本文關(guān)鍵詞： 運(yùn)動(dòng)狀態(tài)識(shí)別 數(shù)字濾波 步伐檢測(cè) BLE協(xié)議棧 狀態(tài)機(jī) 應(yīng)用規(guī)范　出處：《電子科技大學(xué)》2015年碩士論文　論文類(lèi)型：學(xué)位論文

【摘要】：可穿戴式運(yùn)動(dòng)監(jiān)測(cè)系統(tǒng)是當(dāng)下的研究熱點(diǎn),科技公司和國(guó)內(nèi)外學(xué)者在該領(lǐng)域做了大量工作,并已推出一些知名產(chǎn)品。但是,如今的研究成果和已有產(chǎn)品,在信號(hào)處理、運(yùn)動(dòng)狀態(tài)識(shí)別等方面還存在很多缺陷,精確度也參差不齊,而且研究方法較為局限,產(chǎn)品的營(yíng)銷(xiāo)因素遠(yuǎn)大于技術(shù)因素。因此,該技術(shù)領(lǐng)域還處于探索階段,并沒(méi)有完全成熟的解決方案。本文以運(yùn)動(dòng)監(jiān)測(cè)系統(tǒng)的傳感子系統(tǒng)為研究對(duì)象,采用低功耗藍(lán)牙(BLE)作為通信方式,設(shè)計(jì)帶有加速度傳感器的傳感設(shè)備。根據(jù)數(shù)據(jù)分析和實(shí)際條件,對(duì)運(yùn)動(dòng)監(jiān)測(cè)的實(shí)現(xiàn)中涉及到的算法、研究方法、軟硬件設(shè)計(jì)等方面內(nèi)容進(jìn)行深入研究,主要內(nèi)容包含以下幾個(gè)方面:(1)提出一種結(jié)合仿真結(jié)果的數(shù)字濾波器在可穿戴設(shè)備上實(shí)現(xiàn)的方法。根據(jù)仿真實(shí)驗(yàn)結(jié)果,設(shè)計(jì)了適用于運(yùn)動(dòng)監(jiān)測(cè)的數(shù)字濾波器,濾波后能夠獲得平滑、易于處理的信號(hào)波形,排除信號(hào)毛刺的影響。(2)提出一種多閾值-峰值計(jì)步算法,并建立運(yùn)動(dòng)狀態(tài)機(jī)模型。根據(jù)步行中軀干運(yùn)動(dòng)和手臂擺動(dòng)加速度的周期性變化,提取特征值,進(jìn)行步伐檢測(cè)和步態(tài)判別,并排除外界加速度干擾的影響。然后編寫(xiě)Matlab仿真程序進(jìn)行驗(yàn)證。狀態(tài)機(jī)采用嵌套式結(jié)構(gòu),分為母狀態(tài)和子狀態(tài)兩層,并標(biāo)定狀態(tài)轉(zhuǎn)移條件。(3)提出一種基于SVM分類(lèi)算法的特征參數(shù)有效性的驗(yàn)證方法,并以本次設(shè)計(jì)選取的特征參數(shù)為例,進(jìn)行方法的介紹,通過(guò)對(duì)仿真實(shí)驗(yàn)結(jié)果的分析,討論本次設(shè)計(jì)中特征參數(shù)在運(yùn)動(dòng)監(jiān)測(cè)算法中的作用和有效性。此外,也通過(guò)分類(lèi)預(yù)測(cè)結(jié)果對(duì)運(yùn)動(dòng)監(jiān)測(cè)算法提供改進(jìn)意見(jiàn)。(4)完成運(yùn)動(dòng)監(jiān)測(cè)應(yīng)用軟件的實(shí)現(xiàn)。在BLE協(xié)議棧的應(yīng)用層上,結(jié)合操作系統(tǒng)抽象層(OSAL)的運(yùn)行機(jī)制,采用分步執(zhí)行和狀態(tài)轉(zhuǎn)移的方式,在傳感設(shè)備上完成計(jì)步算法和運(yùn)動(dòng)狀態(tài)機(jī)的程序設(shè)計(jì)。(5)完成運(yùn)動(dòng)監(jiān)測(cè)應(yīng)用規(guī)范的設(shè)計(jì)。在BLE協(xié)議棧中定義運(yùn)動(dòng)監(jiān)測(cè)服務(wù),對(duì)該服務(wù)中的特性及特性屬性做出規(guī)范,保證傳感設(shè)備和智能終端對(duì)運(yùn)動(dòng)監(jiān)測(cè)服務(wù)具有統(tǒng)一的理解。(6)采用本系統(tǒng)與其他兩款知名智能腕帶產(chǎn)品進(jìn)行對(duì)比測(cè)試,以驗(yàn)證本文設(shè)計(jì)的運(yùn)動(dòng)監(jiān)測(cè)系統(tǒng)的性能。證明了該系統(tǒng)具有良好的監(jiān)測(cè)性能。
[Abstract]:Wearable motion monitoring system is the current research hotspot, technology companies and domestic and foreign scholars have done a lot of work in this field, and have launched some well-known products. However, today's research results and existing products. In signal processing, motion state recognition and other aspects, there are still many defects, the accuracy is not uniform, and the research method is more limited, the marketing factor of the product is far greater than the technical factor. The technology is still in the exploration stage, and there is no fully mature solution. In this paper, the sensor subsystem of the motion monitoring system is taken as the research object, and