scholarly journals Neural Network with Variable Type Connection Weights for Autonomous Obstacle Avoidance on a Prototype of Six-wheel Type Intelligent Wheelchair

Author(s):  
Toshihiko YASUDA ◽  
Kazushi NAKAMURA ◽  
Akihiro KAWAHARA ◽  
Katsuyuki TANAKA
Author(s):  
Toshihiko Yasuda ◽  
◽  
Hajime Tanaka ◽  
Kazushi Nakamura ◽  
Katsuyuki Tanaka ◽  
...  

We have been studying electrically powered wheelchair operation to make electrically powered wheelchair intelligent and to develop a mobility aid for those who find it difficult or impossible to use conventional electrically powered wheelchairs. Some of the prototypes we have developed use neural networks providing obstacle avoidance. In previous research, we found that by varying neural network connection weight based on obstacles in the wheelchair’s vicinity and its run state, obstacle avoidance is improved. In this research, we discuss the adjustability of neural networks with variant connection weight based on numerical studies.


Author(s):  
Mohammed Alshraideh ◽  
Basel A. Mahafzah ◽  
Saleh Al-Sharaeh ◽  
Ziad M. Hawamdeh

1996 ◽  
Vol 27 (12) ◽  
pp. 102-112 ◽  
Author(s):  
Minoru Kodaira ◽  
Teruhiko Ohtomo ◽  
Atsushi Tanaka ◽  
Masami Iwatsuki ◽  
Takao Ohuchi

Sign in / Sign up

Export Citation Format

Share Document