本文選題：地空通信信號 + K-均值聚類算法　； 參考：《西華大學(xué)》2015年碩士論文

【摘要】：隨著電磁環(huán)境的日趨復(fù)雜,無線電民航地空通信頻段受干擾事件日益嚴(yán)重,給人們的生命財產(chǎn)安全帶來了很大的威脅。如何準(zhǔn)確、高效的識別出地空通信異常信號成為無線電日常監(jiān)測工作的重要目標(biāo),同時具有很高的理論價值及研究意義。地空通信業(yè)務(wù)一般是語音通信,具有偶發(fā)性、出現(xiàn)概率低、危害性強的特點。因此在地空通信異常信號識別中,有效利用直觀的語音信息,選擇合適的分類器成為準(zhǔn)確、快速、高效、自動識別地空通信異常信號的關(guān)鍵。K-均值聚類算(K?means)法在信號特征處理及信號識別中已經(jīng)得到了廣泛的應(yīng)用。但是該算法由于聚類中心初始化問題的存在,使得最終識別效率穩(wěn)定性無法得到保證。而支持向量機(SVM)則擅長于解決復(fù)雜的信號分類問題,在圖像處理、醫(yī)學(xué)研究等領(lǐng)域應(yīng)用廣泛。本文將對基于智能優(yōu)化算法的支持向量機參數(shù)選擇方法做進一步的研究,給出一種識別效率高、消耗時間短的支持向量機分類器,并將其運用到地空通信干擾信號識別當(dāng)中。具體研究內(nèi)容如下:1.探究將無線電地空通信音頻信號作為無線電地空通信異常信號識別依據(jù)的可行性。利用meansK?算法完成對地空通信音頻信號特征集合的構(gòu)建;通過歐式距離判別法進行了判別實驗。2.提出一種基于引力搜索算法的支持向量機分類器。根據(jù)支持向量機分類原理及最優(yōu)化理論建立支持向量機參數(shù)選擇的優(yōu)化模型,依據(jù)該模型完成上述分類器的構(gòu)建。3.將上述分類器用于地空通信音頻信號的識別當(dāng)中。通過對比實驗,證明該方法具有識別率高、耗時短等特點。
[Abstract]:With the increasing complexity of electromagnetic environment, the interference in the frequency band of radio civil aviation ground-to-air communication is becoming more and more serious, which brings a great threat to the safety of people's life and property. How to accurately and efficiently identify the abnormal signals of ground-to-air communication has become an important target in the daily monitoring of radio, and has high theoretical value and research significance at the same time. Ground-to-air communication service is usually voice communication, which has the characteristics of accidental occurrence, low probability of occurrence and strong harmfulness. Therefore, in the recognition of abnormal signals in ground-to-air communication, it becomes accurate, fast and efficient to use the intuitionistic speech information effectively and select the appropriate classifier. The key method of automatic recognition of abnormal signals in ground-to-air communications. The K-Means clustering method has been widely used in signal feature processing and signal recognition. However, due to the problem of clustering center initialization, the stability of the final recognition efficiency can not be guaranteed. Support Vector Machine (SVM) is good at solving complex signal classification problems, and is widely used in image processing, medical research and other fields. In this paper, the parameter selection method of support vector machine based on intelligent optimization algorithm is further studied, and a support vector machine classifier with high recognition efficiency and short time consumption is proposed and applied to ground to air communication interference signal recognition. The specific contents of the study are as follows: 1. To explore the feasibility of using radio ground-air communication audio signal as the basis for the identification of radio ground-air communication abnormal signals. Using Means K? The algorithm is used to construct the audio signal feature set of ground-to-air communication, and the Euclidean distance discriminant method is used in the discriminant experiment. 2. A support vector machine classifier based on gravitational search algorithm is proposed. According to the classification principle of support vector machine and the optimization theory, the optimization model of parameter selection of support vector machine is established, and the construction of the classifier. 3 is completed according to the model. The classifier is used in the recognition of the ground-air communication audio signal. The comparison experiment shows that the method has the advantages of high recognition rate and short time consuming.
【學(xué)位授予單位】：西華大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2015
【分類號】：TN972;TP18

【參考文獻】

相關(guān)期刊論文 前2條

1 賈燕花;徐蔚鴻;;K-means聚類和支持向量機結(jié)合的文本分類研究[J];計算機工程與應(yīng)用;2010年22期

2 陳榮元;蔣加伏;;基于聚類算法和層次支持向量機的人臉識別方法[J];計算技術(shù)與自動化;2006年01期

，

本文編號：2032552