【正文】 本科學(xué)生畢業(yè)設(shè)計 基于 C 語言的 BP神經(jīng)網(wǎng)絡(luò)預(yù)測程序開發(fā) 系部名稱： 汽車與交通工程學(xué)院 專業(yè)班級： 車輛 學(xué)生姓名： 指導(dǎo)教師： 職 稱： 實驗師 The Graduation Design for Bachelor39。s Degree Development on BP Neural Network Prediction Program Based on C Language Candidate： Gao Xiaolin Specialty： Vehicles Engineering Class： 0711 Supervisor： Lecturer Wang Yuexin Heilongjiang Institute of Technology I 摘 要 人工神經(jīng)網(wǎng)絡(luò)理論是新近發(fā)展起來的交叉學(xué)科，采用物理器件或計算機軟硬件模擬生物體中神經(jīng)細胞的某些結(jié)構(gòu)與功能，進而將其應(yīng)用于工程領(lǐng)域，尤其適合高度復(fù)雜的非線性動力學(xué)系統(tǒng)仿真。人工神經(jīng)網(wǎng)絡(luò)已經(jīng)在組合優(yōu)化、模式識別、圖象處理、自動控制、信號處理、機器人和人工智能等領(lǐng)域得到廣泛應(yīng)用，尤其在工程領(lǐng)域逐漸受到廣泛重視。人工神經(jīng)網(wǎng)絡(luò)用于非線性動力學(xué)系統(tǒng)研究，近來在汽車動力學(xué)建模與仿真領(lǐng)域也引起了極大關(guān)注。 許多工業(yè)生產(chǎn)過程存在時滯和大時間常數(shù) , 控制難度較大 , 傳統(tǒng)的控制策略對此類控制問題很難取得滿意的效果。為了解決這類問題 , 預(yù)測控制應(yīng)運而生。預(yù)測控制是一種基于模型的控制策略。 反向傳播 (BP) 神經(jīng)網(wǎng)絡(luò)是當(dāng)前應(yīng)用最為廣泛的一種神經(jīng)網(wǎng)絡(luò) , 它結(jié)構(gòu)簡單 , 工作狀態(tài)穩(wěn)定 , 并且已有大量提高網(wǎng)絡(luò)訓(xùn)練速度的改進算法。 關(guān)鍵字 ： BP 神經(jīng)網(wǎng)絡(luò)；應(yīng)用； C 語言 ；汽車保有量；預(yù)測 全套設(shè)計，加 153893706 II ABSTRACT Artificial neural work theory is newly developed the interdisciplinary, adopting physical device or puter hardware and software simulation of the nerve cells in biological structure and function, and then some of its application in engineering field, especially suitable for highly plex nonlinear dynamic system simulation. Artificial neural work has been in the binatorial optimization, pattern recognition, image processing, automatic control, signal processing, robot, and artificial intelligence and other areas to be wide ly applied, especially in engineering areas has been gradually paid more attention . Artificial neural work for nonlinear dynamic system research in automotive dynamics, recently modeling and simulation field also get great attention. Many industrial production process exist time delay and large time constant, bigger control difficulty, traditional control strategies on such controlling problems is difficult to obtain satisfactory results. In order to solve this kind of problem, predictive control arises at the historic moment. Predictive control is a model based control strategy. Back propagation (BP) neural work is currently the most widely used neural work, it is simple in structure, work, and a large number of stable status work algorithm are proposed to increase training speed. Key words: BP neural work； Application； C language； Auto possession； Prediction 目 錄 摘要 ............................................................................................................................... I ABSTRACT ...............................................................................................................II 第 1 章 緒論 .............................................................................................................. 1 人工神經(jīng)網(wǎng)絡(luò)的發(fā)展歷史 .............................................................................. 1 人工神經(jīng)網(wǎng)絡(luò)的特性 ...................................................................................... 2 BP 神經(jīng)網(wǎng)絡(luò) .................................................................................................. 3 第 2 章 神經(jīng)網(wǎng)絡(luò)預(yù)測基本原理 ......................................................................... 5 生物神經(jīng)元模型 .............................................................................................. 5 人工神經(jīng)元模型 .............................................................................................. 6 人工神經(jīng)網(wǎng)絡(luò)模型 ........................................................................................ 10 人工神經(jīng)網(wǎng)絡(luò)的學(xué)習(xí) .................................................................................... 11 本章小結(jié) ........................................................................................................ 12 第 3 章 BP 神經(jīng)網(wǎng)絡(luò)結(jié)構(gòu)及數(shù)學(xué)模型 ........................................................... 13 BP 神經(jīng)網(wǎng)絡(luò)結(jié)構(gòu) ........................................................................................... 13 BP 神經(jīng)元 ....................................................................................................... 14 BP 網(wǎng)絡(luò) ........................................................................................................... 15 正向傳播 .............................................................................................. 15 反向傳播 .............................................................................................. 16 本章小結(jié) ........................................................................................................ 18 第 4 章 BP 網(wǎng)絡(luò)的自學(xué)習(xí)與預(yù)測編程 ........................................................... 19 BP 網(wǎng)絡(luò)的學(xué)習(xí)與預(yù)測神經(jīng)編程 ................................................................... 19 輸入模式順傳播 ............................................................................................ 19 輸出誤差的逆?zhèn)鞑?........................................................................................ 21 循環(huán)記憶訓(xùn)練 ................................................................................................ 24 學(xué)習(xí)結(jié)果的判別 ............................................................................................ 27 對數(shù)據(jù)的預(yù)測程序編寫 ................................................................................ 32 本章小結(jié) ........................................................................................................ 36 第五章 BP 神經(jīng)網(wǎng)絡(luò)在汽車保有量預(yù)測中的應(yīng)用 ................................. 37 汽車保有量預(yù)測的意義 ................................................................................. 37 基于 BP 神經(jīng)網(wǎng)絡(luò)的汽車保有量預(yù)測模型 .................................................. 37 汽車保有量主要影響因素分 析 ........................................................... 37 汽車保有量預(yù)測模型 ........................................................................... 37 實例分析 ................................................