Speed estimation of the electric-powered wheelchair by proposing the state observer method based on experimental data

Abstract

The percentage of people with disabilities (PWDs) kept increasing in each year, and the demanding usage of wheelchair is also increased. This leads to more innovation which will be focused on wheelchair such as electric-powered wheelchair (EPW) and autonomous wheelchair. The innovations of the wheelchair tend to aid the PWDs with their daily routines, and it seems more acceptable in the society. When the demand for EPW is increased, the numbers of accidents among the EPW might be increased. The objective of this study is to improve the safety system of EPW by estimating the velocity of EPW. From the velocity of EPW, the autonomous system can be embedded in the EPW in order to prevent the collision. The first phase of this study is to conduct the experiment to analyse the rotational speed of the tire during straight line and obstacle avoidance. Then, the simulation is developed to design the EPW and state observer for estimating the speed of EPW. In the modelling of EPW, the input is voltage and the output is the rotational speed of the tire. Then, the modelling is validated by comparing with the experimental results. The last phase of this study is to add the state observer in the modelling to estimate the speed of EPW. In the state observer, the experimental value from the rotational tire speed will be used as a reference and compared with the simulation. The error between the experimental and simulation value is known as estimation error, and the state observer will minimize the error to get the estimated data. By estimating the velocity of the EPW, the speed of the EPW during the cornering, obstacle avoidance and straight line could be known. The focus of this study is at first and second phases, which focus on experiments to measure the rotational speed of the tire and modelling the EPW. The results shows that the error between the experimental and simulation value is below 10%, and the modelling can be used for the state observer to estimate the speed of EPW.