2016 Fiscal Year Annual Research Report
不確かなプラントのためのネットワーク化制御理論の構築：省リソース化の実現に向けて
| Project/Area Number |
16H07412
|
|---|
| Research Institution | National Institute of Informatics |
|---|
Principal Investigator |
岸田 昌子 国立情報学研究所, 情報学プリンシプル研究系, 准教授 (60782141)
|
|---|
| Project Period (FY) |
2016-08-26 – 2018-03-31
|
| Keywords | ネットワーク化制御 / 省リソース化 |
|---|
| Outline of Annual Research Achievements |
Progress has been made toward achieving the resource-aware control for the networked control systems. The objective of the resource-aware control is to minimize the use of resources (here, we consider the number of sampling and control updates) while satisfying the performance conditions.
The proposed strategy is to separately determine the sampling time and the control update time. In order to reduce the number of samplings, the idea of self-triggered control is used for ``sampling''. Namely, we determine online the next time instants of sampling using the plant model and previously received data to guarantee the satisfaction of the specified performance condition. After sampling, the idea of event-triggered control is used for ``control update''. We determine whether the control input should be updated or not by checking pre-defined conditions on the sampled data. This idea has been tailored to three linear uncertain systems for different cases.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
The obtained results are summarized and submitted to two international conferences and one journal. The conference papers have been accepted and the journal paper is under review. The details are described below.
1.Proposed an event-triggered control algorithm that uses self-triggered sampled data for continuous-time linear uncertain systems (parametric uncertainties only). The performance condition was specified by the upper bound on the derivative of Lyapunov function. The event-trigger condition was constructed so that the state error in the actuator state does not become too large. [American Control Conference 2017 (to appear)]
2.Proposed a self-triggered control algorithm for discrete-time linear uncertain systems (both parametric uncertainties and disturbances). The performance condition was specified so as to guarantee the system to be uniformly ultimately bounded. [IFAC 2017 World Congress (to appear)]
3.Proposed an event-triggered control algorithm that uses self-triggered sampled data for continuous-time linear uncertain systems (both parametric uncertainties and disturbances). The performance condition was specified so as to guarantee the finite-gain L2 stability. The event-trigger condition was constructed so as to balance the number of sampling and the number of control signal update to minimize the resource usage. [IET Control Theory & Applications (under review)]
Thus, the project has been progressing satisfactory.
|
| Strategy for Future Research Activity |
The rest of the project focuses on the following two tasks:
1.Extend the results obtained in 2 above by adding an event-trigger condition to for the control update.
2.Extend the results in the first year to nonlinear systems.
|
