论文部分内容阅读
针对穿戴假肢的初期患者跌倒可能性较高的问题,提出一种基于相关性分析的模糊自适应反馈调节跌倒预警方法.应用支持向量机回归方法改进不同受试者的相关性分析模板,并对2种传感器的预测值进行融合以进行跌倒判别.由于不同患者应对跌倒时身体反应不同,采用具有反馈调节单元的模糊自适应变权重组合预测算法对不同患者的融合权重进行参数寻优,以增强下肢假肢跌倒预警系统的灵活性,最后采集多名受试者的数据进行分析论证该方法的可行性.实验结果表明,其预警正确率可达95%.该系统可实际应用于初期患者的康复训练,以提高假肢的安全性.
Aiming at the problem that the initial patients who wear prosthetic limbs fall more likely, this paper proposes a fuzzy adaptive feedback control method based on the correlation analysis to predict the fall of early-warning patients.Using support vector machine regression to improve the correlation analysis template of different subjects, Two kinds of sensors were fused for the fall judgment.Because different patients responded to different body reactions during fall, fuzzy adaptive variable weight combination forecasting algorithm with feedback adjustment unit was used to optimize the fusion weight of different patients in order to enhance Lower extremity prosthesis fall early warning system flexibility, and finally collected more than one subject’s data to demonstrate the feasibility of the method.Experimental results show that the correct rate of up to 95% of the warning.The system can be applied to the initial rehabilitation of patients Training to improve the safety of prostheses.