外文翻譯--有限元法、模糊邏輯以及虛擬現(xiàn)實模型對長壁采煤法的模擬

1、<p>　　中文4037字，2582單詞，13800英文字符</p><p>　　出處：Javier Toraño, Isidro Diego, Mario Menéndez and Malcolm Gent. A finite element method (FEM) – Fuzzy logic (Soft Computing) – virtual reality model ap

2、proach in a coalface long wall mining simulation. 2008, Vol. 17(4): 413-424</p><p><b>　　附錄外文翻譯</b></p><p>　　A finite element method (FEM) – Fuzzy logic (Soft Computing) – virtua

3、l reality model approach in a coalface long wall mining simulation</p><p>　　Javier Toraño, Isidro Diego, Mario Menéndez and Malcolm Gent</p><p>　　Abstract：The modeling of the behavior

4、of a long wall coal mining installation is not an easy task, particularly when we deal with a coal seam exceeding4 m of thickness. The variables playing a role on the performance of the shearer and powered roof support a

5、re both difficult to define and quantify and technologies involved belong to several different scientific areas: geology, rock mechanics, stress-deformation calculations and hydraulics.</p><p>　　State of the

6、 art modeling tools ，fuzzy logic, neural networks and three-dimensional (3D) finite element calculations are employed in order to develop a computerized model that will allow predicting the response of the installation a

7、gainst changing operation conditions. This response is checked against extensive data obtained from deep measurement campaigns and finally shown to the system user through Virtual Reality Modeling Language (VRML) tools.&

8、lt;/p><p>　　Keywords: Fuzzy logic; Coal long wall; Virtual reality; Powered roof support</p><p>　　1. Introduction</p><p>　　The exploitation of coal deposits is a science with a large a

9、mount of variables and a high percentage of uncertainties. The exploitation method to be used will depend mainly on the regularity and shape of the deposit. In case of flat and regular coal scams with slopes lower than 3

10、50 the preferred method is the so called "long wall coalface", a highly productive method which can be almost completely mechanized. It has been widely used all over the world since the 1950s.</p><p&

11、gt;　　In this method a long rectangle of coal, called ”panel” and ,ranging 1000× 80 m is mined in a single pass. The face, 80m wide, is mined using a shearer, a traveling mechanism equipped with a pair of rotating dr

12、ums that extract the coal by means of metallic picks (sec Fig. 2). The height of the coal is called thickness and the long wall method is applied in scams which minimum thickness is 2m. In the case to be presented in thi

13、s paper scam thickness ranges between 3 and 4.2 m. The sheared coal fa</p><p>　　The rocks over the working area in front of the shearer are retained by a Shield support (Fig. 1), which holds the terrain stre

14、sses by means of hydraulic jacks. More than 50 units can be mounted side by side to protect the whole face and these supports advance autonomously following the hollow opened by the shearer by means of another set of hyd

15、raulic jacks. The rock masses collapse behind the face as the supports move following the shearer.</p><p>　　Although all the mining, transport and support systems arc fully mechanized, the changing behavior

16、of the surrounding rock masses as the face passes by and the inherent changing nature of the rock scams force the use of a complete and direct human machine operation and supervision. Even more, the results of the mining

17、 operation can not be adequately predicted. The large number of factors that play a role in the coalface advance makes extremely difficult to predict future advances from very exper</p><p>　　It would be extr

18、emely interesting to have a tool prepared to simulate the behavior of the referred above systems and will open the possibility to a quasi-autonomous control of the systems. But the development of this tool is an extraord

19、inary complex task. Technical skills belonging to very different disciplines are involved: geology, rock mechanics, stress-deformation calculations and hydraulics among others.</p><p>　　This paper shows the

20、development and validation of a computerized model that simulates the response of along-wall exploitation against the changing operation conditions, as well as its visual representation in a clear manner for the operator

21、. The working strategy and the instruments used in this development are:</p><p>　　-The earth pressures arc simulated through neural networks (Cheng and Peng [1,2]).</p><p>　　-An extensive measur

22、ements campaign of the coalface parameters has been carried out (Toraňo et al. [3]).</p><p>　　-The roof support has been simulated by FEM method using the commercial code Cosmos. The software is able to calc

23、ulate the stress-strain state of the support depending on the load previously calculated by the neural networks.</p><p>　　-The complex response of the long wall face has been simulated using a fuzzy model, m

24、odel tuned and validated using the data previously obtained in the measurements campaign.</p><p>　　-The whole installation is being represented virtually through the use of VRML tools. Then all the data prev

25、iously calculated can be shown at the same time and in a comprehensive manner.</p><p>　　The mix of the methods referred above create a working strategy capable of simulate the changing environment of a long

