壓縮包內(nèi)含有CAD圖紙和說明書,均可直接下載獲得文件，所見所得，電腦查看更方便。Q 197216396 或 11970985

附錄A 譯文 釆煤機(jī)自適應(yīng)記憶切割 摘要：針對(duì)以往的采煤機(jī)記憶切割技術(shù)在我國(guó)復(fù)雜地質(zhì)條件下不適用的情況，提出了基于模糊控制理論的采煤機(jī)自適應(yīng)記憶切割技術(shù)，設(shè)計(jì)出了采煤機(jī)位置、姿態(tài)定位系統(tǒng)和采煤機(jī)自適應(yīng)切割模糊控制系統(tǒng)。該系統(tǒng)可以獲取采煤機(jī)任意位置處的姿態(tài)和狀態(tài)信息，自動(dòng)跟蹤所記憶的切割路徑，基于模糊控制理論對(duì)是否截割到巖石進(jìn)行判斷并做出最優(yōu)的處理方案。通過實(shí)驗(yàn)室的路徑跟蹤實(shí)驗(yàn)和西安煤礦機(jī)械有限公司的自適應(yīng)調(diào)節(jié)實(shí)驗(yàn)證明：該技術(shù)在實(shí)現(xiàn)采煤機(jī)記憶切割的基礎(chǔ)上能夠識(shí)別出滾筒到切割巖石時(shí)的異常狀態(tài)，并對(duì)牽引速度和滾筒高度做自適應(yīng)調(diào)節(jié)，能夠滿足復(fù)雜地質(zhì)條件下對(duì)采煤機(jī)的控制要求。 關(guān)鍵詞：采煤機(jī)；自適應(yīng)控制；記憶切割；模糊控制 中圖分類號(hào)：TD421. 6；TP273. 4 文獻(xiàn)標(biāo)志碼：A 采煤機(jī)的自動(dòng)化控制是實(shí)現(xiàn)采煤工作面自動(dòng)化遙控器在近距離操控，大量的煤塵和水霧使得操作的重點(diǎn)和難點(diǎn)，而采煤機(jī)自動(dòng)化的關(guān)鍵則是滾筒的人員很難看清滾筒是否切割到巖石，僅能通過滾筒自動(dòng)調(diào)高。目前國(guó)內(nèi)采煤機(jī)大多依靠人工通過紅外發(fā)出的聲音進(jìn)行判別。為此，國(guó)內(nèi)外學(xué)者提出過利用y射線探測(cè)、雷達(dá)探測(cè)、紅外溫度探測(cè)、截割力分析、振動(dòng)分析等方法來進(jìn)行煤巖識(shí)別[1-5]，但實(shí)際應(yīng) 用的效果都不理想。20世紀(jì)80年代中期西德學(xué)者 首次提出了記憶割煤行程自動(dòng)調(diào)高系統(tǒng)，并成功應(yīng) 用于美國(guó)JOY公司的7LS6型、德國(guó)Eickhoff公司的SL500型和DBT公司的EL3000型等采煤機(jī)上。該方法避免了煤巖識(shí)別這一技術(shù)難題，便于實(shí)現(xiàn)且操作簡(jiǎn)單，但在我國(guó)采煤工作面上的應(yīng)用并不理想，主要原因在于：1) 我國(guó)煤礦地質(zhì)條件復(fù)雜，煤巖界面變化劇烈，所記憶的切割曲線不具有普遍性，當(dāng)記憶路徑失效時(shí)，就需要對(duì)切割路徑重新進(jìn)行記憶。2) 隨著我國(guó)采煤機(jī)功率的不斷增大，同時(shí)為了移架、推溜的順利進(jìn)行和采煤效率的提高，較小的夾矸和斷 層均直接進(jìn)行切割而不必頻繁的調(diào)節(jié)滾筒高度且頂 板和底板應(yīng)盡量切割平整。因此，有必要針對(duì)我國(guó)復(fù)雜地質(zhì)條件和當(dāng)前煤礦生產(chǎn)需求對(duì)采煤機(jī)記憶切割技術(shù)進(jìn)行研究和改進(jìn)。筆者提出了基于模糊理論 的采煤機(jī)自適應(yīng)記憶切割技術(shù)，既遵從記憶切割路徑以避免切割過硬巖石而導(dǎo)致的機(jī)械或電氣部件損壞，又盡可能少的調(diào)節(jié)滾筒高度從而保持頂板和底板的平整性。 1采煤機(jī)自適應(yīng)記憶切割總體控制方案 采煤機(jī)工作過程中最主要的兩個(gè)動(dòng)作是機(jī)身的橫向往復(fù)運(yùn)動(dòng)和搖臂的縱向升降運(yùn)動(dòng)，前者對(duì)應(yīng)于牽引電機(jī)的轉(zhuǎn)速；后者對(duì)應(yīng)于調(diào)高油缸的伸縮量。因此對(duì)采煤機(jī)的控制也主要針對(duì)與牽引電機(jī)和調(diào)高油缸。傳統(tǒng)的記憶切割技術(shù)要求采煤機(jī)切割到巖石后搖臂立即下降，以避免截割電機(jī)堵轉(zhuǎn)或截割齒斷裂。而隨著電機(jī)功率的增大和截割齒材料的改進(jìn)，使得采煤機(jī)可以對(duì)一般硬度的巖石直接進(jìn)行切割。 因此在本方案中，控制器根據(jù)采煤機(jī)的各項(xiàng)傳感數(shù)據(jù)不僅可以識(shí)別出是否切割到巖石，而且還能夠判斷出是否允許對(duì)其直接切割。對(duì)于一般硬度的巖石采取降低牽引速度強(qiáng)行切割的方法，如果巖石硬度過大則采取調(diào)節(jié)滾筒高度進(jìn)行避讓的方法。該控制方案在采煤機(jī)機(jī)械部件和電器部件不受損傷的基礎(chǔ) 上，最大限度地保證了采煤效率。 采煤機(jī)自適應(yīng)記憶切割總體控制流程如圖1所示，可分為3個(gè)階段：人工示教階段、自適應(yīng)切割階段和人工修正階段。每個(gè)階段既相對(duì)獨(dú)立又相互聯(lián)系，如自適應(yīng)切割與人工示教和人工修正都發(fā)生聯(lián)系；人工示教和人工修正共享一個(gè)數(shù)據(jù)記憶集來記錄人工操作。 圖1采煤機(jī)自適應(yīng)記憶切割總體控制流程圖 人工示教由人工操控和數(shù)據(jù)記憶兩部分組成。當(dāng)操作人員控制采煤機(jī)進(jìn)行切割時(shí)，機(jī)載控制器每隔一段時(shí)間便會(huì)記錄下采煤機(jī)當(dāng)前的位置、姿態(tài)、狀態(tài)和動(dòng)作等信息。其中，位置信息是指采煤機(jī)在工作面處的空間坐標(biāo)；姿態(tài)信息是指采煤機(jī)的機(jī)身傾斜角度和滾筒的空間坐標(biāo)；狀態(tài)信息是指采煤機(jī)機(jī)械部件和電器部件運(yùn)行的狀態(tài)參數(shù)；動(dòng)作信息是指操作人員對(duì)采煤機(jī)發(fā)出的控制命令。