Model Predictive Obstacle Avoidance Control for an Electric Wheelchair in Indoor Environments Using Artificial Potential Field Method

In this paper, a model predictive control (MPC) based obstacle avoidance control scheme in the indoor environment is proposed. Since the mobile area is limited, indoor vehicles need to take a path with minimum avoidance while estimating states of obstacles and vehicles sequentially. With the combination of the artificial potential field method, an MPC-based control scheme is introduced to deal with the above problem. To estimate the self-position and obstacle position, a 3D point cloud is clustered, and the Normal Distribution Transform-Simultaneous Localization And Mapping (NDT-SLAM) and the Kalman filter are employed. A Gazebo simulation is conducted to evaluate and compute path and velocity considering obstacles in an environment with multiple unknown stationary obstacles. As a result, the avoidance of many obstacles occluded by other objects can be realized. In addition, an experiment in an actual indoor environment is conducted to verify the motion of the experimental vehicle. It is confirmed that it is possible to drive with MPC using the estimated position even in a narrow environment.