Multi-sensor fusion-based intelligent auxiliary system of power wheelchairs for individuals with limbs disabilities: design and implementation

With the increasing number of individuals with disabilities and an aging population, power wheelchairs have become essential. However, users with upper limb impairments struggle with conventional controls, and customized solutions remain costly. This paper proposes a low-cost, highly functional, and universally applicable intelligent assistive control system for power wheelchairs, tailored for individuals with limbs disabilities. The system integrates multi-sensor information fusion to enhance mobility and environmental awareness. An inertial measurement unit (IMU) on the user's head captures motion states, using an extended Kalman filter (EKF) to generate precise control commands. LiDAR, an odometer, and ultrasonic sensors provide environmental perception, processed via the particle filter Gmapping algorithm and Bayesian theory. Dempster-Shafer (D-S) evidence theory fuses motion intention with environmental data, enabling autonomous path planning. Prototype experiments show that the system ensures accurate wheelchair control while adapting to environmental constraints. The head motion-based directional control achieves 98% average accuracy, with the EKF reducing false triggers by 84%. Steering response time remains below 1.3 s, and obstacle detection reaches 100%. Additionally, the system dynamically optimizes paths, ensuring safe and efficient operation. This research offers a cost-effective mobility solution for individuals with limb disabilities, including those with high-level paralysis, significantly improving independence and quality of life.