這些數(shù)據(jù)經(jīng)過處理后存儲(chǔ)于控制器中以指導(dǎo)采煤機(jī)的自動(dòng)運(yùn)行。顯然，機(jī)載控制器的采集頻率越大就越詳細(xì)的記錄下采煤機(jī)運(yùn)行過程，但同時(shí)又會(huì)產(chǎn)生大量無用信息占用控制器的處理能力和存儲(chǔ)空間；而如果頻率過小，就有可能漏掉采煤機(jī)的一些重要?jiǎng)幼?。針?duì)這一問題，本方案將控制器的采集點(diǎn)分為常規(guī)點(diǎn)和關(guān) 鍵點(diǎn)。常規(guī)點(diǎn)沿采煤機(jī)運(yùn)行方向等距離分布，其間相隔1 m。關(guān)鍵點(diǎn)則是操作人員對(duì)采煤機(jī)發(fā)出控制命令的點(diǎn)，如采煤機(jī)的啟動(dòng)、停止、加速、減速，搖臂的上升、下降等。關(guān)鍵點(diǎn)是人工示教的核心，直接反映了操作人員的操作方式和操作順序。采取常規(guī)點(diǎn)和關(guān)鍵點(diǎn)相結(jié)合進(jìn)行記憶的策略，既確保了記憶質(zhì)量又降低了數(shù)據(jù)量，為后續(xù)的自適應(yīng)切割過程提供了保障。 自適應(yīng)切割是指由機(jī)載控制器控制采煤機(jī)按照人工示教的路徑自動(dòng)切割煤層。首先，機(jī)載控制器根據(jù)人工示教過程中所記憶的操作命令控制采煤機(jī)運(yùn)行，在運(yùn)行過程中機(jī)載控制器在每個(gè)常規(guī)點(diǎn)和關(guān)鍵點(diǎn)處將當(dāng)前采集到的數(shù)據(jù)與所記憶的數(shù)據(jù)進(jìn)行對(duì)比。而后，模糊控制器根據(jù)對(duì)比結(jié)果對(duì)當(dāng)前的運(yùn)行狀態(tài)進(jìn)行判別是否切割到巖石；是否需要停機(jī)；是否需要加速或減速；是否需要上升或下降搖臂；是否需要人工干預(yù)；是否需要重新人工示教等。最后，機(jī)載控制器根據(jù)模糊控制器的判別做出相應(yīng)的控制輸出，對(duì)于常規(guī)操作如加減速、升降搖臂等，可由控制器自行完成；而對(duì)于人工干預(yù)、人工示教操作，則需要向操作人員發(fā)出報(bào)警提示，請(qǐng)求人工介入。 人工修正是指當(dāng)機(jī)載控制器遇到無法解決的故障或者無法識(shí)別的狀態(tài)時(shí)，將控制方式從自動(dòng)轉(zhuǎn)換為手動(dòng)，由操作人員控制采煤機(jī)進(jìn)行切割。人工修正是自適應(yīng)切割的有力補(bǔ)充，確保了在緊急情況下人工操作的優(yōu)先權(quán)。機(jī)載控制器會(huì)記錄下人工修正的操作步驟，作為關(guān)鍵點(diǎn)進(jìn)行存儲(chǔ)，下次遇到類似狀況便可以自行解決。當(dāng)人工修正完成后，操作人員可以將采煤機(jī)的控制方式從手動(dòng)改為自動(dòng)，由機(jī)載控制器根據(jù)記憶數(shù)據(jù)控制采煤機(jī)完成后續(xù)的切割任務(wù)。 2采煤機(jī)的位置和姿態(tài)定位 由上文可知，采煤機(jī)位置和姿態(tài)是自適應(yīng)記憶切割中的重要信息，直接影響到采煤機(jī)的控制效果。國(guó)內(nèi)以往對(duì)于采煤機(jī)記憶切割的研究并不深入，對(duì)于采煤機(jī)位置和姿態(tài)的空間定位問題尚未有完整的 解決方案。曾有學(xué)者提出利用軸編碼器計(jì)算采煤機(jī)行走距離來進(jìn)行位置定位，利用位移傳感器獲取調(diào) 高油缸的伸縮量來進(jìn)行姿態(tài)定位，但得到僅是位置 和姿態(tài)的相對(duì)值而并非三維空間內(nèi)的絕對(duì)值[6-8]。 因此，解決采煤機(jī)位置和姿態(tài)定位問題是實(shí)現(xiàn)自適應(yīng)切割的前提和基礎(chǔ)。 2.1 采煤機(jī)位置定位 要確定采煤機(jī)機(jī)身的位置最直觀的方法就是獲取其在三維空間內(nèi)的坐標(biāo)值，這就需要解決以下問題，選取采煤機(jī)上某一固定點(diǎn)作為位置定位的特征點(diǎn)；定義三維坐標(biāo)系；推導(dǎo)特征點(diǎn)坐標(biāo)值的計(jì)算公式。 1) 特征點(diǎn)的選取。采煤機(jī)的機(jī)體過于龐大，因此需要在采煤機(jī)上選取一個(gè)特征點(diǎn)，并以此特征點(diǎn)的三維坐標(biāo)來唯一確定采煤機(jī)的位置。本文選取采煤機(jī)行走齒輪與刮板運(yùn)輸機(jī)上導(dǎo)軌的接觸點(diǎn)作為特征點(diǎn)進(jìn)行定位計(jì)算，此時(shí)特征點(diǎn)的運(yùn)行軌跡與刮板運(yùn)輸機(jī)的導(dǎo)軌重合。 2) 三維坐標(biāo)系的定義。初始狀態(tài)下采煤機(jī)起始位置處的特征點(diǎn)為系統(tǒng)原點(diǎn)O。重力加速度反方向?yàn)閥軸正方向，重力加速度方向?yàn)閥軸負(fù)方向。平行于刮板運(yùn)輸機(jī)且與y軸垂直方向?yàn)閤軸；面朝煤壁，向右為x軸正方向，向左為x軸負(fù)方向。垂直于xy平面且指向煤壁方向?yàn)閦軸正方向，相反為z軸負(fù)方向。需要注意的是，原點(diǎn)是在系統(tǒng)初始狀態(tài)下設(shè)定的固定點(diǎn)，不隨采煤機(jī)的橫向運(yùn)動(dòng)或縱向運(yùn)動(dòng)而改變，并且原點(diǎn)與所對(duì)應(yīng)刮板運(yùn)輸機(jī)推溜受力點(diǎn)的聯(lián)機(jī)必須平行于yz平面。 3) 特征點(diǎn)坐標(biāo)值的計(jì)算。如圖2所示為采煤機(jī)位置定位示意圖，實(shí)線為刮板運(yùn)輸機(jī)的布置情況，實(shí)心圓點(diǎn)為各節(jié)刮板運(yùn)輸機(jī)間的鉸接點(diǎn)。圖中表示的是在初始狀態(tài)下刮板運(yùn)輸機(jī)從圓點(diǎn)經(jīng)過一次推溜后的情況，實(shí)際工作中要經(jīng)過多次推溜。設(shè)刮板運(yùn)輸機(jī)共有n節(jié)，每節(jié)長(zhǎng)度為h，初始狀態(tài)下第k個(gè)鉸接 點(diǎn)處的坐標(biāo)為(xk,yk,0)，第k節(jié)與x軸的夾角為 αk，其中且n＞0，。