外文資料翻譯用單極電液伺服閥控制軸向柱塞泵畢業(yè)論文

《外文資料翻譯用單極電液伺服閥控制軸向柱塞泵畢業(yè)論文》由會(huì)員分享，可在線閱讀，更多相關(guān)《外文資料翻譯用單極電液伺服閥控制軸向柱塞泵畢業(yè)論文（47頁珍藏版）》請(qǐng)?jiān)谘b配圖網(wǎng)上搜索。

1、 指導(dǎo)教師評(píng)定成績 (五級(jí)制)： 指導(dǎo)教師簽字： 譯文 用單極電液伺服閥控制軸向柱塞泵 愛荷華州立大學(xué)工程研究學(xué)院工程科學(xué)與機(jī)械學(xué)系（愛荷華州） 50011 A.Akers 穆爾黑立大學(xué)工業(yè)研究部（明尼達(dá)州 穆爾黑德）56560S. J. Lin [摘要]最優(yōu)控制理論應(yīng)用于一個(gè)軸向活塞泵和單級(jí)電液伺服閥組合的壓力調(diào)節(jié)器設(shè)計(jì)。該控制閥已建模，最優(yōu)控制的規(guī)則也已經(jīng)制定。為了開環(huán)和優(yōu)化控制系統(tǒng)，已經(jīng)獲得了流量階躍輸入的時(shí)間響應(yīng)曲線和輸入伺服閥的電流強(qiáng)度。實(shí)驗(yàn)結(jié)果已經(jīng)和那些沒有被作為藍(lán)本的斜盤式制動(dòng)器

2、的供應(yīng)閥門進(jìn)行比較。該伺服閥控制系統(tǒng)的建模意味著系統(tǒng)的響應(yīng)頻率和壓力峰值的極大提高。 [引言]軸向柱塞泵在航空、工業(yè)、農(nóng)業(yè)系統(tǒng)中都很重要。該泵可以傳送大量的特殊能量，還可以改變能量的流量。對(duì)軸向柱塞泵的流量和壓力的控制是通過改變斜盤的角度來實(shí)現(xiàn)的。該斜盤驅(qū)動(dòng)器是由單級(jí)或二級(jí)的電液伺服閥進(jìn)行控制的。單級(jí)伺服閥是由一個(gè)力矩馬達(dá)直接連接一個(gè)四通滑閥而組成的。閥芯閥由力矩電機(jī)定位，由液壓執(zhí)行器指揮控制流向（圖1）。二級(jí)伺服閥有一個(gè)用于倍增力矩電機(jī)輸出的前置放大器，足以克服流體黏附力和由加速度或振動(dòng)產(chǎn)生的力。插板，噴氣管，閥門和閥芯可作為第一級(jí)，而第二級(jí)幾乎是普遍的閥芯的類型。 從歷史上看單級(jí)伺

3、服閥的穩(wěn)定性和反應(yīng)都優(yōu)于那些使用二級(jí)的，但是，自從重量在航天系統(tǒng)中變得特別重要，近期的努力重點(diǎn)放在了完善更輕便的二級(jí)伺服閥。然而，工程的緊密公差要求與其他因素導(dǎo)致成本過高，因此，單級(jí)伺服閥更可能用于工業(yè)應(yīng)用，因?yàn)榫哂懈偁幜Φ膬r(jià)格是必要的。此外，流體動(dòng)力元件設(shè)計(jì)者認(rèn)為產(chǎn)生相對(duì)較大的閥芯力量是很有必要的。一些不可避免的出現(xiàn)在液壓油和有時(shí)出現(xiàn)在氣閥座上的力量（約100牛頓），往往會(huì)切斷金屬或其它芯片，這點(diǎn)在二級(jí)閥閥芯的線軸上是不會(huì)出現(xiàn)的。 好幾項(xiàng)致力于研究和改善軸向柱塞泵動(dòng)態(tài)控制系統(tǒng)的研究已經(jīng)在進(jìn)行當(dāng)中。Harpur [1] and Merritt [2] 使用線性擾動(dòng)分析來研究有微分區(qū)插孔的三通

