| Project/Area Number |
23K03913
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 21040:Control and system engineering-related
|
|---|
| Research Institution | Shibaura Institute of Technology |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2023-04-01 – 2026-03-31
|
| Project Status |
Granted (Fiscal Year 2023)
|
| Budget Amount *help |
¥2,600,000 (Direct Cost: ¥2,000,000、Indirect Cost: ¥600,000)
Fiscal Year 2025: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2024: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2023: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
|
|---|
| Keywords | data-driven control / control systems / stochastic systems / Control systems / Data-driven control / Nonlinear control / Machine learning / Stochastic systems |
|---|
| Outline of Research at the Start |
This project aims to utilize switched stochastic systems theory, data-driven control theory, and nonlinear functional analysis to develop new data-collection algorithms and data-driven control methods that provide mathematical guarantees for stabilization of nonlinear systems.
|
| Outline of Annual Research Achievements |
In the initial phase of this project, there were four main research achievements.
1) Data-driven control of unknown systems: A data-driven control design method was proposed for linear systems with unknown dynamics and input quantization. This method can be used when the system model during data collection is different from the model during control execution.
2) Handling noisy data in optimization problems: A method was proposed to solve multi-objective optimization problems despite noisy data.
3) Search-based testing approaches: A method was developed to use data from a simulator to test an automated driving system.
4) Moment propagation of stochastic systems: A framework was developed to calculate the future statistical moments of the states of a stochastic system.
|
| Current Status of Research Progress |
Current Status of Research Progress
1: Research has progressed more than it was originally planned.
Reason
This project involves using data coming from a system to learn more about it in order to develop effective controllers. Collaborations with researchers from control theory and computer science fields yielded new results in data collection approaches and new ways of using of data for learning processes and systems in optimization and testing domains.
|
| Strategy for Future Research Activity |
In the next term of this project, the following research topics will be the main focus:
1) Data-driven control of nonlinear systems with uncertainty will be addressed. A method that stabilizes periodic orbits will be developed.
2) Software will be developed for testing approaches that utilize data obtained from simulators.
3) Testing methods developed for automated driving systems will be expanded to cover other domains including aerial vehicles.
4) Game-theoretical analyses of multi-agent systems is an important topic that will be addressed.
|