則當(dāng)采煤機(jī)行程為s時(shí) (1) 其中為商，0≤p≤h為余數(shù)。由此可知采煤機(jī)特征點(diǎn)位于刮板輸送機(jī)第k節(jié)上的p處。 圖2 采煤機(jī)位置定位示意圖 設(shè)采煤機(jī)起始點(diǎn)經(jīng)過m次推溜，每次推溜的距離推溜方向與z軸方向的夾角為βm，其中且m > 0。則經(jīng)過m次推溜后采煤機(jī)特征點(diǎn)的三維坐標(biāo)值(x0，y0，z0)為 (2) 2.2采煤機(jī)姿態(tài)定位 采煤機(jī)的姿態(tài)信息包括機(jī)身傾角和調(diào)高油缸位移量，其中機(jī)身傾角是由刮板運(yùn)輸機(jī)決定的；只有調(diào)高油缸位移量是可調(diào)的。這里采煤機(jī)的姿態(tài)定位是以滾筒空間坐標(biāo)的形式給出的，因?yàn)榭梢跃C合反映 機(jī)身傾角和調(diào)高油缸位移量；而采煤機(jī)的姿態(tài)控制是以調(diào)高油缸位移量的形式給出的，因?yàn)樽藨B(tài)信息中只有該項(xiàng)是可控的。采煤機(jī)的姿態(tài)定位是建立在 位置定位基礎(chǔ)上的，確定了關(guān)鍵點(diǎn)的坐標(biāo)值后，根據(jù)機(jī)身的橫向傾角和縱向傾角確定滾筒坐標(biāo)值。 1) 只考慮采煤機(jī)橫向傾角情況下根據(jù)關(guān)鍵點(diǎn)坐標(biāo)求出滾筒的x軸和y軸坐標(biāo)。如圖3所示為采煤 機(jī)調(diào)高系統(tǒng)機(jī)構(gòu)簡(jiǎn)圖在xy平面內(nèi)的投影，圖中共有 0~4五個(gè)點(diǎn)，第i個(gè)點(diǎn)的坐標(biāo)用(xi，yi)表示，點(diǎn)2、3間線段為調(diào)高油缸的伸出量，粗實(shí)線為采煤機(jī)機(jī)身，點(diǎn)0為位置定位中使用的特征點(diǎn)。其中左圖為機(jī)身水平時(shí)的姿態(tài)，右圖為機(jī)身前傾α角度時(shí)的姿態(tài)，從圖中可以看出滾筒高度不僅取決于調(diào)高油缸伸出量還取決于機(jī)身的傾角。在圖3中，當(dāng)機(jī)身橫向傾角為α?xí)r，根據(jù)特征點(diǎn)的坐標(biāo)(x0，y0)可以求出固定點(diǎn)1、2的坐標(biāo)(x1，y1)和(x2，y2)。在點(diǎn)1、2、3組成的三角形中，已知點(diǎn)1、2的坐標(biāo)值及其與點(diǎn)3間的距離，可以由二維坐標(biāo)系中兩點(diǎn)間距離的公式列出關(guān)于x3和y3的二元二次方程，解此方程能夠求出點(diǎn)3 的坐標(biāo)值(x3，y3)。同理在點(diǎn)1、3、4組成的三角形中能夠求出點(diǎn)4的坐標(biāo)值(x4，y4)，即為滾筒的旋轉(zhuǎn)中心坐標(biāo)值。 圖3采煤機(jī)姿態(tài)在xy平面內(nèi)的投影簡(jiǎn)圖 2) 考慮采煤機(jī)縱向傾角情況下對(duì)滾筒的坐標(biāo)值進(jìn)行修正。設(shè)在位置定位中求得特征點(diǎn)坐標(biāo)值為 (x0，y0，z0)，采煤機(jī)機(jī)身的縱向傾角為β，上一步得 到的滾筒在xy平面內(nèi)投影坐標(biāo)值為(x4＇，y4＇)，則可由坐標(biāo)投影關(guān)系求得滾筒的三維坐標(biāo)值(x4，y4，z4)為 (3) 3采煤機(jī)的自適應(yīng)切割 自適應(yīng)切割是采煤機(jī)控制部分的核心內(nèi)容，主要由路徑跟蹤和自適應(yīng)調(diào)節(jié)兩大部分組成。路徑跟蹤是指在各設(shè)備工作狀態(tài)正常的前提下，盡可能的按照人工示教時(shí)所記錄的數(shù)據(jù)來復(fù)原采煤機(jī)的運(yùn)行過程。自適應(yīng)調(diào)節(jié)是指在路徑跟蹤過程中判斷出采煤機(jī)所處的運(yùn)行狀態(tài)，根據(jù)不同的情況采取相應(yīng)的措施來將采煤機(jī)調(diào)節(jié)至正常工作狀態(tài)。 3.1 路徑跟蹤策略 路徑跟蹤的依據(jù)是人工示教過程中各個(gè)記憶點(diǎn)中的數(shù)據(jù)，判斷路徑跟蹤效果的指標(biāo)包括滾筒坐標(biāo)和牽引速度兩部分。在未進(jìn)行路徑跟蹤前，機(jī)載控制器根據(jù)當(dāng)前刮板運(yùn)輸機(jī)的布置情況計(jì)算出采煤機(jī)運(yùn)行到每個(gè)記憶點(diǎn)時(shí)的機(jī)身坐標(biāo)值，將其與人工示教時(shí)記憶的機(jī)身坐標(biāo)值進(jìn)行比較，求出機(jī)身的上升高度，并由此計(jì)算出此時(shí)滾筒應(yīng)當(dāng)升高的高度以及 所對(duì)應(yīng)的調(diào)高油缸位移量，作為本次運(yùn)行的理想值存儲(chǔ)至控制器中。在路徑跟蹤階段，當(dāng)采煤機(jī)運(yùn)行到第i個(gè)記憶點(diǎn)時(shí)，機(jī)載控制器會(huì)讀取第z + 1個(gè)記憶點(diǎn)處調(diào)高油缸位移量和機(jī)身牽引速度的理想值；根據(jù)z點(diǎn)與z+1點(diǎn)的間距計(jì)算出調(diào)高油缸的運(yùn)行時(shí) 間和牽引變頻器的加速曲線。當(dāng)采煤機(jī)按照這種控 制方案行駛到i + 1點(diǎn)時(shí)，機(jī)載控制器又會(huì)根據(jù)此時(shí)的調(diào)高油缸位移和運(yùn)行速度來制定z + 1至i + 2點(diǎn)間的運(yùn)行方案。這種控制策略在每個(gè)記憶點(diǎn)處都重新計(jì)算下一點(diǎn)的運(yùn)行方案，從而消除了多點(diǎn)間的累計(jì)誤差，保證了路徑跟蹤的精度。 為了驗(yàn)證路徑跟蹤策略的實(shí)際效果，本課題組研發(fā)了采煤機(jī)記憶切割實(shí)驗(yàn)平臺(tái)，如圖4所示該平臺(tái)具有與真實(shí)采煤機(jī)相同的控制功能，可以模擬采煤機(jī)的工作過程。作者基于該平臺(tái)進(jìn)行了采煤機(jī)路 徑跟蹤效果測(cè)試。