the low power Bluetooth BLEs are used as the communication mode. According to the data analysis and actual conditions, the algorithms, research methods, hardware and software design of motion monitoring are studied. The main contents include the following several aspects: 1) A digital filter based on simulation results is proposed to realize the wearable device. A digital filter for motion monitoring is designed. After filtering, a smooth and easily processed signal waveform is obtained, and the effect of signal burrs is eliminated. (2) A multi-threshold peak-peak step algorithm is proposed. According to the periodic changes of torso motion and arm swing acceleration in walking, the characteristic values are extracted, and the gait detection and gait detection are carried out. And eliminate the impact of external acceleration interference. Then write a Matlab simulation program to verify. The state machine uses nested structure, divided into mother state and sub-state two layers. Calibration condition of state transition. 3) A method to verify the validity of feature parameters based on SVM classification algorithm is proposed, and the method is introduced with the selected feature parameters of this design as an example. Through the analysis of the simulation results, the function and effectiveness of the feature parameters in the motion detection algorithm are discussed. The application software of motion monitoring is also implemented by classifying and forecasting results. It is on the application layer of BLE protocol stack. Combined with the operating system abstract layer (OSAL) running mechanism, the way of step by step execution and state transition is adopted. Complete the design of the step algorithm and the motion state machine on the sensing equipment. 5) complete the design of the motion monitoring application specification. Define the motion monitoring service in the BLE protocol stack. Specifications are made for features and attribute attributes in the service. Ensure that the sensor equipment and intelligent terminals have a unified understanding of motion monitoring services. 6) this system is compared with other well-known smart wrist band products for testing. In order to verify the performance of the motion monitoring system designed in this paper, it is proved that the system has good monitoring performance.
【學(xué)位授予單位】：電子科技大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類(lèi)號(hào)】：TP274

【參考文獻(xiàn)】

相關(guān)期刊論文 前1條

1 高嵩峰;張春林;;人體參數(shù)化運(yùn)動(dòng)學(xué)模型建立方法的研究[J];航天醫(yī)學(xué)與醫(yī)學(xué)工程;2009年05期

相關(guān)碩士學(xué)位論文 前1條

1 王勇;基于力傳感器的飛行模擬器運(yùn)動(dòng)平臺(tái)控制研究[D];南京航空航天大學(xué);2012年

，

本文編號(hào)：1452576