4、伺服閥和有等面積插孔的四通伺服閥的控制系統(tǒng)。Dreymuller [3] 用勞斯系數(shù)數(shù)組研究軸向柱塞泵的最佳性能。Mack et al. [4] 驗(yàn)證了給變量泵安裝微機(jī)接口用于控制泵的流量和壓力以達(dá)到對(duì)泵的動(dòng)作進(jìn)行補(bǔ)償?shù)目尚行?。最近，Zeiger and Akers [5] 應(yīng)用最優(yōu)控制理論為軸向柱塞泵設(shè)計(jì)了一個(gè)壓力調(diào)節(jié)器。他們的研究結(jié)果表明，直線性最優(yōu)控制方法沒有為流量干擾提供足夠的壓力強(qiáng)度。不過，增強(qiáng)最優(yōu)控制通過它的流量干擾抵消能量，提供了良好的解決辦法。他們的工作并沒有考慮到使用的伺服閥的類型，而是提出了這樣一個(gè)疑問：這樣一個(gè)裝置的頻率和阻尼能夠得到什么樣的代表值。相關(guān)伺服閥的短缺使得人

5、們認(rèn)為對(duì)于設(shè)計(jì)泵時(shí)使用不同伺服閥的影響的全面調(diào)查相當(dāng)重要。本文介紹了單級(jí)伺服閥進(jìn)行的工作，考慮第一伺服閥型。 在這項(xiàng)工作中，推導(dǎo)出了軸向柱塞泵系統(tǒng)的狀態(tài)方程。此外，Zeiger and Akers [5]依靠單級(jí)伺服閥，并且對(duì)壓力時(shí)間曲線進(jìn)行比較從而對(duì)泵的斜盤的驅(qū)動(dòng)建模。 圖1 泵系統(tǒng)的物理模型 一、 動(dòng)力系統(tǒng)模型 到控制執(zhí)行器的流量連續(xù)性忽略了可壓縮性的影響，表示為 （1） 當(dāng)流量連續(xù)性的原則是適用于在泵的排放量控制線，我們得到 （2） 關(guān)于斜盤活塞的任何角位置的瞬時(shí)扭矩在參考文獻(xiàn)（6）中已得出。通過該模型計(jì)算出的力矩的準(zhǔn)確性總在實(shí)驗(yàn)值的10%以。研究

6、結(jié)果還表明，扭矩和壓力之間的關(guān)系因?yàn)楸玫牟煌煌北P傾角和斜盤角速度在實(shí)際圍大致呈線性關(guān)系。這種分析使我們能夠編寫在一個(gè)線性方程形式扭矩，由于斜盤施加的扭矩是由執(zhí)行機(jī)構(gòu)平衡（有彈簧和壓力的力量對(duì)他們進(jìn)行作用）。因此 （3） 該永磁力矩電機(jī)用于移動(dòng)的伺服閥閥芯產(chǎn)生轉(zhuǎn)矩由下式給予 （4） 對(duì)轉(zhuǎn)子運(yùn)用牛頓第二定律，我們得到 （5） 對(duì)作用在閥芯上的流體壓力也同樣進(jìn)行了分析，我們可以把方程（5）寫成 （6） 因?yàn)? 狀態(tài)變量的分配如下： （7） （8） 兩個(gè)控制輸入如下

7、 （9） 方程（2），（3）與（6） - （9）可變成狀態(tài)和輸出方程形式： （10） （11） 該軸向活塞泵控制框圖單級(jí)伺服閥如圖2所示 ，它顯示了功能組件是如何聯(lián)接的，并且確定了每個(gè)組件的響應(yīng)數(shù)學(xué)方程。 二、最優(yōu)控制設(shè)計(jì) 該控制系統(tǒng)的主要目標(biāo)是設(shè)計(jì)一個(gè)控制規(guī)則，使輸出（整個(gè)泵壓力變化的不同）接近于0，這樣可以流量和其他因素變化的時(shí)候保持恒定的壓力。此外，過度的壓力峰值能夠維持在可承受的限度。線性動(dòng)態(tài)系統(tǒng)的最優(yōu)調(diào)節(jié)器問題表現(xiàn)為包括確定一個(gè)矢量控制u（t）的最小化功能 （12） 根據(jù)已知條件 如果是一個(gè)真正的對(duì)稱，半正定矩陣

