本文選題：憶阻器 + 神經(jīng)形態(tài)系統(tǒng)�。� 參考：《西南大學》2017年碩士論文

【摘要】：神經(jīng)形態(tài)系統(tǒng)是仿人工神經(jīng)網(wǎng)絡(luò)構(gòu)建的硬件系統(tǒng),可以實現(xiàn)更高的信息處理和容錯能力,被廣泛應(yīng)用于模式識別、機器學習、信號處理、圖像處理等領(lǐng)域。憶阻器作為一個納米級元件,具有非易失性/易失性、記憶性、可塑性、低功耗等特點,可以作為一個天然的突觸。基于憶阻器的交叉架構(gòu)則可以作為神經(jīng)形態(tài)系統(tǒng)中天然的權(quán)值矩陣。由于不同的憶阻器具有不同的特性,為滿足不同的需求,基于各類憶阻器設(shè)計的神經(jīng)形態(tài)系統(tǒng)也越來越豐富。但這些系統(tǒng)多數(shù)存在三個問題,其一,通過傳統(tǒng)工具對測試樣本進行了大量的預(yù)處理;其二,網(wǎng)絡(luò)的訓(xùn)練過程通常是通過線下系統(tǒng)實現(xiàn)的,而測試過程則是在神經(jīng)形態(tài)電路系統(tǒng)上實現(xiàn)。其三,大多神經(jīng)形態(tài)系統(tǒng)仿照傳統(tǒng)數(shù)字系統(tǒng)的處理方法,而忽略了人腦的獨特特性,比如遺忘特性。針對以上問題,本文構(gòu)建了一種基于方差相關(guān)學習算法的遺忘憶阻神經(jīng)形態(tài)系統(tǒng),并將該系統(tǒng)成功應(yīng)用于模式識別。在進行有效的手寫數(shù)字圖像識別之外,還研究了遺忘速率與識別效率之間的關(guān)系。本文的具體研究內(nèi)容和成果如下:1、對經(jīng)典惠普憶阻器和遺忘憶阻器的內(nèi)部機制進行了闡述說明,并對其數(shù)學模型進行了理論推導(dǎo)。在一維遺忘憶阻器模型的基礎(chǔ)上,介紹了改進后的三維遺忘憶阻器模型,并給出了三維遺忘憶阻器模型的單、雙極和單雙可逆條件。通過建立SPICE仿真模型,對這三種憶阻器模型的內(nèi)部特性和突觸行為進行了詳細的比較和分析。2、基于一維遺忘憶阻器模型設(shè)計了一種神經(jīng)形態(tài)電路系統(tǒng)。該系統(tǒng)是包含自學習電路系統(tǒng)、訓(xùn)練電路系統(tǒng)及識別電路系統(tǒng)的多層集成系統(tǒng),可以實現(xiàn)樣本的在線訓(xùn)練和識別功能。針對系統(tǒng)不同層所實現(xiàn)功能的不同,給出了各個部分的電路原理設(shè)計和功能仿真。3、基于樣本的群體特征和個體特征,提出方差相關(guān)學習算法,實現(xiàn)對憶阻交叉架構(gòu)矩陣在線的訓(xùn)練。該方法可以有效簡化樣本的預(yù)處理工作,同時便于電路系統(tǒng)的實現(xiàn)。4、將神經(jīng)形態(tài)系統(tǒng)應(yīng)用于手寫數(shù)字圖像的模式訓(xùn)練及識別。通過仿真驗證了系統(tǒng)的功能和有效性。另外,進一步研究了遺忘憶阻器的遺忘因子?對識別結(jié)果的影響,發(fā)現(xiàn)不同區(qū)間的遺忘因子對識別效果具有不同程度的影響。
[Abstract]:Neural Morphology system is a hardware system constructed by artificial neural network, which can achieve higher information processing and fault tolerance. It is widely used in pattern recognition, machine learning, signal processing, image processing and other fields. As a nanoscale device, the memory device has the characteristics of non-volatile / volatile, memory, plasticity and low power consumption, so it can be used as a natural synapse. The cross structure based on amnesia can be used as a natural weight matrix in neural morphological system. Because different amnesia devices have different characteristics, in order to meet different needs, neural morphological systems based on various kinds of amnesia devices are more and more abundant. However, most of these systems have three problems: first, a large number of test samples are preprocessed by traditional tools; second, the training process of network is usually realized by offline system. The test process is implemented on the neural morphological circuit system. Third, most neural morphological systems mimic the traditional digital systems, while ignoring the unique characteristics of the human brain, such as forgetting. In order to solve the above problems, a forgetfulness and amnesia neural morphological system based on variance correlation learning algorithm is constructed in this paper, and the system is successfully applied to pattern recognition. In addition to effective handwritten digital image recognition, the relationship between forgetting rate and recognition efficiency is also studied. The specific research contents and results of this paper are as follows: 1. The internal mechanisms of the classic Hewlett-Packard (HP) amnesia and the amnesia amnesia are explained, and its mathematical model is theoretically deduced. On the basis of one dimensional amnesia model, the improved three dimensional amnesia model is introduced, and the single, bipolar and single double reversible conditions of the three dimensional amnesia model are given. The internal characteristics and synaptic behavior of the three kinds of amnesia models are compared and analyzed in detail by establishing the SPICE simulation model. A neural morphological circuit system is designed based on the one-dimensional amnesia model. The system is a multi-layer integrated system including self-learning circuit system, training circuit system and recognition circuit system, which can realize on-line training and recognition of samples. Aiming at the different functions realized by different layers of the system, the circuit principle design and function simulation of each part are given. Based on the group and individual characteristics of the sample, the variance correlation learning algorithm is proposed. The online training of memory cross-architecture matrix is realized. This method can simplify the preprocessing of the sample effectively and is convenient for the realization of the circuit system. The neural morphological system is applied to the pattern training and recognition of handwritten digital image. The function and effectiveness of the system are verified by simulation. In addition, the forgetfulness factor of amnesia is further studied. It is found that the forgetting factors of different intervals have different effects on the recognition results.
【學位授予單位】：西南大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：TN60;TP391.4

【參考文獻】

相關(guān)期刊論文 前3條

1 邵楠;張盛兵;邵舒淵;;具有突觸特性憶阻模型的改進與模型經(jīng)驗學習特性機理[J];物理學報;2016年12期

2 閔國旗;王麗丹;段書凱;;離子遷移憶阻混沌電路及其在語音保密通信中的應(yīng)用[J];物理學報;2015年21期

3 許小勇;鐘太勇;;三次樣條插值函數(shù)的構(gòu)造與Matlab實現(xiàn)[J];兵工自動化;2006年11期

，

本文編號：1866524