2005 Fiscal Year Final Research Report Summary
Universal Design of Four-Wheel Independent Steering System
| Project/Area Number |
16560209
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Dynamics/Control
|
|---|
| Research Institution | Graduate School of Informatics, Kyoto University |
|---|
Principal Investigator |
NISHIHARA Osamu Kyoto University, Graduate School of Informatics, Associate Professor, 情報学研究科, 助教授 (00218182)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KUMAMOTO H Kyoto University, Graduate School of Informatics, Professor, 情報学研究科, 教授 (10109019)
HIRAOKA T Kyoto University, Graduate School of Informatics, Research Associate, 情報学研究科, 助手 (30311749)
|
|---|
| Project Period (FY) |
2004 – 2005
|
| Keywords | Four-Wheel Independent Steering / Active Steering / Direct Yaw-moment Control / Minimax Optimization / Tire Workload / Front Independent Steering / Real-time Processing / Driving Simulator |
|---|
| Research Abstract |
A major study subject for the first fiscal year (2004) was a theoretical investigation on integrated control of steering and traction/braking force distribution of the four-wheel independent steering vehicle. We studied a cascade control system that consists of two levels. The upstream side is an ordinary controller for active steering vehicle. The steering angles and longitudinal tire forces are optimized at the downstream side in reference to the minimax criterion of the tire workload. We have confirmed that the problem is reduced to a convex optimization and solved efficiently with a combination of binary search and the golden section method. The numerical simulation was performed with the CarSim in the MATLAB/Simulink environment. In 2005, we concentrated on a real-time implementation of the tire force optimization, because the distributor must process a sequence of static optimization problems so that the system can synchronize with the rather simple feedback controller. It turned out that the original algorithm was not efficient enough for the Real-time Workshop that were available to us in the MATLAB/Simulink environment. Therefore, we developed an approximate algorithm that asymptotically converges to the exact solution. The new algorithm is faster enough for our CarSim-based driving simulator. Real-time simulations with experimental subjects including double lane changes were carried out and the results confirmed their contribution to the driveabilities.
|
