Obstacle crossing intention recognition method, and an exoskeleton robot obstacle crossing control method and equipment

Abstract

The invention discloses an obstacle crossing intention recognition method, and an exoskeleton robot obstacle crossing control method and equipment. The obstacle crossing intention recognition method comprises the steps: collecting gait data and electroencephalogram data of an observed person; acquiring an electroencephalogram data window in the electroencephalogram data corresponding to the gait data according to the gait data; and training an obstacle crossing model for identifying an obstacle crossing intention based on the electroencephalogram data window. According to the obstacle crossingintention recognition method, the exoskeleton robot with a simple structure can be controlled to walk and cross obstacles with the good man-machine collaboration; the training model algorithm is nothigh in hardware requirement, information redundancy is not prone to being caused, system delay is not prone to being generated, so that the real-time performance of the exoskeleton robot is improvedand the application difficulty is reduced; the user can control the exoskeleton robot to execute the obstacle crossing action only with the obstacle crossing intention without extra external stimulation for guidance.

Description

technical field [0001] The present application relates to the field of robot control, in particular to an obstacle-crossing intention recognition method, an exoskeleton robot obstacle-crossing control method and equipment. Background technique [0002] With the deepening of population growth and aging in our country, the number of people with lower limb disabilities will continue to grow. Relying on technological innovation to ensure and improve the quality of life of people with lower limb disabilities has risen to national strategic needs. With the development of robotics technology, lower limb exoskeleton robots can effectively help people with lower limb disabilities to perform rehabilitation exercises and even restore walking ability. Exoskeleton robots based on brain-computer interface control can perform movements that are more in line with the natural rhythm of human movement, thereby enhancing the degree of human-machine coordination; on the other hand, it can enhan...

AI Technical Summary

Problems solved by technology

[0003] Existing lower extremity exoskeleton robots based on traditional control methods have poor obstacle-surmounting ability, and basically can only walk on flat ground; even if they have the obstacle-surmounting function, the mechanical structure is too complex, which makes the human-machine collaboration performance poor

However, the current lower extremity exoskeleton robot system based on the brain-computer interface mainly uses steady-state visual evoked potentials, motor imagery and other paradigms to identify the user's movement intention. Make the user's visual fatigue or even damage; while the paradigm based on motor imagery requires the user to carry out training with a high concentration of attention, and is less applied to the intention recognition of crossing obstacles

There are also some systems that try to combine multiple signals such as vision and voice with EEG signals to identify intentions, but they have high requirements for algorithms and hardware, which may easily cause information redundancy and increase system delay, which will affect the real-time performance of the exoskeleton system. Increase the difficulty of practical application

Images