8、，是一個(gè)真正的對(duì)稱，正定矩陣。 最優(yōu)控制法把價(jià)值函數(shù)降低到最小值，方程（12）可以被轉(zhuǎn)化為 （13） 當(dāng)反饋增益矩陣為 (14) 并且當(dāng)是對(duì)稱的，正面的穩(wěn)態(tài)方程定解 （15） 當(dāng)矩陣和在方程（12）中所代表的相對(duì)比較重要的值都被精確確定，把方程（13）中的最優(yōu)控制規(guī)則作為條件借出方程（15），這些值不斷的傳送給數(shù)據(jù)處理器，產(chǎn)生反饋信號(hào)對(duì)伺服閥進(jìn)行控制。 最佳閉環(huán)系統(tǒng)，可得到如下 (18) 特征值是矩陣和閉環(huán)系統(tǒng)的極點(diǎn)。該比重可能會(huì)取得理想的瞬時(shí)變化的反應(yīng)[7]。 圖2 單級(jí)伺服閥軸向柱塞泵的結(jié)構(gòu)框圖 三、結(jié)果 開環(huán)系統(tǒng)的

9、初始數(shù)據(jù)，我們可以單獨(dú)的使用一個(gè)干擾步驟輸入包括。下游量的數(shù)值可以假設(shè)為0.5， 1.0和2，壓力波動(dòng)，斜盤角，斜盤角速度就可以求解。正如預(yù)期的達(dá)到峰值壓力干擾的時(shí)間隨著V的增加而增加，隨著初始速率的變化，壓力始終保持不變。當(dāng)0.01/s用作一個(gè)輸入步驟時(shí)，穩(wěn)定的斜盤角度被看作與下游量成正比。而且當(dāng)下游量增長的時(shí)候，斜盤也會(huì)有一個(gè)超過起穩(wěn)態(tài)位置的增長趨勢(shì)。當(dāng)使用干擾電流的時(shí)候，初速度的變化是與下游量無關(guān)的。 四、最優(yōu)控制系統(tǒng) 對(duì)最優(yōu)控制器的性能已經(jīng)進(jìn)行了研究，它的計(jì)算結(jié)果和參考文獻(xiàn)[5]所展現(xiàn)的一樣，缺乏對(duì)現(xiàn)存控制系統(tǒng)進(jìn)行改善的穩(wěn)健性。因此，它不會(huì)被完全的用作為一個(gè)控制器。為本文所使用

10、的配置響應(yīng)計(jì)算結(jié)果再次證實(shí)，此方面還需要增強(qiáng)。這個(gè)增強(qiáng)只須將最佳比例控制器轉(zhuǎn)換成比例加積分控制器。這種將有助于增強(qiáng)響應(yīng)時(shí)間，超調(diào)量，并減少穩(wěn)態(tài)誤差。本文中使用的這種控制器，如圖（3）所示，可以看出，使用增量積分使這個(gè)增強(qiáng)達(dá)到了一個(gè)相當(dāng)完美的程度。這個(gè)過程導(dǎo)致了一個(gè)六指令系統(tǒng)。 該方法用于分析選擇矩陣Q和由產(chǎn)生的不同控制規(guī)則組成的標(biāo)量、工程量R的影響，然后計(jì)算性能指標(biāo)和策繪了從方程解最優(yōu)控制泵的根位置。為Q矩陣和R選擇合適的值以達(dá)到減少穩(wěn)定時(shí)間、超調(diào)量和穩(wěn)態(tài)錯(cuò)誤的系統(tǒng)響應(yīng)的結(jié)果。挑選的值如下： 用來對(duì)干擾進(jìn)行響應(yīng)的數(shù)據(jù)是與一個(gè)系列22泵和一個(gè)典型的單級(jí)電液伺服閥 聯(lián)系的。這些數(shù)據(jù)在

11、表一中給出，其中為流動(dòng)性能的穩(wěn)定值也一并給出。 圖3 加強(qiáng)的最優(yōu)控制 五、響應(yīng)圖 對(duì)于一個(gè)階躍的響應(yīng)已經(jīng)表示在圖4至7里面。該系統(tǒng)最佳下游量再次假定為。對(duì)于有下游量和的次優(yōu)的系統(tǒng)，同樣進(jìn)行了研究。壓力反應(yīng)圖4所示的是V的較小速度和較小峰值。此外，在參考文獻(xiàn)[5]中可以知道，當(dāng)泵在最優(yōu)化控制與正確的單級(jí)閥模型條件下時(shí)，響應(yīng)頻率大約是平時(shí)的三倍，而峰值的壓力也同樣因?yàn)檫@個(gè)三倍的變化而減少了。 當(dāng)循環(huán)是封閉的時(shí)候，響應(yīng)是相當(dāng)?shù)目?。從圖6和7可以觀察到V值對(duì)斜盤角速度、閥芯移位和峰值速度有非常重要的影響。此外，當(dāng)下游量減半或變成最優(yōu)值的兩倍時(shí)，頻率響應(yīng)也有±20%的變化。 圖8給出了

