Novel Fuzzy Logic Controller based on Time Delay Inputs for a Conventional Electric Wheelchair

 

M. Rojas, P. Ponce, A. Molina

 

 

This work proposes a Dynamic fuzzy logic Controller for the navigation problem of an electric wheelchair. The controller uses present data from three ultrasonic sensors as the main source of information from the environment. However other inputs, named as “dynamic time delay”, are obtained from past samples of those static data and are used to design the rule base. Although fuzzy logic controllers with static inputs could solve basic navigation problems, the proposed structure with dynamic inputs gets an excellent performance for more complex navigation problems. There were designed static and dynamic navigation strategies, which were first deployed in software just to evaluate their behavior. They were tested in a maze and their trajectories were compared to select the best. For improving its response, the dynamic fuzzy logic strategy was deployed in hardware. The paper presents a comparison between the software and hardware applications to illustrate the possibility of implementing the proposed methodology in different platforms. The dynamic fuzzy logic controller led the electric wheelchair without colliding against walls, and is a high performance navigation system. Moreover, this controller could solve the sensor limitations.