實(shí)驗(yàn)所得的路徑跟蹤曲線如圖5 所示，圖中實(shí)線部分為記憶路徑，其上的實(shí)心點(diǎn)為記憶點(diǎn)，包括常規(guī)點(diǎn)和關(guān)鍵點(diǎn)；虛線部分為實(shí)際運(yùn)行路徑。實(shí)驗(yàn)結(jié)果表明：該路徑跟蹤策略可有效跟蹤所記憶的切割路徑，但在路徑的拐點(diǎn)處尚存在一定的滯后，這主要是由于調(diào)高油缸對(duì)控制命令的響應(yīng)具有一定的延時(shí)性。 圖4 采煤機(jī)記憶切割實(shí)驗(yàn)平臺(tái) 圖5采煤機(jī)路徑跟蹤曲線 3.2自適應(yīng)調(diào)節(jié)策略 自適應(yīng)調(diào)節(jié)的依據(jù)是路徑跟蹤過程中各設(shè)備的工作狀態(tài)；調(diào)節(jié)的對(duì)象是采煤機(jī)的運(yùn)行速度和滾筒高度。實(shí)踐證明當(dāng)采煤機(jī)切割到巖石后其截割電機(jī)溫度、電流以及搖臂振動(dòng)都將加大并超出正常范圍，此時(shí)應(yīng)首先降低牽引速度，之后如果采煤機(jī)狀態(tài)恢復(fù)正常則直接切割巖石；如果持續(xù)降低牽引速度一段時(shí)間后采煤機(jī)狀態(tài)仍然異常則降低滾筒高度；如果持續(xù)降低滾筒高度一段時(shí)間后采煤機(jī)狀態(tài)仍無法恢復(fù)正常則向操作人員發(fā)出警報(bào)請(qǐng)求人工干預(yù)。 自適應(yīng)調(diào)節(jié)策略強(qiáng)調(diào)以降低牽引速度作為應(yīng)對(duì)切割巖石時(shí)產(chǎn)生的電流、振動(dòng)增大等問題的首選，而不是單一的降低滾筒高度。 由于對(duì)采煤機(jī)狀態(tài)是否異常的判斷主要來自于生產(chǎn)經(jīng)驗(yàn)，而且很難建立采煤機(jī)的數(shù)學(xué)模型，因此本文采用模糊控制方法來實(shí)現(xiàn)采煤機(jī)的自適應(yīng)調(diào)節(jié)。模糊控制的概念是由美國(guó)加州大學(xué)教授L. A. Zadeh首先提出的，其基本思想是將操作人員的控制經(jīng)驗(yàn)用具有模糊含義的語言、變量加以描述，用一組條件語句構(gòu)成控制規(guī)則以及相應(yīng)的模糊推理，最終通過模糊決策得到精確控制量[9-10]。模糊控制具有如下特點(diǎn)：1) 不需要建立被控對(duì)象的數(shù)學(xué)模型，只需掌握現(xiàn)場(chǎng)操作人員或有關(guān)專家經(jīng)驗(yàn)、知識(shí)和數(shù)據(jù)[11-12]；2) 具有較強(qiáng)的魯棒性，尤其適應(yīng)于非線性時(shí)變、滯后系統(tǒng)的控制[13-14]；3) 不用數(shù)值而用語言式的模糊變量來描述系統(tǒng)，使得操作人員易于使用自然語言進(jìn)行人機(jī)對(duì)話[15-16]。 結(jié)合采煤機(jī)具體情況，本模糊控制系統(tǒng)的輸入量包括截割電機(jī)電流、搖臂振動(dòng)幅值兩部分。其中截割電機(jī)電流C的論域?yàn)榫轠0，2]，搖臂振動(dòng)幅值V論域?yàn)榫轠0，3]，模糊控制輸出量O的論域?yàn)閇-2，2]，其模糊子集均為{NB,NM,ZO,PM,PB }，分別對(duì)應(yīng)“負(fù)大”、“負(fù)中”、“零”、“正中”、“正大”。該系統(tǒng)的模糊控制規(guī)則如表1所示。 表1 自適應(yīng)調(diào)節(jié)模糊控制規(guī)則度 VOC NB NM ZO PM PB NB PB PB PB PB NB NM PB PM PM NM NB ZO PB PM ZO NM NB PM PB NM NM NM NB PB NB NB NB NB NB 其中，C，V，O中每個(gè)模糊子集的取值都需要結(jié)合操作人員和生產(chǎn)廠家的經(jīng)驗(yàn)來確定。本系統(tǒng)結(jié)合西安煤機(jī)廠MU900/2210-WD型電牽引采煤機(jī)的相關(guān)參數(shù)和設(shè)計(jì)人員的經(jīng)驗(yàn)總結(jié)出了每個(gè)模糊子集的取值及其所對(duì)應(yīng)的控制操作，如表2-4所示。其中截割電流的模糊子集ZO取值為1. 00表示正常工作時(shí)的電流，NM取值為0. 90表示正常工作電流的0. 90倍，其他取值同理。最終的模糊控制輸出如圖6所示。 表 2 截割電機(jī)電流的模糊子集 模糊子集 NB NM ZO PM PB 取值 0.80 0.90 1.00 1.30 2.00 控制操作 報(bào)警 加速 /減速/降低滾筒 停車 表 3 振動(dòng)幅值的模糊子集 模糊子集 NB NM ZO PM PB 取值 0.50 0.80 1.00 1.80 3.00 控制操作 報(bào)警 加速 /減速/降低滾筒 停車 表 4 控制輸出的模糊子集 模糊子集 NB NM ZO PM PB 取值 -2.00 -1.00 0 1.00 2.00 控制操作 停車 減速/降低滾筒 / 加速 報(bào)警 為了驗(yàn)證采煤機(jī)自適應(yīng)調(diào)節(jié)策略的控制效果，作者在西安煤機(jī)廠的采煤機(jī)工況參數(shù)模擬實(shí)驗(yàn)臺(tái)上進(jìn)行了仿真測(cè)試。如圖7所示，該實(shí)驗(yàn)臺(tái)可以模擬采煤機(jī)在不同負(fù)載下的工況參數(shù)。實(shí)驗(yàn)過程中系統(tǒng)模擬出采煤機(jī)的截割電機(jī)電流、搖臂振動(dòng)幅值，而后將其輸入到模糊控制器中，模糊控制器根據(jù)模糊判別規(guī)則控制采煤機(jī)的牽引速度。如圖8所示，當(dāng)截割電機(jī)電流和搖臂振動(dòng)幅值急速增加時(shí)模糊控制器控制牽引電機(jī)減速;隨著牽引速度的降低截割電機(jī)電流和搖臂振動(dòng)幅值均有所下降;當(dāng)截割電機(jī)電流和搖臂振動(dòng)幅值趨于正常值時(shí)模糊控制器不再降低牽引電機(jī)速度，此時(shí)牽引速度趨于平穩(wěn)。 