26、wall coal exploitation. This method gives a very usable tool to schedule the exploitation of future coal panels, as it can predict the foreseeable performance of the installation when applied to other areas of the coal d

27、eposit.</p><p>　　It could also ease the operation of the machinery as direct information about the inferred state of the surrounding rock mass can be presented to the operator, thus diminishing, although nev

28、er eliminating, the need of man force in the coal face. It will also open path for a possible way of autonomous operation of the installation, which cannot be checked in the simulated installation due to lack of electron

29、ics associated to the hydraulics systems that control the supports.</p><p>　　2. Excavation and roof support equipment</p><p>　　The exploitation method through long wall mining using Glinik-18/41

30、-Poz roof support and double drum shearer [3], is applied to coal seam exploitations in very difficult geological and mining conditions, particularly when the roof and floor rock masses have very low Geomechanical proper

31、ties.</p><p>　　The long wall equipment was sized according to the higher requirements and was supplied by Polish manufacturers.</p><p>　　The powered support consists of 53 two leg lemniscate shi

32、eld support Glinik 18/41 Poz(Fig.1)ranging from 1.80 to 4.10m in height and 1.50m in width. The coal-winning machine is a bidirectional double drum shearer loader Famur KGS-324 with 2×132 kW cutting power and the fa

33、ce conveyor is a Glinik 260/724 BP1 with maximum carrying capacity up to 1000 t per hour. Main characteristics of the mining equipment are summarized in Tables 1 and 2.Several factors were taken into account in the selec

34、tion of </p><p>　　a) The canopy would not be rigid; a sliding front part (fore pole) will adjust its length in order to support the immediate roof as soon as possible</p><p>　　b) The face guard

35、(flipper) not only would push against the coal face, but it would have maximum deflection (up to take horizontal position) in order to cover the roof when it could not been done by the extendable canopy </p><p

36、>　　c) The support density would be high both at set and at yield to ensure stability in the immediate roof</p><p>　　d) The advancing ram (conveyor pushing system) would have auxiliary hydraulic cylinders

37、to bring up or to get down the conveyor and the cutting horizon</p><p>　　e) The shield base would transfer a low pressure on the floor</p><p>　　f) The shield would have hydraulic cylinders to co

38、rrect its lateral position to avoid tilting of the shield.</p><p>　　Fig.2 shows the shield support, the shearer and the face conveyor are shown in operation in the long wall face. Details of the front part o

39、f the canopy are shown in Fig.3.</p><p>　　The planning and operation of those mechanized installations have been traditionally based in the experience of the mine technicians and in the results of detailed d

40、aily measurements campaigns, that in the actual stage and after two and a half mined panels(about 2000 m)summarize around 165.000 measured values [4].</p><p>　　The daily measurements data base contains: date

41、, long wall advance, mined thickness, coal presence in the floor, longitudinal dip ，cross dip ，support mean load density, coalface sterility and shattering.</p><p>　　3. FEM dynamic modeling</p><p&

42、gt;　　3.1. Dynamic FEM</p><p>　　Finite Element Modelling,FEM,comprises:3D design and modelization of the different equipment parts, assembling those parts and the roof support units, simulating their movement

43、s in the mine and the 4D stress calculation of the moving roof support([5,6]).To design the several pieces there have been used powerful·3D modeling tools as are SolidWorks and CosmosM software which allow us to mak

44、e the 3D design and assembly of each equipment component, its animation through programming and the calculati</p><p>　　The long wall has 58 powered roof supports. We will analyze only one support as they are

45、 all identical, being all installed side by side and continuously in front of the mined coalface. Each joint between the 89 pieces that creates each support has to be modeled and its physics defined, reaching a final ass

46、embly shown in Fig.4 totalizing 236 relationships between elements. Solidworks has a render module that can assign textures to each of the pieces for visualization purposes, as can be seen in </p><p>　　Once

47、the roof support is assembled the process continues with the simulation of the movements using Cosmos Motion software. The generation of the support movement has been done through the introduction of the cinematic equati

48、on of all components using the Cosmos Motion feature called“expressions”(by means of FORTRAN language),thus better defining the relative motion of each piece.</p><p>　　The structure of this language is the f

49、ollowing: if(time?x1:y1,y1,if(time?x2:y2,y2,if(time?x3:y3,y3,0))).</p><p>　　3.2. Stress-displacements calculation</p><p>　　Once the roof support has been design in three dimensions a stress–stra

50、in calculation will be made, taking as boundary conditions the loads determined by</p><p>　　the long wall characteristics. One of the boundary conditions most difficult to </p><p>　　quantify is

