基于足底压力分布时空HOG特征的步态识别方法

提出一种基于足底压力分布时空HOG的步态识别算法,在特征层对足底压力的时间域和空间域信息进行融合。首先寻找足底总压力时间曲线上的极大值和极小值等几个特征点,利用这几个特征点所对应时刻的足底压力分布来构建时空HOG特征向量,最后采用SVM进行步态识别。采集不同行走速度下30人的单步足底压力分布数据进行实验,在不区分样本速度的情况下,该方法的识别率为93。5%。实验结果表明足底压力分布时空HOG特征能较好地刻画步态动力学特征,且具有良好的速度适应性。     

基于足底压力与支持向量机的步态相位识别研究

随着两足机器人、人工假肢技术以及为行走困难病人康复设计的康复训练机器人的发展,在线的步态相位识别方法越来越重要。本文提出的基于足底压力与支持向量机(SVM)的步态相位识别算法主要由五部分组成,即数据采集、数据预处理、特征提取、训练分类器和分类识别。实验表明：该方法能够对运动中的步态相位进行准确的判断。     

电动跑步机上步态特征获取系统的设计

将柔性阵列压力传感技术应用到跑步机上,设计了一种新的在跑步机上获取步态特征的系统。该系统可以准确获取步长、步频、腾空时间和支撑时间等4个步态特征指标,用于跑步机上跑步的步态特征分析。通过研究不同速度下训练者的步态特征,发现训练者会通过调整步频和步长以达到设定的跑步机跑速;同时,训练者在不同速度下调整的方式是不同的。因此,该研究为今后在电动跑步机上实现速度自适应控制提供了很好的参考。     

Gender and walking speed effects on plantar pressure distribution for adults aged 20-60 years

This study investigates gender and walking speed (80%, 100%, 120% and 140% of preferred walking speed (PWS)) effects on plantar pressure parameters. In total, 30 healthy males and females, aged between 20 to 60 years, participated in this study. A plantar pressure measurement device was used to measure the peak pressure, peak force and contact area in six plantar zones. The results indicate that males had higher peak pressure and peak force in the medial toe and forefoot, as well as greater contact area in the central forefoot and heel areas. Females had greater contact area in the midfoot. Increased walking speed caused a significant increase in most of the response measures and the increase became more obvious when the speed was higher than 120% PWS. Although there was no significant interaction between gender and PWS, some gender differences were found. PRACTITIONER SUMMARY: Using percentage PWS provides a new perspective to discuss the effects of gender and walking speed on plantar pressure distribution. This study's findings can be very useful for footwear and orthotics design for different genders.     

Effects of different loads and carrying systems on selected biomechanical parameters describing walking gait

The effects of two different systems on selected biomechanical parameters of walking gait, while carrying loads of varying magnitude, were investigated.

Ten healthy males who were not regularly engaged in carrying tasks walked a distance of 20 m for ten trials for each of the following five conditions: (i) normal walking without any external load; (ii) 20% and (iii) 40% body weight carried using a backpack system; and (iv) 20% and (v) 40% body weight carried using a doublepack system which distributed the load equally between the front and back of the subjects. The experimental set-up consisted of a Kistler force platform interfaced to a Tektronix 4051 Graphic Calculator, two super 8 mm movie cameras and a photoelectric timing system. Force data (417 Hz) were obtained for ten trials along with side- and rear-view film data (100 fps) for three of the trials for each of the subject conditions. In addition, selected aspects of foot-position data were acquired from a minimum of six footprints from one trial for each subject condition. Walking speed was controlled at 4·5 ± 0·3km/h. Parameters describing the temporal relationship of the gait pattern and values describing the spatial relationship of foot position were evaluated. Selected variables describing the components of the ground-reaction-force-time curves were also examined. Finally, selected kinematic and kinetic parameters were evaluated for four functional subphases of the support period.

Comparisons using a one-way ANOVA with repeated measures were conducted to examine differences between parameters describing the load-carrying conditions and normal gait. Results from the analysis revealed that both the light and heavy loads substantially modified the normal walking gait pattern. Interactions between the load conditions and carrying systems were tested using separate two-way ANOVA with repeated measures. Significant ordinal interactions as well as significant main effects were found between the two carrying systems for some parameters, suggesting that the doublepack system was more effective than the conventional backpack system, especially for carrying the heavy load.     

