2004 Fiscal Year Final Research Report Summary
Fault-tolerant Coordinated Motion Control for a Modular Mobile Robot
| Project/Area Number |
14550236
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent mechanics/Mechanical systems
|
|---|
| Research Institution | DOSHISHA UNIVERSITY (2004)
Hiroshima University (2002-2003) |
|---|
Principal Investigator |
HASHIMOTO Masafumi Doshisha University, Faculty of Engineering, Professor, 工学部, 教授 (10145815)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
OBA Fuminori Hiroshima University, Graduate School of Engineering, Professor, 大学院・工学研究科, 教授 (10081267)
EGUCHI Tohru Hiroshima University, Graduate School of Engineering, Research Associate, 大学院・工学研究科, 助手 (80253566)
|
|---|
| Project Period (FY) |
2002 – 2004
|
| Keywords | Modular robot / Multi-sensor system / Dead reckoning system / Fault detection and diagnosis / Fault-tolerant sensing / Fault-tolerant control / Path tracking control / Leveling control |
|---|
| Research Abstract |
This project studied a fault-tolerant control for a modular mobile robot. The project is concluded as follows:
1. A 3D dead reckoning system for the robot was designed based on a decentralized data association. The dead reckoning algorithm is formulated based on the information filter, the Covariance Intersection and the unscented transformation. This algorithm brings highly modularity and extendibility in the dead reckoning system.
2. A motion control system was designed for the robot moving on undulating area. The motion control system consists of chassis-unit controller and wheel-unit controller; the chassis-unit controller decides the desired motion of the robot, while each of wheel-unit controllers controls its own mobile base and chassis tilting mechanism to produce the desired motion of the robot. The control system makes it possible to keep the chassis in a horizontal position while the robot tracking its prescribed path on undulating area.
3. A multi-model based fault detection and diagnosis (FDD) system of internal sensors was designed. Three fault modes (hard fault, noise fault, and scale fault modes) of each sensor are handled. The FDD algorithm is formulated based on the variable structure interacting multiple-model algorithm and the Kalman filter. The FDD algorithm is embedded into the dead-reckoning system of our modular robot. It allows the robust self-localization of the robot subject to sensor faults.
4. A fault-tolerant control system for a modular robot was designed. The hard fault of sensors and actuators in two drive/steering units is handled. The fault detection and diagnosis are achieved via the interacting multiple-model estimator and heuristic rule. A fault-tolerant controller is designed based on the Ackerman geometry. The control system achieves the safe motion of the robot subject to sensor/actuator fault.
|
