from simplest single track model through double track up to sophisticated multibody models. First tests of our new controller concepts have been carried out for yaw rate control based on a special “disturbance observer” and for rollover avoidance via coordinated steering/braking. The library VehicleControls is based upon the VehicleInterfaces library and isdeveloped within the scope of project EuroSysLib.
Diagram view of a Modelica model of a parallel hybrid vehicle with optimised operating strategy
Hybrid electric vehicles (HEV) offer the opportunity of fuel reduction by torque allocation of the electric motor and the engine and by taking advantage of temporary energy storage in a battery. The ability of fuel reduction is based on the varying efficiencies of the propulsion aggregates. These efficiencies depend on the operating states, recuperation and boost which means, that the engine can be sized down when it is supported by the motor to meet the peaks of power demand. In this context, the library also contains examples of fuel consumption optimised operating strategies.