本文關鍵詞：基于改進GA-BP算法的SMAW工藝參數(shù)的優(yōu)化　出處：《內蒙古科技大學》2015年碩士論文　論文類型：學位論文

  更多相關文章： 手工電弧焊 焊接變形 GA-BP算法 工藝參數(shù)優(yōu)化

【摘要】：焊接技術是現(xiàn)代機械制造的關鍵工藝技術，其中以手工電弧焊最為常見，但容易產(chǎn)生較嚴重的焊接變形是手工電弧焊最大的缺陷。本文應新宏昌重工有限責任公司技術部的研發(fā)需求，針對該公司生產(chǎn)的自卸車車廂焊接變形較為嚴重的問題，結合公司焊接生產(chǎn)實際，以實用、可行為基本原則，，在深入了解焊接變形形成機理、影響因素和控制技術的基礎上，從優(yōu)化焊接工藝參數(shù)的角度出發(fā)，采用改進型GA-BP算法來對該公司手工電弧焊焊接工藝進行優(yōu)化，以期達到解決產(chǎn)品焊接質量缺陷的目的。 本文首先對新宏昌重工有限責任公司現(xiàn)有的焊接工藝進行全面分析，客觀判斷導致焊接變形產(chǎn)生的工藝因素，確定工藝優(yōu)化方向。在此基礎上，針對性的選擇基于改進GA-BP算法的焊接變形預測策略和多目標工藝參數(shù)優(yōu)化相結合的策略，來實現(xiàn)對焊接工藝參數(shù)的優(yōu)化。 因焊接工藝參數(shù)與焊接變形之間的非線性關系高度復雜，本文以焊接電流、焊接電壓、焊接速度和冷卻速度為輸入，以橫向收縮變形和角變形為輸出，構建了小樣本、高精度、基于改進GA-BP算法的變形預測模型，在各項實測數(shù)據(jù)的基礎上，利用該GA-BP模型強大的非線性關系識別能力，來尋找焊接工藝參數(shù)與焊接變形之間的非線性關系，奠定多目標工藝參數(shù)優(yōu)化的基礎。 在多目標工藝參數(shù)優(yōu)化中，以GA-BP變形預測模型所尋找得到的工藝參數(shù)與焊接變形之間的非線性關系，來代替多目標工藝參數(shù)優(yōu)化中的目標函數(shù)，并結合工廠焊接生產(chǎn)實際情況選定優(yōu)化變量約束條件、即焊接工藝參數(shù)的取值范圍，構建基于正交原則的優(yōu)化求解空間，以橫向收縮變形和角變形最小為優(yōu)化目標，建立多目標工藝參數(shù)優(yōu)化模型，用于實現(xiàn)焊接工藝參數(shù)的優(yōu)化。 此外，本文合理設計試驗方案，利用真實的實測數(shù)據(jù)對基于改進GA-BP算法的變形預測模型的預測效率、多目標工藝參數(shù)優(yōu)化模型優(yōu)化結果的有效性進行了驗證，驗證表明本文所提出的焊接工藝參數(shù)優(yōu)化策略可以通過實現(xiàn)對焊接工藝參數(shù)的優(yōu)化有效減小焊接變形量。并在此基礎上，利用Matlab平臺開發(fā)出基于改進GA-BP算法的SMAW焊接工藝參數(shù)優(yōu)化系統(tǒng)，大大提升了本文核心理論的的工程應用價值。
[Abstract]:Welding technology is the key technology of modern mechanical manufacturing, in which manual arc welding is the most common. But more serious welding deformation is the biggest defect of manual arc welding. This paper should meet the research and development requirements of the technical department of Xinhongchang heavy Industry Co., Ltd. Aiming at the serious problem of welding deformation of dump truck car produced by the company, combined with the actual welding production of the company, taking practical and feasible as the basic principle, the forming mechanism of welding deformation is deeply understood. On the basis of influencing factors and control technology, from the point of view of optimizing welding process parameters, the improved GA-BP algorithm is adopted to optimize the welding process of manual arc welding in this company. In order to solve the welding quality defects. In this paper, the existing welding process of Xinhongchang heavy Industry Co., Ltd. is analyzed comprehensively, the process factors that lead to welding deformation are judged objectively, and the direction of process optimization is determined. The welding deformation prediction strategy based on the improved GA-BP algorithm and the multi-objective process parameter optimization strategy are selected to optimize the welding process parameters. Because the nonlinear relationship between welding parameters and welding deformation is highly complex, this paper takes welding current, welding voltage, welding speed and cooling speed as input, and takes transverse shrinkage deformation and angular deformation as output. The deformation prediction model based on the improved GA-BP algorithm is constructed with small sample and high precision. Based on the measured data, the strong nonlinear relationship recognition ability of the GA-BP model is utilized. To find the nonlinear relationship between welding process parameters and welding deformation, and lay the foundation of multi-objective process parameters optimization. In the optimization of multi-objective process parameters, the nonlinear relationship between process parameters and welding deformation obtained by GA-BP deformation prediction model is used to replace the objective function in multi-objective process parameter optimization. Combined with the actual situation of factory welding production, the constraints of optimization variables, that is, the range of welding process parameters, are selected, and the optimization solution space based on orthogonal principle is constructed. Aiming at minimum transverse shrinkage and angle deformation, a multi-objective process parameter optimization model is established to optimize welding process parameters. In addition, the experiment scheme is designed reasonably, and the prediction efficiency of the deformation prediction model based on the improved GA-BP algorithm is obtained by using the real measured data. The effectiveness of the optimization results of multi-objective process parameters optimization model is verified. The results show that the proposed optimization strategy can effectively reduce the welding deformation by optimizing the welding parameters. The optimization system of welding parameters for SMAW welding based on improved GA-BP algorithm is developed by using Matlab platform, which greatly enhances the engineering application value of the core theory of this paper.
【學位授予單位】：內蒙古科技大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：TG404

【參考文獻】

相關期刊論文 前5條

1 蔡廣宇;楊家軍;李永久;;弧焊機器人焊接工藝參數(shù)設計系統(tǒng)研究[J];組合機床與自動化加工技術;2009年11期

2 吳濤;;用遺傳算法優(yōu)化神經(jīng)網(wǎng)絡權值[J];湛江師范學院學報;2007年03期

3 郭釗;;Q345鋼的焊接工藝及其應用研究[J];科技創(chuàng)新導報;2008年26期

4 劉松;賈東樂;;自卸車焊接變形的控制和矯正[J];專用汽車;2011年06期

5 梁毅;劉世洪;;基于遺傳算法優(yōu)化的BP神經(jīng)網(wǎng)絡的組合預測模型方法研究[J];中國農業(yè)科學;2012年23期

本文編號：1436334