本文選題：神經網(wǎng)絡 + 故障診斷��； 參考：《湖南師范大學》2015年碩士論文

【摘要】：信息處理技術在當今得到了快速發(fā)展,電子設備中的電路變得日益復雜,由模擬電路引起的設備故障問題,要得到有效處理卻日益棘手。集成電路集成度的變高,元器件本身固有的不穩(wěn)定性等原因給快速定位故障及處理故障帶來更大挑戰(zhàn)。面對眾多出現(xiàn)的問題,傳統(tǒng)故障診斷方法已經不能滿足社會發(fā)展需求,新的診斷技術迫在眉睫。各國研究者開始嘗試新的理論研究,其中神經網(wǎng)絡作為智能技術運用于模擬電路故診斷研究得到快速發(fā)展,在新的診斷技術方面開辟了新路徑,并在一段時間內取得了豐碩的成果。現(xiàn)如今,廣大學者開始重視將小波分析,遺傳算法等多種理論及其融合理論結合神經網(wǎng)絡進行故障診斷的新技術,這為智能化故障診斷技術提供了新的思路。LabVIEW軟件作為一款功能強大的圖形編程軟件,可以提供良好的人工交互界面,已經開始運用于故障診斷技術中,為實現(xiàn)故障診斷的簡易化提供了便捷之路。本文以新的診斷技術為背景,將小波分析,遺傳算法理論融合到神經網(wǎng)絡,結合虛擬儀器(Lab VIEW平臺),實現(xiàn)電路故障的可視化診斷。介紹了模擬電路故障診斷的研究背景意義、國內外發(fā)展現(xiàn)狀、存在問題及分類方法。概述人工神經網(wǎng)絡理論,包括其特點、應用以及學習方式。以BP神經網(wǎng)絡理論為基礎,對小波神經網(wǎng)絡結構進行構造及其改進算法進行詳細講解,通過仿真實例進行驗證所提算法的正確性,其中包括使用軟件ORCAD10.5對待診斷電路進行原始數(shù)據(jù)提取;利用MATLAB軟件平臺編程對數(shù)據(jù)進行多分辨分析,提取故障特征值,構造樣本集;基于小波神經網(wǎng)絡故障診斷方法的實現(xiàn):針對神經網(wǎng)絡權值問題,利用遺傳算法進行優(yōu)化,改善網(wǎng)絡性能,最后通過Lab VIEW軟件平臺實現(xiàn)編寫程序的圖形化,搭建神經網(wǎng)絡模擬電路故障診斷系統(tǒng)界面,實現(xiàn)診斷過程的可視化,操作簡易化。
[Abstract]:With the rapid development of information processing technology, the circuits in electronic devices are becoming more and more complex, but the problems caused by analog circuits are becoming more and more difficult to deal with effectively. The high integration of integrated circuits and the inherent instability of components bring greater challenges to fast fault location and fault handling. In the face of many problems, the traditional fault diagnosis method can not meet the needs of social development, new diagnosis technology is urgent. Researchers all over the world began to try new theoretical research, in which neural network as an intelligent technology used in analog circuits so that the rapid development of diagnostic research, in the new diagnostic technology opened up a new path. And in a period of time has achieved fruitful results. Nowadays, many scholars begin to attach importance to the new technology of fault diagnosis, which combines wavelet analysis, genetic algorithm and fusion theory with neural network. This provides a new idea for intelligent fault diagnosis technology. LabVIEW software, as a powerful graphical programming software, can provide a good interactive interface, and has been used in fault diagnosis technology. It provides a convenient way to realize the simplification of fault diagnosis. In this paper, based on the new diagnosis technology, wavelet analysis and genetic algorithm theory are combined into neural network, and the visual diagnosis of circuit fault is realized by combining virtual instrument with LabLab VIEW platform. This paper introduces the research background significance, development status, existing problems and classification methods of analog circuit fault diagnosis. This paper summarizes the theory of artificial neural network, including its characteristics, applications and learning methods. Based on BP neural network theory, the structure of wavelet neural network structure and its improved algorithm are explained in detail, and the correctness of the proposed algorithm is verified by a simulation example. The software ORCAD10.5 is used to extract the original data from the diagnosis circuit, the multi-resolution analysis of the data is carried out by using the MATLAB software platform, the fault characteristic value is extracted, and the sample set is constructed. The realization of fault diagnosis method based on wavelet neural network: aiming at the weight problem of neural network, the genetic algorithm is used to optimize the network performance and improve the network performance. Finally, the graphical programming is realized through Lab VIEW software platform. The interface of the neural network analog circuit fault diagnosis system is built to realize the visualization of the diagnosis process and the simplicity of operation.
【學位授予單位】：湖南師范大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：TP183;TN710

【參考文獻】

相關期刊論文 前1條

1 楊士元;一種新的模擬電路K故障診斷方法[J];清華大學學報(自然科學版);1992年01期

，

本文編號：1782249