基于双流步态网络的跨视角步态识别

通过增强样本数据和网络特征,提出双流步态网络,增强模型对携带物、衣物变化影响的鲁棒性.首先构造双流步态网络,分别提取步态视频数据中的全局特征和协变量影响范围外的局部判别信息.再将两组网络的特征信息相加融合后,得到步态的双流特征表达.提出的限制随机遮挡策略增广用于训练样本的难度和多样性,提高网络对局部特征的学习能力,减弱协变量的不利影响.另外,改进三元组损失采样方法,加速网络模型的训练收敛速度.在大型步态数据集CASIA-B和OU-MVLP上的实验表明,在携带背包和穿着不同衣物的行走状态下,双流步态网络步态识别准确率较高.     

基于足底压力和关节运动信息的步态检测系统设计

作为老年人高发疾病,脑卒中有着较高的几率造成患者行动障碍。及时检测、获取偏瘫患者行走时步态信息,是治疗师为患者制定康复计划的重要环节。目前,较新的步态检测系统大多采用三维分析,三维步态分析系统体积较大、成本高、操作不便,在患者主动康复训练中存在较大的局限。针对现今步态检测系统的局限,设计了一种结合无线传输技术,检测患者关节运动角度及足底压力的运动信息检测系统。系统采用薄膜压力传感器、惯性传感器分别采集足底压力及关节运动信息,实现了对患者的8路足底压力信息及下肢关节运动角度信息的快速、有效采集。实验表明,系统可以有效采集关节运动角度及足底压力信息,在未来的康复训练中有着很好的应用前景。     

结合无符号Laplace谱特征的触觉步态识别算法

针对单纯利用压力点分布特征进行触觉步态识别的不足,提出了一种结合无符号Laplace谱特征的动态触觉步态识别算法。利用足底压力数字化场地采集常速、快速和慢速三种情况下的触觉步态数据,生成足底压力分布图像,并根据足底解剖学的结构划分区域;以足底压力图像各区域为节点构造结构图,并采用无符号Laplace矩阵表示;通过对该矩阵进行奇异值分解（Singular Value Decomposition,SVD）获取谱特征,并结合形状特征得到触觉步态特征;选择“一对一”的支持向量机（Support Vector Machine,SVM）多分类方法,按照人在行走过程中不同的速度分别构造分类器,从而实现动态触觉步态的识别。实验结果表明该识别算法对不同速度样本数据的触觉步态识别正确率都较高。     

基于步态检测算法的辅助行走系统设计

为了保证盲人和视障患者安全方便出行和减少误报警,设计一种基于步态检测算法的辅助行走系统。信号处理算法上利用小波包对信号进行降噪,提取三轴加速度的均值、标准差、方差与小波能量谱,并结合足底压力信息组成多维参数作为步态特征,选取基于遗传算法优化支持向量机进行步态识别。硬件部分主要包括三轴加速度信号和足底压力信号采集硬件设计。软件部分主要包括对足底压力信号和三轴加速度信号的采集和处理以及一款能够自动拨号安卓手机APP。实验结果表明,该系统基于多传感器融合的步态检测技术,步态检测识别率平均达90.48%。该系统便携性好、功耗低、测障效果好,在辅助行走领域具有一定的研究意义和实用价值。     

Intent Pattern Recognition of Lower-limb Motion Based on Mechanical Sensors

Based on the regularity nature of lower-limb motion, an intent pattern recognition approach for above-knee prosthesis is proposed in this paper. To remedy the defects of recognizer based on electromyogram (EMG), we develop a pure mechanical sensor architecture for intent pattern recognition of lower-limb motion. The sensor system is composed of an accelerometer, a gyroscope mounted on the prosthetic socket, and two pressure sensors mounted under the sole. To compensate the delay in the control of prosthesis, the signals in the stance phase are used to predict the terrain and speed in the swing phase. Specifically, the intent pattern recognizer utilizes intraclass correlation coefficient (ICC) according to the Cartesian product of walking speed and terrain. Moreover, the sensor data are fused via DempsterShafer's theory. And hidden Markov model (HMM) is used to recognize the realtime motion state with the reference of the prior step. The proposed method can infer the prosthesis user's intent of walking on different terrain, which includes level ground, stair ascent, stair descent, up and down ramp. The experiments demonstrate that the intent pattern recognizer is capable of identifying five typical terrain-modes with the rate of 95.8%. The outcome of this investigation is expected to substantially improve the control performance of powered above-knee prosthesis.      

