论文部分内容阅读
针对人体运动随机性和非结构地面等因素造成在不同类型地面上足底力(GRF)差异大的情况,研发了一套配有鞋底压力传感器、用于实时检测足底力变化的实验靴,提出了基于PSO-SVM(基于粒子群优化算法的支持向量机)的步态分类方法.根据足底受力云图,该实验靴中冗余布置了7枚压力传感器.对人行走在步行机(5 km/h)、水平硬路面和野外草地上的足底力进行了采集和处理.将基本组的足底力作为训练集,预设对应的标签值.基于这些训练集,构建了普通的分类器I和基于粒子群优化算法的支持向量分类器II,并分别利用它们对水平硬路面和野外草地行走的数据进行分类检验.实验结果表明,与普通的分类器I相比,该优化算法不仅对复杂地面上的行走步态分类具有明显的优势,对水平硬路面和野外草地的行走步态识别正确率分别提升了32.9%和42.8%,而且能在某些鞋底传感器发生故障后保持较快的寻优速度和较好的鲁棒性.
In view of the large discrepancy of plantar power (GRF) on different types of ground due to human motion randomness and unstructured ground and so on, a set of test boots equipped with a sole pressure sensor for detecting the change of plantar force in real time has been developed Gait classification method based on PSO-SVM (support vector machine based on particle swarm optimization algorithm) According to the plantar stress cloud map, seven pressure sensors are redundantly arranged in this test shoe, / h), the horizontal hard pavement and the plantar power on the wild grassland were collected and processed.The plantar force of the basic group was used as the training set, and the corresponding label value was preset.On the basis of these training sets, an ordinary classifier I and Particle Swarm Optimization (PSO) -based support vector classifier (II) is used to classify the data of horizontal hard pavement and grassland walking respectively.The experimental results show that compared with the conventional classifier I, On the walking gait classification has obvious advantages, the level of hard pavement and grassland walking gait recognition accuracy increased by 32.9% and 42.8%, respectively, and in some shoe sensors occur After the fault to maintain a faster search speed and better robustness.