A method of obstacle avoidance is proposed aimed at the phenomenon of collision accident likely caused by the non-stationary nature of the electroencephalogram (EEG) signals while the users steer the brain-controlled wheelchair. Multi-channel ultrasonic sensors obstacle avoidance system and a feedback interface used for obstacle avoidance alert are also designed. The prompt messages displayed on the feedback interface will assist the users to make effective obstacle avoidance decisions. If the wheelchair can't avoid obstacle successfully when the users apply their EEG signals to steer the wheelchair autonomously, the system will send messages to stop the wheelchair immediately to avoid collision with obstacles, which ensures the personal safety of the wheelchair users. The experimental results of brain - controlled wheelchair show that when the safety area critical value and the emergency stop distance critical value of the front and back are respectively set as 500 millimeters, 400 millimeters and 200 millimeters, the obstacle avoidance system with good stability and strong real -time property can meet the needs of brain -controlled wheelchair users' safety navigation.%针对脑控轮椅行驶时因用户脑电信号的不稳定性可能引起碰撞事故发生的现象,提出了一种避障方法,并设计了多路超声波传感器避障系统及避障提示反馈界面；反馈界面上实时显示的提示信息将辅助用户做出有效的避障决策；当用户采用脑电自主控制轮椅运行时,若无法成功避障,则该避障系统立即启动紧急停车功能以避免与障碍物发生碰撞,保证了脑控轮椅用户的人身安全；实验结果表明:当把安全区域临界值、前后及左右紧急停车距离临界值分别设置为500mm、400mm和200mm时,该超声波传感器避障系统稳定性好、实时性强,能够满足脑控轮椅的用户安全导航的需要.
展开▼