下肢助行外骨骼步态规划与仿真研究

针对目前大多数下肢助行外骨骼自身不能维持稳定行走与步态规划方法存在缺陷的现状，设计一种10自由度主动驱动下肢外骨骼；其步态规划基于静稳定性判据，采用重心投影法，利用构造函数规划重心及踝关节轨迹，进一步通过运动学计算获得完整步态周期内所有关节的运动轨迹；运用ADAMS构建人-外骨骼耦合运动虚拟样机模型，进行平地步行仿真，仿真得到关键点的空间轨迹与动力学参数，通过与规划的重心和踝关节轨迹对比，外骨骼能够实现预期运动并稳定行走，仿真结果表明所设计的步态规划方法合理有效。     

Prospective gait changes as a function of shifting perceptions of slipperiness: effects of visual and somatosensory cues

Forty participants, ages 18–45 years, rated perceived slipperiness before and after walking on five different floors under three different surface conditions. The before-ratings were taken as a proxy for visual cues to slipperiness, while after-ratings were taken as a proxy for somatosensory feedback received while walking on the surface. Before and after ratings of slipperiness were used to predict gait parameters, as a function of trial, during repeated walking. Effects of after-ratings of slipperiness were observed beginning on the second trial, and continued through the fifth trial, while effects of before-ratings of slipperiness were most apparent on the first trial. When perceived slipperiness increased (or decreased) from before to after walking on the surface, gait became more (or less) protective across trials. It is concluded that both visual cues, as well as somatosensory feedback, are used in the prospective control of gait.

Practitioner Summary: Effects of visual and somatosensory cues to slipperiness on gait were disentangled using floor surfaces varying in the slipperiness suggested by those cues. Visually based ratings of slipperiness predicted gait parameters on earlier trials, while somatosensory-based ratings predicted gait parameters on subsequent trials. Flooring design should provide reliable information regarding slipperiness.     

Wave CPG model for autonomous decentralized multi-legged robot: Gait generation and walking speed control

In this paper, we propose a method to control gait generation and walking speed control for an autonomous decentralized multi-legged robot by using a wave Central Pattern Generator (CPG) model. The wave CPG model is a mathematical model of nonlinear oscillators and generates rhythmic movements of the legs. The gait generation and the walking speed control are achieved by controlling the virtual energy of the oscillators (Hamiltonian). A real robot experiment showed the relationship to the Hamiltonian, the actual energy consumption and the walking speed, and the effectiveness of the proposed method was verified.     

Silhouette-based gait recognition via deterministic learning

In this paper, we present a new silhouette-based gait recognition method via deterministic learning theory, which combines spatio-temporal motion characteristics and physical parameters of a human subject by analyzing shape parameters of the subject?s silhouette contour. It has been validated only in sequences with lateral view, recorded in laboratory conditions. The ratio of the silhouette?s height and width (H–W ratio), the width of the outer contour of the binarized silhouette, the silhouette area and the vertical coordinate of centroid of the outer contour are combined as gait features for recognition. They represent the dynamics of gait motion and can more effectively reflect the tiny variance between different gait patterns. The gait recognition approach consists of two phases: a training phase and a test phase. In the training phase, the gait dynamics underlying different individuals? gaits are locally accurately approximated by radial basis function (RBF) networks via deterministic learning theory. The obtained knowledge of approximated gait dynamics is stored in constant RBF networks. In the test phase, a bank of dynamical estimators is constructed for all the training gait patterns. The constant RBF networks obtained from the training phase are embedded in the estimators. By comparing the set of estimators with a test gait pattern, a set of recognition errors are generated, and the average L₁ norms of the errors are taken as the similarity measure between the dynamics of the training gait patterns and the dynamics of the test gait pattern. The test gait pattern similar to one of the training gait patterns can be recognized according to the smallest error principle. Finally, the recognition performance of the proposed algorithm is comparatively illustrated to take into consideration the published gait recognition approaches on the most well-known public gait databases: CASIA, CMU MoBo and TUM GAID.     

