本文關鍵詞： 工程裝備 液壓系統(tǒng) 泄漏仿真 故障診斷 AMESim 支持向量機　出處：《國防科學技術大學》2012年碩士論文　論文類型：學位論文

【摘要】：工程裝備是我軍武器裝備的重要組成部分，主要用以軍隊遂行工程保障任務，其工作穩(wěn)定性、戰(zhàn)場生存能力將直接關系到戰(zhàn)斗工程保障任務的順利完成乃至戰(zhàn)爭的成敗。液壓傳動廣泛應用于我軍現役工程裝備的動力傳遞中，而液壓泄漏一直是工程裝備技術保障的難題之一。開展液壓系統(tǒng)泄漏故障診斷和防泄漏技術研究，探尋有效降低工程裝備液壓泄漏故障的途徑和方法，對提高工程裝備保障能力具有重要的理論指導意義和軍事應用價值。 本文針對液壓系統(tǒng)泄漏故障診斷手段單一、故障樣本難以獲取等問題，以典型工程裝備某重型機械化橋為研究對象，采用面向對象建模仿真和支持向量機故障診斷方法，對其工作裝置液壓系統(tǒng)進行了建模仿真和液壓泄漏故障診斷，，有效提高了小樣本事件故障診斷的準確率，為液壓系統(tǒng)泄漏故障診斷提供了一種新的智能診斷方法。論文主要研究工作及成果如下： （1）系統(tǒng)概括了液壓系統(tǒng)泄漏故障診斷的研究現狀，對引起泄漏的原因和泄漏故障機理進行了詳細分析，將外泄漏和內泄漏分別等效為孔口出流和縫隙流動，分析得到油液黏度和縫隙高度等因素對泄漏量有較大影響，提出了液壓泄漏故障診斷策略。 （2）根據上述制定的泄漏故障診斷策略，運用液壓機械系統(tǒng)建模仿真軟件AMESim建立了某重型機械化橋首跨架設液壓系統(tǒng)工作回路模型，引入頂推液壓缸工作回路泄漏和鎖緊液壓缸內泄漏等泄漏故障進行仿真分析，得到不同泄漏情況下的故障特征，為泄漏故障診斷和預防提供了理論依據。 （3）從泄漏故障仿真中提取壓力、流量等特征參數，導入Matlab求取時域特征值，作為支持向量機進行泄漏故障診斷的數據樣本，運用SVM模式識別軟件包LibSVM訓練所得模型準確診斷出了泄漏故障發(fā)生的部位和引起泄漏的原因，從而驗證了本文將泄漏故障建模仿真與支持向量機故障診斷相結合的方法的可行性和有效性。 （4）從控制液壓油黏度、控制液壓系統(tǒng)縫隙高度和液壓管路的正確裝配等環(huán)節(jié)對如何做好液壓系統(tǒng)防泄漏工作進行了分析研究，提出了預防泄漏故障的對策和方法。
[Abstract]:Engineering equipment is an important part of our army's weapons and equipment. It is mainly used for the military to carry out engineering support tasks and its work stability. Battlefield survivability will be directly related to the successful completion of combat engineering support tasks and even the success or failure of war. Hydraulic transmission is widely used in the power transmission of our active engineering equipment. Hydraulic leakage has always been one of the difficult problems in the technical support of engineering equipment. The research on hydraulic system leakage fault diagnosis and leakage prevention technology is carried out to explore the ways and methods to effectively reduce the hydraulic leakage failure of engineering equipment. It has important theoretical guiding significance and military application value to improve the engineering equipment support ability. Aiming at the problems of single method of hydraulic system leakage fault diagnosis and difficulty in obtaining fault samples, this paper takes a heavy mechanized bridge of typical engineering equipment as research object, adopts object-oriented modeling and simulation and support vector machine fault diagnosis method. The hydraulic system of its working device is modeled and simulated and the fault diagnosis of hydraulic leakage is carried out, which effectively improves the accuracy of fault diagnosis for small sample events. This paper provides a new intelligent diagnosis method for hydraulic system leakage fault diagnosis. The main research work and results are as follows:. 1) the present situation of hydraulic system leakage fault diagnosis is summarized. The causes of leakage and the mechanism of leakage fault are analyzed in detail. The external leakage and internal leakage are equivalent to orifice outlet flow and slot flow, respectively. Based on the analysis of oil viscosity and gap height, the leakage rate is greatly affected, and the fault diagnosis strategy of hydraulic leakage is put forward. 2) according to the leak fault diagnosis strategy, a hydraulic system working loop model of a heavy-duty mechanized bridge is established by using the hydraulic mechanical system modeling and simulation software AMESim. The leakage faults such as the leakage in the working loop of the thrust cylinder and the leakage in the lock hydraulic cylinder are simulated and analyzed, and the fault characteristics under different leakage conditions are obtained, which provides a theoretical basis for the diagnosis and prevention of the leakage fault. (3) extracting characteristic parameters such as pressure and flow rate from leakage fault simulation, and importing Matlab to obtain time domain eigenvalue, which can be used as data sample for leak fault diagnosis by support vector machine (SVM). By using the LibSVM training model of SVM pattern recognition software package, the location of leakage fault and the cause of leakage are accurately diagnosed. Thus, the feasibility and effectiveness of the method combining leakage fault modeling and simulation with support vector machine fault diagnosis are verified in this paper. 4) from the aspects of controlling the viscosity of hydraulic oil, controlling the height of hydraulic system crevice and the correct assembly of hydraulic pipeline, this paper analyzes and studies how to prevent the leakage of hydraulic system, and puts forward the countermeasures and methods to prevent leakage fault.
【學位授予單位】：國防科學技術大學
【學位級別】：碩士
【學位授予年份】：2012
【分類號】：TH137

【參考文獻】

相關期刊論文 前10條

1 董選明，裘麗華，王占林;故障診斷神經網絡系統(tǒng)的專家知識表達方法[J];北京航空航天大學學報;1996年01期

2 張建華,王占林;基于模糊神經網絡的故障診斷方法的研究[J];北京航空航天大學學報;1997年04期

3 李新德;液壓系統(tǒng)管路漏油原因探討及對策[J];工程機械;2003年06期

4 秦家升 ,游善蘭;AMESim軟件的特征及其應用[J];工程機械;2004年12期

5 劉汝臣;李新德;;淺談液壓系統(tǒng)故障診斷的方法[J];裝備制造技術;2009年07期

6 劉汝臣;李新德;;液壓系統(tǒng)漏油原因分析及控制措施[J];裝備制造技術;2009年08期

7 胡建軍;;工程機械液壓系統(tǒng)內泄漏成因及處理措施分析探討[J];廣東建材;2008年08期

8 趙四軍;王少萍;吳柯銳;;基于粗糙集和支持向量機的航空液壓泵故障診斷[J];中北大學學報(自然科學版);2010年03期

9 周仁斌;謝鐵邦;崔清斌;;自行火炮液壓系統(tǒng)泄漏機理及影響因素研究[J];湖北工業(yè)大學學報;2006年02期

10 彭國朋;黃海濤;;雷達天線車液壓故障分析及其動態(tài)仿真[J];火控雷達技術;2009年02期

相關博士學位論文 前1條

1 賀湘宇;挖掘機液壓系統(tǒng)故障診斷方法研究[D];中南大學;2008年

本文編號：1514600