12、將泵的轉(zhuǎn)速從210rad/s減少到126 rad/s對(duì)響應(yīng)的影響。通過對(duì)Akers and Zeiger [5]的工作的再次直接比較，可以看出壓力峰值的過沖大幅減 圖5 斜盤角度的時(shí)間響應(yīng)（最優(yōu)系統(tǒng)）， 圖4 優(yōu)化系統(tǒng)的壓力——時(shí)間響應(yīng)曲線， 圖6 斜盤角速度的時(shí)間響應(yīng)（最優(yōu)系統(tǒng)）， 圖7 閥芯位移的時(shí)間響應(yīng)（最優(yōu)系統(tǒng)）， 少而頻率卻增加了。參考文獻(xiàn)[5]中提與，該曲線也顯示出這樣一個(gè)一樣的趨勢(shì)：當(dāng)轉(zhuǎn)速提高時(shí)，較小峰值和頻率同樣也會(huì)增大。 六、結(jié)論 用一個(gè)單級(jí)伺服閥以驅(qū)動(dòng)斜盤并且控制泵壓來建立一個(gè)商用的軸向柱塞泵模型是可行的。 開環(huán)系統(tǒng)已經(jīng)開始研究，并且最優(yōu)控制準(zhǔn)則也已

13、經(jīng)在那時(shí)制定。最優(yōu)閉環(huán)系統(tǒng)的時(shí)間響應(yīng)曲線已經(jīng)提出。 閉環(huán)系統(tǒng)、優(yōu)化系統(tǒng)時(shí)間響應(yīng)之間的比較同時(shí)在開環(huán)系統(tǒng)和一個(gè)用工程實(shí)踐[5]假定頻率值和斜盤控制執(zhí)行機(jī)構(gòu)的阻尼值的系統(tǒng)中進(jìn)行。在每一個(gè)比較當(dāng)中，單級(jí)閥的生產(chǎn)納入了顯著低峰值壓力、更高響應(yīng)頻率的指標(biāo)以改進(jìn)其性能。單級(jí)伺服閥性能提升的重要原因是因?yàn)橛袦?zhǔn)確的建模以進(jìn)行狀態(tài)變量分析。 圖8 最優(yōu)系統(tǒng)下，泵不同轉(zhuǎn)速的壓力時(shí)間響應(yīng)，v=12， 表一 使用的數(shù)據(jù)系列22軸向柱塞泵和一個(gè)典型的單級(jí)電液伺服閥： = 1000 MPa (油壓縮) （泄露系數(shù)） （量排放系統(tǒng)） （軸轉(zhuǎn)動(dòng)） （具體的體積位移） (斜盤慣性) （斜盤力矩系數(shù)

14、） （執(zhí)行器彈簧剛度） （活塞球半徑） （斜盤力矩系數(shù)） （斜盤力矩系數(shù)） （流量壓力系數(shù)） （位移執(zhí)行器的控制容積） （活塞領(lǐng)域） （流量增益） （慣性電樞） （電樞支點(diǎn)的距離） （閥芯質(zhì)量） (粘性阻尼系數(shù)) (閥芯阻尼系數(shù)) （力矩電機(jī)彎矩） (流通領(lǐng)域梯度) (恒轉(zhuǎn)矩電機(jī)) (供應(yīng)壓力) 七、答 感愛荷華州Sundstrand HydroTransmission公司提供泵的幾何數(shù)據(jù)和一種單級(jí)伺服閥的有關(guān)數(shù)據(jù)。此項(xiàng)成果由工程科學(xué)和機(jī)械系支持，愛荷華州立大學(xué)工程研究所提供資金和編輯幫助。對(duì)于他們的支持非常感。 參考文獻(xiàn)： [1]