圖6 采煤機(jī)模糊控制輸出圖 圖7 采煤機(jī)工況參數(shù)模擬實(shí)驗(yàn)臺(tái) 4 結(jié)論 作為綜采工作而的關(guān)鍵設(shè)備，采煤機(jī)的自動(dòng)化是實(shí)現(xiàn)綜采工作而自動(dòng)化和少人化的重點(diǎn)和難點(diǎn)。記憶切割技術(shù)是被實(shí)踐證明最有效的采煤機(jī)自動(dòng)控制方法，但在我國(guó)復(fù)雜的地質(zhì)條件下煤巖界而變化劇烈，僅依靠記憶切割技術(shù)并不適用。針對(duì)這一問題，筆者提出了 基于模糊控制理論的采煤機(jī)自適應(yīng)記憶切割技術(shù)，設(shè)計(jì)出了采煤機(jī)位置、姿態(tài)空間定位系統(tǒng)和自適應(yīng)切割模糊控制系統(tǒng)。在保證采煤機(jī)工作狀態(tài)正常的前提下，該系統(tǒng)可以避免搖臂的頻繁升降，確保了頂板、底板的平整性并提高了采煤效率。目前該系統(tǒng)在MU900/2210 WD型電牽引采煤機(jī)1:6樣機(jī)模型構(gòu)成的實(shí)驗(yàn)平臺(tái)上進(jìn)行了路徑跟蹤測(cè)試，證明了路徑跟蹤策略的可行性。并且在西安煤礦機(jī)械有限公司的采煤機(jī)工況模擬實(shí)驗(yàn)臺(tái)上進(jìn)行了自適應(yīng)調(diào)節(jié)測(cè)試，實(shí)驗(yàn)數(shù)據(jù)表明:模糊控制系統(tǒng)可以識(shí)別采煤機(jī)的異常工作狀態(tài)，并采取相應(yīng)的控制方法將其恢復(fù)至正常狀態(tài)。接下來將進(jìn)行采煤工作而的現(xiàn)場(chǎng)實(shí)驗(yàn)，并根據(jù)實(shí)驗(yàn)結(jié)果對(duì)系統(tǒng)做進(jìn)一步的改進(jìn)和完善。 圖8 采煤機(jī)自適應(yīng)調(diào)節(jié)控制效果 附錄B 外文文獻(xiàn) Modelling and Simulation on Shearer Self-adaptive Memory Cutting Abstract:Automation of shearer is the key point to realize the fully mechanized coal face. According to the complicated geological condition in our country, this paper built a shearer self-adaptive memory cutting model based on fuzzy control theory. This model contains shearer positioning system and fuzzy control system which can get the message of shearer's position and attitude at any point, trace the memorial cutting path automatically, judge whether the shearer cuts rocks based on fuzzy control theory and find the optimal scheme. The author simulated the working environment in laboratory and factory, did experiment to test whether the model can adapt complicated geological condition. Key words: shearer; fuzzy control; modelling; simulation 1. Control model of shearer self-adaptive memory cutting For the shearer there are two important movements when it working: the horizontal reciprocating motion and the longitudinal direction of the rocker arm movements. The former corresponds to the speed of traction motor, and the latter corresponds to the telescopic amount of the height adjusting oil cylinder. So the mining machine control is mainly determined by the traction motor and the height adjusting oil cylinder. Conventional memory cutting technology requirements of shearer arm dropping immediately when cutting to the rock, so as to avoid the cutting motor blocking or cutting tooth fracture. As the electric power increases and the cutting tooth material improvement, it makes the shearer can be cut directly on the general hardness ofrock. So in this scheme, the controller can not only identify whether or not cutting into the rock, but also able to determine whether to allow the direct cutting according to the coal mining machine of the sensing data. The control model of shearer self-adaptive memory cutting is shown in Fig.l , which can be divided into three stages: artificial teaching stage, the adaptive