电子测量与仪器学报

李沐,马超,姚杰,苏鹏,王梦迪,张海洋,徐浩文.基于运动姿态和表面肌电数据融合的多数据流神经网络步态识别[J].电子测量与仪器学报,2026,40(5):69-77

基于运动姿态和表面肌电数据融合的多数据流神经网络步态识别

Gait recognition based on a multi-stream neural network fusing motionposture and surface electromyography data

DOI：

英文关键词:gait recognition movement posture surface electromyography data fusion multi-stream neural network

基金项目:

作者	单位
李沐	1.北京信息科技大学机电工程学院北京100192；2.北京信息科技大学现代测控技术教育部重点实验室北京100192
马超	1.北京信息科技大学机电工程学院北京100192；2.北京信息科技大学现代测控技术教育部重点实验室北京100192
姚杰	北京航空航天大学生物与医学工程学院北京100191
苏鹏	1.北京信息科技大学机电工程学院北京100192；2.北京信息科技大学现代测控技术教育部重点实验室北京100192
王梦迪	1.北京信息科技大学机电工程学院北京100192；2.北京信息科技大学现代测控技术教育部重点实验室北京100192
张海洋	1.北京信息科技大学机电工程学院北京100192；2.北京信息科技大学现代测控技术教育部重点实验室北京100192
徐浩文	1.北京信息科技大学机电工程学院北京100192；2.北京信息科技大学现代测控技术教育部重点实验室北京100192

Author	Institution
Li Mu	1.School of Mechanical and Electrical Engineering, Beijing Information Science and Technology University, Beijing 100192, China; 2.Key Laboratory of Modern Measurement and Control Technology, Ministry of Education, Beijing Information Science and Technology University, Beijing 100192, China
Ma Chao	1.School of Mechanical and Electrical Engineering, Beijing Information Science and Technology University, Beijing 100192, China; 2.Key Laboratory of Modern Measurement and Control Technology, Ministry of Education, Beijing Information Science and Technology University, Beijing 100192, China
Yao Jie	School of Biological Science and Medical Engineering, Beihang University, Beijing 100191, China
Su Peng	1.School of Mechanical and Electrical Engineering, Beijing Information Science and Technology University, Beijing 100192, China; 2.Key Laboratory of Modern Measurement and Control Technology, Ministry of Education, Beijing Information Science and Technology University, Beijing 100192, China
Wang Mengdi	1.School of Mechanical and Electrical Engineering, Beijing Information Science and Technology University, Beijing 100192, China; 2.Key Laboratory of Modern Measurement and Control Technology, Ministry of Education, Beijing Information Science and Technology University, Beijing 100192, China
Zhang Haiyang	1.School of Mechanical and Electrical Engineering, Beijing Information Science and Technology University, Beijing 100192, China; 2.Key Laboratory of Modern Measurement and Control Technology, Ministry of Education, Beijing Information Science and Technology University, Beijing 100192, China
Xu Haowen	1.School of Mechanical and Electrical Engineering, Beijing Information Science and Technology University, Beijing 100192, China; 2.Key Laboratory of Modern Measurement and Control Technology, Ministry of Education, Beijing Information Science and Technology University, Beijing 100192, China

摘要点击次数: 224

全文下载次数: 82

中文摘要:

针对步态识别方法中单模态信号在前进、后退与转弯等动态任务中易误判的问题，提出一种基于运动姿态信号与表面肌电(sEMG)信号融合的多数据流神经网络模型方法。采集下肢运动姿态与表面肌电信号，通过步态周期划分和对齐策略完成数据处理，从数据特性信息在时序结构、信号类型与表达维度上的异质性角度出发构建包括运动姿态的动态时序特征、sEMG的频域倍频程谱和时域直方图统计的特征信息。采用Transformer捕捉运动姿态信号的动态演化模式、双路多层感知机（MLP）分别提取sEMG在时域与频域上的局部响应特性，构建进行功能互补的多数据流神经网络模型，从而实现运动姿态与局部肌纤维活动信息的融合表达。结果表明，所建立的模型在前进、后退与转弯3类任务中的识别准确率分别提升了8%、6%与24%，显著改善了由动作边界模糊带来的误判问题，在准确性、泛化性方面具有更好的表现。该方法为多模态融合在复杂步态识别中的应用提供了研究参考和技术支撑，为高精度的步态识别系统设计提供了参考。

英文摘要:

To address the problem of single-modal signals being prone to misclassification in dynamic tasks such as forward, backward, and turning, this paper proposes a multi-stream neural network model based on the fusion of motion posture signals and surface electromyography (sEMG) signals. Lower-limb motion posture and sEMG signals are collected and processed using a gait cycle segmentation and alignment strategy. Based on the heterogeneity of data characteristics in terms of temporal structure, signal type, and expression dimensions, feature information is constructed, including dynamic temporal features of motion posture, frequency-domain octave spectrum of sEMG, and time-domain histogram statistics. A Transformer is used to capture the dynamic evolution of motion posture signals, and a two-way multilayer perceptron (MLP) is used to extract local response characteristics of sEMG in the time and frequency domains, forming a multi-stream neural network structure with complementary feature representations to achieve the fused expression of motion posture and local muscle fiber activation information. Results show that the proposed model improves recognition accuracy by 8%, 6%, and 24% in forward, backward, and turning tasks, respectively, significantly reducing misclassification caused by blurred motion boundaries and demonstrating superior accuracy and generalization performance. This method provides research reference and technical support for the application of multimodal fusion in complex gait recognition, and provides a reference for the design of high-precision gait recognition systems.

查看全文查看/发表评论下载PDF阅读器