15、Harpur, N. F., "Some Design Considerations ofHydraulic Servos of the Jack Type," Proc. Conf.Hydraulic Servomechanism, Vol. 41, I Mech E(1953). [2]Merritt, H. E., Hydraulic Control Systems, JohnWiley & Sons lnc. (1967). [3]Dreymuller, J., "Pilot-operated and DirectlyActuated Pressure Control with V

16、ariableDelivery Axial Piston Pumps," Proc. 4th International Fluid Power Symposium, pp. B1-1 toBI-B20 (1975). [4]Mack, P., et al., "Microcomputer Control of aVariable Displacement Pump," Proc. 40th Nat,Conf. on Fluid Power, Vol. XXXVIII, 55-61(1984). [5]Zeiger, G., and Akers, A., "Optimal Control

17、ofan Axial Piston Pump," Proc. 7th InternationalFluid Power Symposium Paper No. 7, 57-64(1986). [6]Zeiger, G., and Akers, A., "Torque on theSwashplate of an Axial Piston Pump," ASMEJournal of Dynamic Systems Measurement andControl, Vol. 107, No. 3, 220-226 (1985). [7]D'Azzo, J. J., and Houpis, C.

18、H. Linear Control System Analysis and Design, 2nd editionMcGraw-Hill Book Company (1981). CONTROLOF AN AXIAL PISTON PUMP USING ASINGLE-STAGE ELECTROHYDRAULIC SERVOVALVE A. AkersEngineering Research Institute andDepartment of Engineering Science and Mechanics IowaStateUnivers

19、ityAmes, Iowa50011 S.J.LinDepartment of IndustrialStudiesMoorheadStateUniversityMoorhead, Minnesota56560 ABSTRACT Optimal control theory is applied to the design of a pressure regulator for an axial pistonpump and single-stage electrohydraulic valve combination. The control valve has been modeled

20、 andanoptimal control law has been formulated. Thetime response curves due to a step input in flowrate and in current input to the servovalve havebeen obtained for the open loop and for the optimalcontrol system. The results have been compared tothose in which the supply valve to theswashplateactuat

21、ors was not modeled. Controlled systemmodeling of the servovalve significantly improvesthe system's response frequency and pressure peaks. INTRODUCTION Axial piston pumps are important in aircraft,industrial, and agricultural systems; they cantransmit large specific powers and their flow ratecan

22、 be varied. The control of flow or pressure ofaxial piston pumps is achieved by changing theswashplate angle. The swashplate actuator is controlled by an electrohydraulic servovalve, whichmay be either single-stage or two-stage. Single-stage servovalves consist of a torque motordirectly attached to

23、a four-way spool valve. Thespool valve, positioned by the torque motor,directs controlled flow to the hydraulic actuator(Fig. 1).Two-stage servovalves have a hydraulicpreamplifier that multiplies the force output ofthe torque motor sufficiently to overcome flow andstiction forces and forces resultin

24、g from acceleration or vibration. Flapper, jet pipe, and spoolvalves may be used as a first-stage, while thesecond-stage is almost universally of the spooltype. Historically, the stability and response ofthe single-stage servovalves have been superior tothose using two stages, but since weight is p

25、aramount in aerospace systems, recent effort hasfocused on perfecting the lighter, two-stage servovalve. However, the close engineering tolerancesrequired and other factors have led to high costs;thus, it is more likely that single-stage valveswill be used for industrial applications wherecompetitiv

26、e pricing is essential. In addition,designers of fluid power components see a need togenerate relatively large spool forces. Suchforces (approximately 100N), not achievable in thespools of two-stage valves, are required to severmetal or other chips, which inevitably are presentin hydraulic oil and w

27、hich sometimes lodge at thevalve seats. Several studies have been conducted in anattempt to investigate and improve the dynamic control systems of axial piston pumps. Harpur [1] andMerritt [2] used linearized perturbation analysisto investigate a three-way servovalve with differential area jack and

28、 a four-way servovalve withequal-area jack control systems. Dreymuller [3]analyzed optimal performance of axial piston pumpsby use of a Routh coefficients array. Mack et al.[4] examined the feasibility of interfacing amicrocomputer to a variable displacement pump tocontrol flow rate and provide pres

29、sure compensationfor the pump action. More recently, Zeiger andAkers [5] applied optimal control theories to thedesign of a pressure regulator for an axial pistonpump. Their results showed that a straight-linear,optimal control method did not provide adequatepressure stiffness to flow disturbances.

30、However,the augmented optimal control provides a goodsolution to the pump regulator problem because ofits capability to offset flow disturbances. Theirwork did not take into account the type of servovalve to be used; an assumption was made as to whatrepresentative values could be obtained for freque

31、ncy and damping of such a device. The lack ofan associated servovalve was considered sufficiently serious to warrant a full investigationinto the effects of using different servovalvedesigns with the pump. This paper describes thework conducted with a single-stage servovalve, thefirst servovalve typ

32、e considered. In this work, the state equations were derivedfor the axial piston pump system. In addition,actuation of the swashplate of the pump by means ofa single-stage servovalve was modeled, and a comparison was made between the pressure time-responsecurves obtained and those obtained by Zeige

33、r andAkers [5]. DYNAMICAL MODEL OF THE SYSTEM [2] The flow continuity into the control actuator neglecting the effect of compressibility, is expressed as (1) When the principle of flow continuity isapplied to the control volume in the discharge lineof the pump, we have （2） The instanta

34、neous torque on the swashplate atany angular position of the pistons has been obtained in Ref. [6]. The accuracy of the computedtorques from the model is shown to be everywherewithin 10% of experimental values. The resultsalso indicate that relationships between torque andpressure differential acro

35、ss the pump, swashplateangle, and angular velocity of the swashplate areapproximately linear over the practical range.That analysis permits us to write the torque equation in a linearized form, since the torque exertedon the swashplate is balanced by the actuators(which have spring and pressure forc

36、es acting onthem). Then （3） The permanent-magnet torque motor used to movethe spool of the servovalve produces a torque givenby （4） Applying Newton's second law to the armature,we obtain （5） The stroking flow forces acting on the valvespool have also been analyzed, and we may w

37、riteEq. (5) as （6） Where The assigned state variables are as follows: （7） （8） and two controlled inputs are （9） Equations (2), (3) and (6)-(9) can be put intothe state and output equation form: （10） （11） The block diagram of the axial piston pumpcontrolle

38、d by a single-stage servovalve is shown inFig. 2, which shows how the functional componentsare connected and the mathematical equations thatdetermine the response of each component. OPTIMAL CONTROL FORMULATION The main objective of the control system is todesign a control law that will bring the o

39、utput (the variation of pressure differential across thepump) close to zero, so that constant pressure maybe maintained while flow rate and other variableschange. In addition, over-pressure peaks may bekept within acceptable limits.The optimal regulator problem for linear dynamic systems shown below

40、 consists of determining a vector control u(t)that minimizes the functional （12） Subject to the restrictions where is a real symmetric, positive semi-definitematrix, and is a real symmetric, positivedefinite matrix. The optimal control law that minimized thecost function, Eq. (12), ca

41、n be specified as （13） where the feedback gain matrix is (14) and where is the symmetric, positive definitesolution of the steady-state Riccati equation （15） Once the matrices and that representassessment of the relative importance of the various terms in Eq. (12) have been specifi

42、ed, thesolution of Eq. (15) specifies the optimal controlarelaw in Eq. (13); these values continuouslyto the value of theprocessed and fed back generatecontrol current into the servovalve. The optimal closed-loop system can be obtainedas (18) The eigenvalues of are poles ofmatrices and clo

43、sed loop system.The weightingmay be to obtain desirable transientchangedresponse [7]. RESULTS Open Loop SystemFor the first set of data we use a disturbances comprising, separately, a step input of.Values of downstreamvolume of 0.5, 1.0, and 2were assumed, and theresulting pressure fluctuations,

44、 swashplate angle,and swashplate angular velocity were evaluated.Asexpected, the time to achieve peak pressure disturbance is increased with increasing V, with theinitial rate of change of pressure being approximately constant.The steady swashplate angle isseen to be proportional to the downstream v

45、olumewhen a step input of 0.01/s is used, and as downstream volume increases there is an increasing tendency for the swashplate to overshoot its steadystate position. When the disturbance current isapplied, the initial rate of change of is independent of downstream volume. Optimally Controlled Sys

46、tem The performance of the nonaugmented optimalcontroller has been investigated.Its computedresults lack robustness for improving the existingcontrol system as shown in Ref. [5]. It thereforewould not perform adequately as a controller. Thecomputed results for response for the configurationused in

47、this paper confirmed that once more augmentation was required.The augmentation simply converts the optimal proportional controller into aproportional plus integral controller. Such anaugmentation will aid in response time, overshoot,and reducing steady state error. The controllerused in this paper,

48、as shown in Fig. 3, shows whereaugmentation has been achieved by using the integral of the increment of .This procedure givesrise to a sixth-order system.. The method used for the analysis on the effectof the selection for the Q matrix and scalar quantities R consisted of generating different con

49、trollaws and then computing the performance index andplotting the root location of the optimally controlled pump from the solution of the Riccati equation. The appropriate values selected for the Qmatrix and R result in reduced settling time, overshoot and steady state errors of the systemresponses.

50、 Selected values appear below. The data used for the response to disturbanceswere those associated with a Series 22 pump and atypical single-stage electrohydraulic servo-valveand are given in Table 1, where the steady valuesfor flow properties are also given. Response Diagrams The respons

51、es to a step input of are shown in Figs. 4 through 7. Thedownstream volume for the optimal system was againassumed to be . Suboptimal systems havingdownstream volumes of and were also investigated. The pressure response shown in Fig. 4 ismore are smallerrapid and the peaks less forvalues of V.In a

52、ddition, it can be seen that when the pump control is optimized with the correctmodel of the single-stage valve, then the responsefrequency is roughly three times that shown inRef. [5]; the peak pressures are also reduced by afactor of three. —————————————————————————————————— Table 1. Data used

53、 for the Series 22 axial pistonpump and a typical single-stage electrohydraulicservovalve. = 1000 MPa (oil compressibility) —————————————————————————————————— The response is considerably faster when theloop is closed. From Figs. 6 and 7 it is obser

54、vedthat the value of V has a significant effect uponthe angular velocity of the swashplate, the spooldisplacement, and velocity peak values; in addition, the response frequencies are changed by ±20%when the downstream volume is halved or doubledfrom the optimal value. On responseFigure 8 gives the

55、effect on the response ofreducing the pump rotational speed from 210 to126 rad/s. Once more a direct comparison with thework of Akers and Zeiger [5] illustrates that thepressure-peak overshoots are much reduced and thefrequency is increased. The curves also show atendency identical to that in Ref. [

56、5] whereincrease in rotational speed gives rise to asmaller peak response and a higher frequency. CONCLUSIONS It has been possible to model a commerciallyavailable, axial piston pump by using a single-stageservovalve to drive the swashplate and socontrol the pump pressure. The open-loop syste

57、m was investigated and anoptimal control law was then formulated. Timeresponse curves have been presented for the closedloop optimal system. Comparison between the closed-loop, optimal-system time responses have been made with both theopen loop system and with a system where values offrequency and

58、damping for the swashplate actuatorcontrol were assumed by using engineering practice[5]. In each comparison, incorporation of thesingle-stage valve improves the performance by producing significantly smaller peak pressures andhigher frequencies. The principal reason for performance improvement is t

59、he fact that the single-stage servovalve has been correctly modeled forinclusion in the state variable analysis. ACKNOWLEDGMENTS Gratitude is expressed to Sundstrand HydroTransmission Company in Ames, Iowa, for furnishingpump geometrical data and relevant data for atypical single-stage servovalve.

60、 The work wassupported by the Department of Engineering Scienceand Mechanics, and financial and editorial help wasprovided by the Engineering Research Institute ofIowa State University. This support is gratefullyacknowledged. REFERENCES 1. Harpur, N. F., "Some Design Considerations ofHydraulic Ser

61、vos of the Jack Type," Proc. Conf.Hydraulic Servomechanism, Vol. 41, I Mech E(1953). 2. Merritt, H. E., Hydraulic Control Systems, JohnWiley & Sons lnc. (1967). 3. Dreymuller, J., "Pilot-operated and DirectlyActuated Pressure Control with VariableDelivery Axial Piston Pumps," Proc. 4th Internation

62、al Fluid Power Symposium, pp. B1-1 toBI-B20 (1975). 4. Mack, P., et al., "Microcomputer Control of aVariable Displacement Pump," Proc. 40th Nat,Conf. on Fluid Power, Vol. XXXVIII, 55-61(1984). 5. Zeiger, G., and Akers, A., "Optimal Control ofan Axial Piston Pump," Proc. 7th InternationalFluid Powe

63、r Symposium Paper No. 7, 57-64(1986). 6. Zeiger, G., and Akers, A., "Torque on theSwashplate of an Axial Piston Pump," ASMEJournal of Dynamic Systems Measurement andControl, Vol. 107, No. 3, 220-226 (1985). 7. D'Azzo, J. J., and Houpis, C. H. Linear Control System Analysis and Design, 2nd editionM

64、cGraw-Hill Book Company (1981). 指導(dǎo)教師評(píng)定成績 (五級(jí)制)： 指導(dǎo)教師簽字： 附件C：譯文 壓電瓷活塞驅(qū)動(dòng)液壓泵的發(fā)展 G. W. Woodruff 學(xué)校機(jī)械工程學(xué)院，喬治亞理工學(xué)院（喬治亞，亞特蘭大）30332-0405William S. Oates, Lisa D. buck 和 Christopher S. Lynch [摘要]壓電材料在高頻率和大型電場(chǎng)工作時(shí)可以產(chǎn)生非常高的功率密度。在馬達(dá)中利用次功率密度只會(huì)由壓電材料產(chǎn)生很小的行程。行程的整改是為了得到

65、一個(gè)高功率密度的設(shè)備。這一點(diǎn)已經(jīng)通過用一個(gè)活塞驅(qū)動(dòng)液壓泵來實(shí)現(xiàn)了。更改流體動(dòng)力也已通過使用止回閥實(shí)現(xiàn)。泵的設(shè)計(jì)和特性詳情已經(jīng)得到。壓電堆棧的熱力循環(huán)在壓電泵中的應(yīng)用也已闡明。 [引言]許多應(yīng)用智能材料需要龐大的力量和大的位移。單行程器可以在一個(gè)很小的距離產(chǎn)生一個(gè)很大的力。超聲波馬達(dá)利用共振波來趨勢(shì)軸旋轉(zhuǎn)。其結(jié)果是固態(tài)馬達(dá)做得非常小，但是，只具有很低的扭矩輸出。分布重復(fù)設(shè)備依靠壓電直接驅(qū)動(dòng)能力工作時(shí)低于共振頻率。這兩個(gè)設(shè)備分別是直線電機(jī)和液壓泵。本文對(duì)液壓泵的發(fā)展和性能進(jìn)行了詳細(xì)的討論。 一、分步重復(fù)執(zhí)行器 分步重復(fù)執(zhí)行器利用壓電的高頻率性能。壓電致動(dòng)器在每個(gè)過程中只能做數(shù)量有限的

66、工作。用與執(zhí)行器阻抗相匹配的負(fù)載使工作量達(dá)到最大化。在單位時(shí)間的能量就是工作量。而這個(gè)能量可以通過提高頻率來增加。 該系統(tǒng)利用一個(gè)堆棧執(zhí)行器來控制液壓泵里的活塞。系統(tǒng)以提高到頻率的潛力為重點(diǎn)來彌補(bǔ)堆棧的小排量問題。通過在泵中增加一個(gè)入口和出口單向閥來糾正流體流動(dòng)以得到直線運(yùn)動(dòng)的液壓執(zhí)行器。此外，為了實(shí)現(xiàn)泵的進(jìn)程，利用了機(jī)械載荷和液壓壓力來保證堆棧的情況受到壓縮。在現(xiàn)行系統(tǒng)中，9.5mm（3/8”）口徑的驅(qū)動(dòng)器可以到達(dá)280 N (62 Ibs.)的阻力和7.25 cm/sec (2.86 in/s)的自由變速。 二、壓電液壓工作周期 Mauk, Menchaca, and Lynch [2]的成果闡明了壓電液壓泵設(shè)計(jì)的效率。該壓電液壓泵的效率將會(huì)通過一個(gè)完整的熱力學(xué)工作周期來說明。 理論上，液壓泵的壓電量輸出應(yīng)該是匹配彈簧負(fù)載壓電阻抗的兩倍。在經(jīng)典的阻抗匹配中，一個(gè)彈簧原件代表這個(gè)負(fù)載會(huì)被壓電瓷的直接驅(qū)動(dòng)所代替。如圖1所示，動(dòng)作是力的作用和實(shí)際位移的結(jié)果。最大壓電負(fù)載的阻力可以得到()。自由位移和零載荷是的最大應(yīng)變()是緊密關(guān)聯(lián)的。實(shí)際負(fù)荷將存在介于這兩個(gè)極端。 圖1 壓