【作者】 李芳宇；

【导师】 孙守迁；

【作者基本信息】 浙江大学 ， 数字化艺术与设计， 2009， 博士

【摘要】 随着全民健身运动的展开,普通人群日益关注通过运动达到健身目的,个性化的运动健身是一直得到关注却不易解决的问题。本文进行了基于主客观信息反馈的运动健身服务技术的研究,主要工作如下:开展了基于主客观信息反馈的运动健身疲劳实验研究。以人体蹬踏健身运动为例设计了运动疲劳实验,通过主客观信息获取方式,在实验过程中获取被试者的表面肌电、心电信号及主观疲劳自觉量值。进行了主观疲劳自觉量和摄氧量的相关性分析,对测试肌肉的表面肌电时频域指标及心率和心率变异性指标进行了统计处理和分析,找出了蹬踏健身有氧运动中表征运动疲劳的指标。实验结果表明,当运动强度增加时,受试者的主观疲劳感受逐渐增大,股外侧肌表面肌电信号的RMS和MPF,以及心电信号中的心率、LF、HF、LF/HF指标具有显著变化,可以识别疲劳状态。研究了基于主客观信息反馈的运动疲劳识别,通过对时频域和小波域的生物电信号进行分析处理,提取出与运动疲劳有关的特征信息,在此基础上建立基于隐马尔科夫模型的疲劳识别方法。对时频域和小波域生物电信息特征融合下的疲劳识别结果比较并和客观数据对比,采用小波域特征分解的隐马尔科夫模型对健身者的运动疲劳状态识别结果准确。通过实验及数据处理获得健身者在未疲劳、适度疲劳和过疲劳三种状态下的心率、心率变异性和肌肉疲劳相关的生物电信息,提出了一种在多生理参数综合考虑下,对健身者的运动能力动态调整的模糊控制方法,确立模糊隶属度函数,从而建立起保持适度健身强度的信息反馈控制策略。提出了一种基于遗传算法优化神经网络预测主观疲劳量值的方法,该法通过缩小基于客观信息得出的预测RPE值与主观实测RPE之间的差距,为健身者提供了主客观统一的运动强度,有效扩大了有氧运动的时间,增强了有氧健身运动的效果。在分析面向服务的运动健身必要性的基础上,结合智能体和网络服务的各自优点,提出了一种多智能体支持的运动健身服务中间件,分析了服务中多智能体的服务协调过程,讨论了面向服务的运动健身策略,并提出基于面向服务架构多智能体支持的运动健身服务,在此基础上探讨了多智能体支持的运动健身服务定制。以智能服务模型指导健身者锻炼,提高了健身者有氧运动的耐受能力。在分析了可扩展的运动健身服务资源的组成的基础上,对运动健身服务的参与各方进行了分析和扩展,提出了基于特定器械的个性化运动健身服务流程。利用前面研究得到的关于运动健身主客观疲劳评价的知识,以功率自行车健身为例,设计了一个下肢蹬踏运动健身服务系统,使基于主客观评价运动疲劳的方法和技术得到了应用,以验证方法的有效性。更多还原

【Abstract】 With the implementation of national fitness program, people pay more attention to fitness to keep health. However personal customizing exercise is a hard problem to solve. This dissertation focus on research on the technology of fitness service based on subjective and objective information feedback through bodies’ fatigue experiments. The main work of this dissertation has been shown as following:An experiment on muscle fatigue of low limbs on general population was taken to observe the signal from surface electromyography (EMG) and electrocardiogram (ECG) based on subjective and objective information feedback to research on fatigue during fitness. The correlations of values between rating of perceived exertion (RPE) and V02max were analyzed by correlation analysis. Statisic method was also used ot analyzed the indices of EMG in time frequecncy domain, heart rate and the indices of heart rate variability, and significant indices were found to evaluate fatigue during aerobics. The results showed that with the growth in exerse intensity, RPE increased, the RMS and MPF in EMG in vastus lateralis, heart rate, and LF, HF, LF/HF in ECG can be used to indicate fatigue stages.A time frequency domain and wavelet domain based analysis were given to process the biological signal via extracting features from subjective and objective information concerned with exercise induced fatigue. A hidden markov model was put forward to recognize fatigue stage.Through experiments and data process of bioelectric information, the features of muscle fatigue during the stages of pre-fatigue, moderate fatigue and extreme fatigue were extracted from the signal of heart rate (HR), heart rate variability (HRV) and EMG. A fuzzy based control method on work load according to the exercise capacity was proposed, fuzzy membership functions of multi physiological parameters were established in order to finish the strategy to maintain moderate fitness.A method based on genetic algorithm to optimize artificial neural network of predicting RPE was proposed to narrow the gap between measured and predictive RPE. It provided a united subjective and objective work load during exercise, thus expanded the time during aerobic exercise and enhanced the effect of aerobic exercise.On the basis of analysis the need of service oriented fitness, a middleware combined with the advantage of multi agent and web service for fitness was put forward. The coordination of service process was analyzed and the strategy for fitness service was demonstrated. Then the architecture of multi agent supported fitness service was put forward, a customized fitness service was used on the basis of this architecture. With the guide of intelligent service model, the exerciser enhanced their endurance capacity in aerobic exercise.On the basis of analyzing the extended compositions of fitness resource, a personalized service flowchart for specific fitness equipment was put forward. In order to validate the method this paper suggested, the knowledge of fatigue evaluation in fitness field was used, and a prototype fitness service system for exercise in bicycle ergometer was developed with the method of evaluation of fatigue based on subjective and objective information feedback.更多还原