Coordination of ground, marine and aerial robots in disaster response
碩士 === 國立臺灣科技大學 === 自動化及控制研究所 === 105 === This thesis describes a heterogeneous robotic collaboration system. It includes 1. Air quality assurance using AMR (aerial mobile robot)/UAV (unmanned aerial vehicle) with the onboard PM2.5 detectors and the intelligent turbulence avoidance strategy, 2. Inte...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2016
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/hk5s2m |
| Summary: | 碩士 === 國立臺灣科技大學 === 自動化及控制研究所 === 105 === This thesis describes a heterogeneous robotic collaboration system. It includes 1. Air quality assurance using AMR (aerial mobile robot)/UAV (unmanned aerial vehicle) with the onboard PM2.5 detectors and the intelligent turbulence avoidance strategy, 2. Intelligent water quality assurance by using the GMR (ground mobile robot) with onboard micro-spectrometer. 3. Heterogeneous robotic collaboration, Aerial-Ground, Aerial-Water (AMR autonomous landing on GMR and WMR (Water Mobile Robot)) and Aerial-Ground (localization, mapping and path planning)
This thesis successfully used artificial neural network (ANN) to classify the direction of turbulence so that the AMR can know where the turbulence from and fuzzy logic control to making decision whether to go through or avoid it depends the different situation of turbulence.
The empirical result shows the ANN method can precise detect and recognize the liquid material and concentration. And the new ideal about liquid detect is the image based method about PSNR. It can quickly classify whether the liquid is pure water or not. This can quickly help people in disaster response area to classify whether it is the drinking water or not.
The empirical result shows the global aerial SLAM for the ground robot is more effective than the conventional local SLAM conducted by single ground robot alone. The visual servo based AMR landing on GMR and WMR is more reliable and precise than the conventional GPS and barometer based localization.
|
|---|