During the validation of vehicles or components proving grounds are often used to generate well defined and repeatable loads within the vehicle. The picture shows the proving ground in Wörth at the Mercedes-Benz plant.

Optimal Track Mixing

Multi criteria optimization
© Photo ITWM

Since the full vehicle load is described by a large set of sensor points, a complex multi criteria optimization problem has to be solved.

Multi criteria optimization
© Photo ITWM

Groups of criteria and corresponding priorities can be defined.

During the validation of vehicles or components proving grounds are often used to generate well defined and repeatable loads within the vehicle. In addition, the proving ground is designed in such a way, that typical customer loads in the field should be reproducible in a highly condensed way. It consists of several lanes (tracks) with different surfaces, slopes, etc. which can be driven with different speeds or different steering and braking maneuvers.

Once representatives for the customer loading have been derived by statistical reasoning, a test schedule for the proving ground has to be found, which should reflect the customer load as close as possible. At the same time, it should be as short as possible to save testing time.

Multi criteria optimization

Since the full vehicle load is described by a large set of sensor points (wheel forces, strains, accelerations at various spots etc.), a complex multi criteria optimization problem has to be solved.

There is no simple overall procedure delivering an acceptable solution in any case. Instead, the different criteria (sensor points) have to be balanced against each other and in addition the total duration of the schedule has to be taken into account.

Depending on the specific task, linear and non-linear optimization methods are supplied. Groups of criteria and corresponding priorities can be defined.

Some of the criteria can be optimized, while the deviation to the target can be constrained for the others. This can be done interactively and the groups can be adapted in a flexible way. Results from a sensitivity analysis of the so called active constraints help a lot during the optimization process.