DECISION MAKING UNDER UNCERTAINTY USING A VEHICLE MULTILEVEL MODEL: APPROACH TO TARGETS ALLOCATION

In this paper, we consider setting performance targets for a vehicle design. The vehicle is modeled by a multilevel hierarchical tree structure. We have considered that each leaf of the structure may have several characteristics, and that for each characteristic a target is defined. Experts’ opinions are expressed with uncertainty regarding the feasibility of achieving these targets. Experts’ opinions are given in the form of intervals associated with their subjective beliefs for the possible values of characteristics. The collected information is propagated in the model to determine the plausibility and the belief for characteristics at the vehicle level. Using this information, five target allocation approaches are discussed which can be applied to three vehicle design strategies.

Download Full-text

Target Cascading: A Design Process for Achieving Vehicle Ride and Handling Targets

10.1115/imece2001/dsc-24533 ◽

2001 ◽

Author(s):

D. Geoff Rideout ◽

Jeffrey L. Stein ◽

John B. Ferris

Keyword(s):

Design Process ◽

Design Methodology ◽

Vehicle Design ◽

Design Strategies ◽

Tire Pressure ◽

Simulation Tools ◽

Performance Targets ◽

Overall Design ◽

Target Cascading ◽

Ride And Handling

Abstract Vehicle dynamics are well understood by both academic researchers and automotive industries. And while modeling and simulation tools are still underutilized, they are becoming more frequently used in the vehicle design process. However, there is still lacking an overall design methodology that can link and integrate in a systematic fashion the design tasks of individual components or systems such that the vehicle performs as intended with a minimal number of design iterations. A process called Target Cascading, applied in the early stages of vehicle design, might serve as this systematic design methodology. In this paper, Target Cascading is evaluated for its ability to propagate top-level design specifications down to specifications for various subsystems and components in a vehicle design problem. More specifically, general ride and handling targets are set for a vehicle and these are cascaded down through the suspension, tire pressure and spring design levels by partitioning the original problem into a hierarchical set of subproblems. At a given level, an optimization problem is formulated to minimize deviations from the proposed targets and thus achieve intersystem compatibility. A coordination strategy links all subproblem decisions so that the overall supersystem performance targets are met. Results are presented that demonstrate Target Cascading’s utility in unearthing tradeoffs and incompatibilities among initial targets early in the vehicle development cycle. Throughout the paper, the Target Cascading process is compared to traditional vehicle design strategies for achieving ride and handling targets. Target Cascading appears to be a promising systematic technique for the design of vehicles to meet ride and handling specifications.

Download Full-text