cutting stage and manual correction phase. Each stage is a relatively independent and interrelated. Fig. l Control model of shearer self-adaptive memory cutting 2. Position and attitude model of shearer Shearer position and attitude is the important information of self-adaptive memory cutting, and affect the mining machine control effect directly. Some scholars have proposed that we can use shaft encoders calculate the walking distance to locate the coal mining location, use isplacement sensor get the high oil cylinder's expansion amount for attitude positioning. However, in this method as mentioned in reference [1] and [2], we can only get the relative value of the position and attitude rather than the absolute value of three-dimensional. Thus, the solution of shearer position and attitude positioning is the prerequisite and basis for achieving the daptive cutting. Fig. 2 Model of shearer position Fig. 3 Model of shearer attitude The most intuitive way to determine the position of shearer body is to get its coordinates in 3D space,and this requires solving the following problems: Select a fixed point on the shearer as a feature point for the location positioning; defined three-dimensional coordinate system; derived calculation formula of feature point's coordinates. Fig 2 shows the model of shearer position. The solid line is the arrangement of the scraper conveyor; solid dots are the hinge point of each ection between the scraper conveyors. In the case of only considering the coal mining machine horizontal angle, we can get the X axis and Y axis according to key point coordinates of a drum. Fig.3 shows the projection of the shearer hydraulic system schematic diagram of mechanism in XY plane. Fig.3 have a total of five points(0-4), the coordinates of the number i point is (xi,yi), that the line segment between points 2 and 3 is protrusion length of the cylinder, thick solid line is shearer body, Point 0 is the feature points used in location positioning. In Fig 3, left figure is the body's level attitude, right figure is the body attitude forwarding angle α. 3. Simulation of shearer self-adaptive memory cutting Shearer self-adaptive memory cutting is the key point of shearer control, which contains path tracking and adaptive adjustment. Path tracking refers to that we can recover the coal mining process with manual data recorded when the operation as much as possible under the premise of the normal working state, Adaptive adjustment refers to that we can determine the shearer operating state in the process of path tracking, and take appropriate measures to adjust shearer to normal working condition depending on the situation. 