Origin of the Problem:
This problem was developed in collaboration with the University at Buffalo Mechanical Engineering Department and engineers at General Motors Corporation. In particular, we thank Joseph Donndelinger (General Motors), Scott Ferguson (University at Buffalo), Kemper Lewis (University at Buffalo), and Ashwin Gurnani (University at Buffalo) for their contributions to this project.
Objective in this Study:
1. Link the GM Technical Feasibility Model(TFM) [4] to ATSV and "shop" or visually steer through the trade space to generate Pareto solutions for a passenger vehicle design.
2. Compare the ATSV obtained Pareto solutions and the "true" Pareto solutions generated by a Multi-Objective Genetic Algorithm to determine if ATSV can be used to achieve the Pareto front more efficiently.
Description of the Problem:
Based on eleven vehicle design parameters:
1) Ten continuous variables (overall exterior dimensions and postions of the occupants)
2) One discrete variable (vehicle's powertrain)
The model includes five measures of performance:
1) Fuel Economy
2) Acceleration
3-5) Meausres of Interior Accommodation
Also includes a constraint function, ConVio, which measures thetotal violation of all constraints in the model. Enables user to explore a broader range of design points [5].
The end goal is to find the Pareto front, in the trade space using ATSV, which allows designers to "shop" for the best design. The "shopping" process is effectively done with the use of Visual Steering Commands with samplers to guide the the trade space exploration process using a vehicle configuration model [5].
Download Technical Feasibility Model (TFM)
Click here to download a copy of the GM TFM file or UPDATED ATSV file.
Research and Experimental Results:
Results from the MOGA study by University at Buffalo and GM for vehicle design can be found in: