【摘要】：構(gòu)建適應(yīng)性強、精確度高的定位模型及定位算法是聲源定位系統(tǒng)要解決的核心問題。然而,目前對機器人聽覺感知的研究主要側(cè)重于對語音的辨識和理解上,在定位模型及算法方面由于缺少機理性的研究。在實際應(yīng)用中只初步實現(xiàn)了對遠場聲源的定向,缺少對高精度定距以及近場聲源定位的處理方法。在聲信號處理領(lǐng)域中對這一問題的研究由于受模型復(fù)雜度和傳聲器陣列結(jié)構(gòu)(尺寸和維數(shù))的限制,一直沒能從理論上突破這一瓶頸,這大大降低了聲源定位系統(tǒng)的適應(yīng)性和實用性。本文研究了一種多元傳聲陣列結(jié)構(gòu)優(yōu)化及其定位算法,重點分析了不確定條件下多元傳聲器陣列的適應(yīng)性、適用性和準確性;研究了不確定條件下的傳聲器陣列的定位模型與定位穩(wěn)定性,并且提出了可行的定位模型及計算方法;設(shè)計了傳聲器陣列的結(jié)構(gòu),通過時延估計的定位方法,采用五元陣列交叉定位的方法解決了聲源定位系統(tǒng)的適應(yīng)性和實用性的問題;仿真驗證了交叉定位算法的有效性并總結(jié)出一些定位規(guī)律;借助移動機器人實驗平臺,實際運用了優(yōu)化陣列及其定位算法,實驗測試表明該方法中距離對聲達時間差誤差不敏感,論文研究的五元陣列結(jié)構(gòu)及定位算法可實現(xiàn)較高精度的聲源目標定位。
[Abstract]:It is the core problem to solve the problem of sound source location system to construct the localization model and algorithm with high adaptability and high accuracy. However, at present, the research of robot auditory perception mainly focuses on the recognition and understanding of speech, due to the lack of mechanical research on localization model and algorithm. In practical application, only the orientation of far-field sound source is realized preliminarily, and the processing method of high-precision distance fixing and near-field sound source location is lacking. Due to the limitation of model complexity and microphone array structure (dimension and dimension), the research on this problem in the field of acoustic signal processing has not been able to break through this bottleneck theoretically. This greatly reduces the adaptability and practicability of the sound source location system. In this paper, the structure optimization and location algorithm of multivariate acoustic array are studied, and the adaptability, applicability and accuracy of multivariate microphone array under uncertain conditions are analyzed. The localization model and stability of microphone array under uncertain conditions are studied, and a feasible location model and calculation method are proposed. The structure of microphone array is designed. The problem of adaptability and practicability of sound source location system is solved by the method of time delay estimation and five-element array cross location. Simulation results show that the algorithm is effective and some localization rules are summarized. With the help of the mobile robot experimental platform, the optimized array and its localization algorithm are used in practice. The experimental results show that the distance in this method is not sensitive to the error of the time difference of acoustic arrival. The five-element array structure and localization algorithm studied in this paper can achieve high accuracy target location of sound source.
【學(xué)位授予單位】：河北工業(yè)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2015
【分類號】：TN641

【參考文獻】

相關(guān)碩士學(xué)位論文 前1條

1 葉一枝;基于時延估計的聲源定位系統(tǒng)的設(shè)計與實現(xiàn)[D];上海交通大學(xué);2011年

，

本文編號：2414276