3.1 Simulation of path tracking In order to verify the practical effect of the path tracking strategy, the group developed a experiment platform of shearers memory cutting. Fig.4 shows that the platform has the same control as real shearer, and can simulate the working process. Author had a path tracking test results based on the platform of the shearer. Experimental curve from the path tracking shown in Fig.S, solid line portion is the path memory, and its solid points are memory points, including conventional point and critical point, dashed part is the actual operation path. The results show that: the path tracking strategy can be effectively tracked by the memory of the cutting path, but the inflection point in the path are still some lag, which was mainly due to that the response of height adjusting oil cylinder to the command control has a certain latency. Fig. 4 Experimental platform of memory cutting Fig. 5 Path tracing curve of the shearer 3.2 Simulation of adaptive adjustment Because of the judgment of state about shearer whether is abnormal mainly from production experience it is difficult to establish mathematical model of shearer, so this paper adopts fuzzy control method to realize shearer's self-adaptive regulation strategies. Concept of fuzzy control was firstly proposed by L.A.Zadeh professor at University of California. Its basic idea is to describe the operator's control experience with a language and variables having vague meaning. The control rules and corresponding fuzzy reasoning are constituted by a set of conditional statement. Finally we get accurate control variable through fuzzy decision. Fuzzy control as mentioned in references [3-6] has the following characteristics: there is no need to establish mathematical model of controlled object, just master knowledge, experience and data of operators or concerned expert; it has strong robustness especially apply to control of nonlinear time-varying and delay system; we describe the system not with numerical but fuzzy variable of language allowing it's easy to achieve man-machine interaction with natural language for operators. Combined with the specific circumstances, the input of the fuzzy control system including: cutting motor current and vibration amplitude of rocker arm. The universe of discourse of cutting motor current C is [0,2]; the universe of discourse of vibration amplitude of rocker arm V is [0,3]; the universe of discourse of output of fuzzy control. Theirs fuzzy subsets are {NB, NM, ZO, PM, PB{ corresponding to "negative big", "negative medium", "zero", "positive medium", "positive big". The system's fuzzy control rules are shown in table 1. The value of each fuzzy subset in C, V, O is determined by combining the experience of both operators and manufacturers. Tab. 1 Fuzzy control rules of adaptive adjustment VOC NB NM ZO PM PB NB PB PB PB PB NB NM PB PM PM NM NB ZO PB PM ZO NM NB PM PB NM NM NM NB PB NB NB NB NB NB The authors have done the test on a simulation