| Project/Area Number |
16560215
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Dynamics/Control
|
|---|
| Research Institution | Nihon University |
|---|
Principal Investigator |
HORIUCHI Shinichiro Nihon University, College of Science and Technology, Professor, 理工学部, 教授 (30181522)
|
|---|
| Project Period (FY) |
2004 – 2005
|
| Project Status |
Completed (Fiscal Year 2005)
|
| Budget Amount *help |
¥1,800,000 (Direct Cost: ¥1,800,000)
Fiscal Year 2005: ¥600,000 (Direct Cost: ¥600,000)
Fiscal Year 2004: ¥1,200,000 (Direct Cost: ¥1,200,000)
|
|---|
| Keywords | Nonlinear Control / Stochastic Robustness / Optimization / Genetic Algorithm / Monte Carlo method / Vehicle Dynamics Control / 非線形統合制御 / 確率的評価基準 / 最悪入力 / 同時設計 / 非線形予測制御 / 入出力安定性 |
|---|
| Research Abstract |
For the design of the nonlinear control systems, robustness is necessary to determine the possibility of instability or inadequate performance in the face of uncertain variation of system dynamics. From this point of view, several design method for nonlinear control system were proposed based on robust control theories such as H_∞ control and μ analysis/synthesis. These approaches assume that the dynamics of controlled system can be described by the linear-time-invariant model and using a singular-value as a metric, the robustness to system dynamics variations is treated as a deterministic way. However it is difficult to express real uncertainty in deterministic form that is required by the theories, and the relationship between the metric of robustness and parameter variations in the physical system is weak.
This study investigates the feasibility of applying the stochastic method to the evaluation and design of nonlinear control systems. The controlled system dynamics is modeled by no
… More
nlinear dynamic equations containing uncertain system parameters. The system robustness is characterized by the probability of instability and probabilities of violation of prescribed performance indexes, subject to the variation of uncertain system parameters. These probabilities are estimated by applying Monte Carlo method to the results of a large number of simulations. To reduce the computational time, a parallel processing system composed of a master computer and nine slave computers is constructed. A performance index consists of weighted sum of the estimated probabilities is optimized by Genetic Algorithm.
The proposed method is applied to the design of vehicle dynamics control system. Four types of vehicle control strategies are chosen to demonstrate the effectiveness of stochastic method to analyze the robustness of the control systems. Then the stochastic method is applied to the design of the nonlinear vehicle control systems. A stochastic cost function containing engineering design criterion is minimized by Genetic Algorithm, producing optimal controller parameters for given control structures. The stochastic approach improves the robustness of the vehicle control systems Less
|
