Autonomous smart wheelchair: Dual-axis electromagnetic harvesting and IoT-based applications with deep learning AI

Advancements in science and technology have improved wheelchairs, including self-sensing and control, but they still need improvements in energy enhancement. To meet the requirement of continuous power supply and smooth operation, this research proposes a dual-axis electromagnetic energy harvesting system capturing the kinetic energy of wheelchairs moving on the paths. The proposed system consists of three modules: a dual-axis wheel energy-capturing module, an electromagnetic induction module, and a power storage module. The dual-axis wheel energy capture module grabs front wheel movement along two axes, while the electromagnetic induction module converts the wheelchair's motion into electrical energy through the rotation of the magnetic ring against the coil ring. The energy storage module stores the power generated by rotating and turning electromagnetic energy harvesters. Analytical modeling, simulation, and experiments were conducted to investigate the system's performance. The system attained a peak voltage of 20.4 V with an RMS power of 0.238 W for Part A and 1.53 V and an RMS power of 0.63 mW for Part B. Additionally, a deep learning method has been used to assess the speed and turning angle voltage signal data from a generator, achieving training accuracy of 99.35 % and 99.37 %. The proposed system can help disabled victims navigate with continuous power supply without battery depletion and to monitor their health conditions with self-powered sensors.