experiment table in Xi'an Coal mining Machinery Co, Ltd. As shown in Fig 5 this experiment table can simulate shearer working parameters at different loads. Cutting motor current and vibration amplitude of rocker arm are simulating by this system during the experiment, and then we put them into the fuzzy controller. Fuzzy controller control the traction speed of shearer based on fuzzy judgment rules. As shown in Fig.6 when cutting motor current and vibration amplitude of rocker arm are rapidly increasing the speed of traction motor is reduced controlled by fuzzy controller; with the reduction in traction speed cutting motor current and vibration amplitude of rocker arm are decreased; when the value of cutting motor current and vibration amplitude of rocker arm tend to be normal fuzzy controller no longer reduce the speed of traction motor, at this time the value of traction speed is tend to stabilized. Fig. 5 Experimental platform of shearer adaptive adjustment Fig. 6 Effect of shearer adaptive adjustment 4. Conclusion Memory cutting has been proven the most effectively for shearer's automatic control. In China geological conditions is very complex and coal-rock interface change rapidly so depending only on memory cutting technology is not applicable. In order to solve this problem, the author put forward a shearer self-adaptive memory cutting model based on fuzzy control theory, which contains the position model, attitude model and the fuzzy control model. At present, the path tracking was tested in experimental platform constituted by 1:6 prototype model of MG900/2210-WD AC electric haulage shearer. And the author has demonstrated the feasibility of the path tracking strategy and conducted a self-adaptive test on shearer's working parameters simulation experiment table in Xi'an Coal mining Machinery Co., Ltd. Next there will be a field experiment in coalface and further improvement will be done on this model according to the results. References [1] F. Xinqiu, H. Jie, Z. Bin, et al., Self-positioning system of the shearer in unmanned workface. Journal of Xi'an University of Science and Technology, 28 (2), 2008, p. 349-353 [2] X. Huguo, "Principle and application of shearer position monitoring device," Mining&Proce- ssing Equipment, 38(11), 2007, p25-27. [3] C. Kaiyuan, Z. Lei, Fuzzy reasoning as a control problem, Fuzzy Systems, 16 (3), 2008, p. 600-614 [4] Laurent Foulloy, Sylvie Galichet, Fuzzy control with fuzzy inputs, Fuzzy Systems, 11 (4),2003, p. 437-449 [5] Rodolfo E. Haber, Jose R. Alique, Fuzzy logic-based torque control system for milling process optimization, Systems Man and Cybernetics, 37 (5), 2007, p. 941-950 [6] H.K. Lam, L.D. Seneviratne, Tracking control of sampled-data fuzzy-model-based control systems, Control Theory and Applications, 3 (1), 2009, p. 56-57 - 12 -