基于改进鲸鱼优化算法的外骨骼机器人步态检测

针对传统的外骨骼机器人步态检测算法中的信息单一化、准确率低、易陷入局部最优等问题，提出基于改进鲸鱼算法优化的支持向量机（IWOA-SVM）的外骨骼机器人步态检测算法，即在鲸鱼优化算法（WOA）中引入遗传算法（GA）的选择、交叉、变异操作，进而去优化支持向量机（SVM）的惩罚因子与核参数，再使用参数优化后的SVM建立分类模型，从而扩大算法的搜索范围，减小算法陷入局部最优的概率。首先，使用混合传感技术采集步态数据，即通过足底压力传感器和膝关节、髋关节角度传感器采集外骨骼机器人的运动数据，并作为步态检测系统的输入；然后，使用门限法对步态相位进行划分并标记标签；最后，将足底压力信号与髋关节、膝关节角度信号融合作为输入，使用IWOA-SVM算法完成对步态的检测。对6个标准测试函数进行仿真实验，并与GA、粒子群优化（PSO）算法、WOA进行比较，数值实验表明，改进鲸鱼优化算法（IWOA）的鲁棒性、寻优精度、收敛速度均优于其他优化算法。通过分析不同穿戴者的步态检测结果发现，准确率可达98.8%，验证了所提算法在新一代外骨骼机器人中的可行性和实用性，并与基于遗传优化算法的支持向量机（GA-SVM）、基于粒子群优化算法的支持向量机（PSO-SVM）、基于鲸鱼优化算法的支持向量机（WOA-SVM）算法进行比较，结果表明，该算法识别准确率分别提高了5.33%、2.70%、1.44%，能够对外骨骼机器人的步态进行有效检测，进而实现外骨骼机器人的精确控制及稳定行走。     

可穿戴下肢外骨骼足底压力检测系统设计与研究

可穿戴下肢外骨骼能够柔性跟随人体运动的前提是足底压力检测数据符合人体步态特征,因此研究外骨骼足底压力检测系统具有重要意义。针对现有穿戴式下肢外骨骼系统压力传感器输入输出存在非线性误差,不能准确检测足底压力变化的缺点,选择电阻式压力传感器,通过理论计算确定曲线拟合方法可行性,并将曲线拟合方法引入足底压力检测系统设计中。研究结果表明,该方式采集数据特征明显,曲线拟合的系统线性度较小,可靠性高,适用足底压力检测系统。     

一种基于隐马尔科夫模型的步态识别算法

步态是指人体走路时的姿态,步态识别是近年来生物特征识别领域一个备受关注的研究方向。步态阶段的区分是步态识别的重要内容。以隐马尔科夫模型(HMM)为基础,基于安装在膝关节的编码器和大腿部的加速度传感器,在外骨骼辅助行走中识别步态的不同阶段。首先进行数据预处理和特征提取;其次对隐马尔科夫步态识别算法进行设计,包括结构的建立、参数的训练和最终的识别;最后对性能进行评估,总体正确率达到91.06%,说明HMM用于步态阶段识别具有较好的性能。     

基于柔性力敏传感器的左右脚动态识别方法

在利用柔性力敏传感器获取动态足底压力分布数据时,能够准确快速自动区分左右脚的数据将极大提升数据的可视性和分析的便利性。为此,提出了一种基于足底压力和脚印外观形状的左右脚动态识别方法。首先,基于足底动力学原理,利用连通域的图像分割算法对足底压力数据进行聚类分析,得到每一步压力脚印的时间和坐标范围;在此基础上进一步分离出完整的单步压力数据;最后利用单步压力数据刻画脚印轮廓,并根据轮廓的外观特征进行左右脚识别。本文提出的方法可应用于步态分析、临床辅助诊断、步态识别等领域。通过108个实测数据样本的测试表明：本文方法的识别率高达94.5%,并具有较好的鲁棒性。     

A Novel Method of Gait Synthesis for Bipedal Fast Locomotion

Common methods of gait generation of bipedal locomotion based on experimental results, can successfully synthesize biped joints’ profiles for a simple walking. However, most of these methods lack sufficient physical backgrounds which can cause major problems for bipeds when performing fast locomotion such as running and jumping. In order to develop a more accurate gait generation method, a thorough study of human running and jumping seems to be necessary. Most biomechanics researchers observed that human dynamics, during fast locomotion, can be modeled by a simple spring loaded inverted pendulum system. Considering this observation, a simple approach for bipedal gait generation in fast locomotion is introduced in this paper. This approach applies a nonlinear control method to synchronize the biped link-segmental dynamics with the spring-mass dynamics. This is done such that while the biped center of mass follows the trajectory of the mass-spring model, the whole biped performs the desired running/jumping process. A computer simulation is done on a three-link under-actuated biped model in order to obtain the robot joints’ profiles which ensure repeatable hopping. The initial results are found to be satisfactory, and improvements are currently underway to explore and enhance the capabilities of the proposed method.