Operation state observation and condition recognition for the control of power-assisted wheelchair

A wheelchair is a mandatory device for people with mobility problems. Typical wheelchairs are maneuvered by muscular forces transmitted to wheels. Such wheelchairs, however, are physically demanding in various conditions, e.g., uphill or step. For this purpose, electric wheelchairs have been developed and commercialized for enhancement of the maneuverability and safety of people with muscular weakness. Such systems involve complete human–machine interaction, and thus it is necessary to accurately observe the physical state and the environmental condition in real-time. For this reasoning, a sensor-fusion method and a decision making algorithm for the complete observation of the physical states of the wheelchair and the detection of the operation conditions are introduced in this paper. The physical states to be observed include the wheel speed, the pitch angle and the external forces, which are the most important physical quantities for the control of a power-assisted wheelchair (PAW) system in daily-life conditions. For the acquisition of such information, multiple motion sensors (i.e., encoders, gyroscopes and accelerometers) are utilized and intelligently fused. Then the estimated physical quantities are further processed in real-time in order to detect the environmental conditions. The proposed methods are all verified by experiments in this paper.