One-wheel electric detachable scooter design for manual wheelchair with autonomous intelligent control

• The one-wheel electric detachable scooter driving module is an inclusive solution that transforms a conventional wheelchair into a tricycle electric wheelchair. • The design of mechanical and electrical systems is proposed as a framework for constructing an electric scooter for wheelchairs with an intelligent navigation feature. • Our study focuses on integrating and evaluating a combination of adaptive pure pursuit and fuzzy logic to control the steering angle and velocity of an embedded system-based electric scooter for a wheelchair. The existing intelligent electric wheelchair is generally developed in the form of an electric motor that is integrated with a wheelchair. Thus, not only making it expensive but also heavy and bulky. To address these challenges, a design for a one-wheel electric detachable scooter (EDS) is proposed. This study designed mechanical and electrical systems for EDS. The system designed could provide manual and autonomous driving modes. The EDS was equipped with steering control, velocity control, and sensors that enable evaluation of the basic function of autonomous intelligent navigation. The evaluation for the autonomous intelligent vehicle was done by combining two control algorithms to provide trajectory tracking ability namely, the adaptive pure pursuit algorithm to calculate the optimal steering angle, and the fuzzy logic algorithm to control the vehicle velocity. Simulation and field tests were conducted to test the performance of control algorithms for a one-wheel EDS. This study result showed the EDS hardware implementation. For the proposed trajectory tracking control algorithm, the appropriate parameter configuration has been evaluated. Based on the field test and the limitations of the sensor, the proper configuration was 5 m for the look-ahead distance, 5 km/h for the maximum velocity, and 0.5 km/h for the minimum velocity. To complete the task, the vehicle got 2.06 km/h as the average velocity, which is still a reasonable one for a wheelchair. This configuration achieved 75.1 % lateral error performance, 80.8 % velocity, and 89.0 % time performance in comparison with other parameter configuration scenarios in this research.