通过人体生物电信号解码运动意图，让机器自然理解人的运动意图，发展基于神经信号的新一代人机接口，已成为研究热点。针对现有方法从生物电信号的统计特性难以反映人体运动的复杂本质，解码运动意图遇到了瓶颈。本研究从人体运动意图相关的大脑皮层和神经肌肉机理出发，致力于研究和构建一种新的计算模型与方法，创新性地提出脑肌功能网络模型，在人体运动机理和脑肌模型的高度解码运动意图，有望取得开创性突破。项目基于复杂网络理论研究多变量脑肌电信号的因果关系测度和相关性建模，定量分析脑肌神经信号间耦合的非线性、高阶等复杂特性，在不同时空层次上耦合分析中枢神经与肌肉、神经肌肉-神经肌肉之间的运动功能联系，构建脑肌功能网络模型，提取并融合脑肌功能网络特征，应用机器学习算法预测和感知人体运动意图，从本质上建立更加有效的人体运动意图感知方法，突破现有方法的瓶颈，有望解决外骨骼机器人、康复机器人、智能假肢的人机自然交互核心问题

Decoding the movement intent using the human bioelectric signals to develop a new human-machine interface is a hot research topic. The fact that the statistical characteristics of bioelectrical signals cannot easily reflect the complex nature of the human motion, decoding the movement intent has encountered some challenges. This research work investigates a new computational method based on the cerebral cortex and neuromuscular mechanism of the movement. In the current study, we proposed a novel brain muscular functional network model. It is expected to achieve the accurate results through decoding the movement intent from the movement mechanism and the brain muscle model. Then, we studied the causality and correlation between multichannel EEG and EMG based on complex network theory. We analyzed the quantitative nonlinear high order complex correlation between multichannel EEG and EMG to construct the brain muscular functional network model. In the next step, we extracted features from the brain muscular functional network, and employed machine learning algorithms to predict the movement intent. Using our proposed method, it is expected to solve the core problem of human-machine interface about exoskeleton robots, rehabilitation robots, and intelligent prosthesis.