51、the roof load, characterized by the support capacity, the distance</p><p>　　among immediate roof weighing and the frequency of the overloads. The method used is the neural networks, the Cheng and Peng one[1]

52、or another one developed by us using the commercial code “SPSS Easy NM”, networks that in both cases have been trained to our exploitation conditions.</p><p>　　Following the Cheng and Peng neural network the

53、 output values relative to their long wall installations are:1)Mean load density required in the roof support of Sd, de 86.18 t/m2, 2)a distance among immediate roof loads, L, of 18.42 m and 3)the overload frequency fy,

54、of 43%.Similar results are obtained using commercial software, “EasyNN” from SPSS, where the data of the American mines have been introduced in the network design, obtaining the outputs for the American mines where Cheng

55、 and Peng dev</p><p>　　The outputs of the roof support are Sd=90.6t/m2, L=19.8m and fy=44.2% which do not deviate from the matrix calculations. The adaptation of these neural networks to our working conditio

56、ns(more exploitation depth, softer coals and different stress distribution in the rock mass)has allowed the achievement of results similar to the ones obtained from the experience and measurements: roof load of Sd=48t/m2

57、,overload frequency of fy=34% and distances among immediate roof load of L=21 m. Once the loads </p><p>　　4.Fuzzy logic model development</p><p>　　4.1.Fuzzy logic model development</p>&l

58、t;p>　　The FEM calculations shown above provide a computational frame that allows us to get a good understanding of the structural mechanic behavior of the self-advancing system at mine. However, another set of modelin

59、g parameters plays a key role in the global understanding of our coal wining system.</p><p>　　More precisely, the interplay between the structural behavior of the coal-winning system and the geological and m

60、orphological features of the coal seam represent a system of high complexity that has no formal representation neither in mathematical closed form nor using traditional computing algorithms.</p><p>　　Such in

61、terplay offers, as a result, the daily advance in meters of the coal wining system as a function of several variables representing an “a-priori” unknown 5-D hyper surface that can be expressed in its general form by the

62、following expression:</p><p>　　S=f (x1;x2;x3;x4;x5),</p><p><b>　　where</b></p><p>　　S Daily advance</p><p>　　x1 Thickness of coal-seam</p>&

63、lt;p>　　x2 Transversal coal-seam dip</p><p>　　x3 Longitudinal coal-seam dip</p><p>　　x4 Roof fractures(upper region of the coal-seam)</p><p>　　x5 Coal hardness<

64、/p><p>　　The importance of obtaining a good estimation for the advance of the coal winning system can be easily understood if we take into account that it serves as one of the fundamental parameters for forecas

65、ting future mine developments. In the coal mining business, an exquisite planning usually means successful results when confronted not only to other coalmining companies, but to competing energy sources, too.</p>

66、<p>　　Soft computing solutions based on Fuzzy-Logic theories have revealed themselves as a useful tool for representing complex geological phenomena[7].This approach has allowed us to represent the hyper surface S w

67、ith a set of linguistic variables composed by fuzzy sets, a knowledge database expressed by fuzzy rules and a suitable model topology.</p><p>　　There have been used triangular and trapezoidal membership func

68、tions for defining the fuzzy-sets that form the linguistic variables. While bell-shaped or other more sophisticated membership functions can hypothetically provide a better performance in some applications [8],we have ch

69、osen these geometric shapes in order to allow its use in a robust and quick inference engine that we have implemented on Microsoft(MS)Excel. The numerical structure of every fuzzy-set is formed by its four singular po<

70、;/p><p>　　Fig.12 shows the table of the chosen numerical values for defining the x1…x5 linguistic variables at play and their related fuzzy-sets. Of special importance is the definition of S as a fuzzy variable

71、, both for input and output.</p><p>　　This deceptively double behavior will be explained in the next section.Fig.13 shows a graphic representation of the numerical values from Fig.12.Note the existence of de

72、ceptively two different versions of the linguistic variables for “Daily Advance” defined for Input/Output.</p><p>　　6. Conclusions</p><p>　　The combined use of several advanced simulation techni

73、ques allows the autonomous control of a long wall mining in a coal exploitation.</p><p>　　The use of neural networks allows the knowledge of the stress state of the surrounding ground. This knowledge and the

74、 detailed structural and constructive characteristics of the support systems allow the simulation of the behavior of the roof supports through finite element method.</p><p>　　Starting from extensive field me

75、asurements a fuzzy model is developed and validated, capable of simulating the response of the shearer-power roof support system to the conditions of the surrounding ground. The graphical representation of the obtained d

76、ata and calculations is done by means of a three dimensional virtual reality model, procedure that allows to the operators of the mining production system the real time visualization of the exploitation state.</p>

77、<p>　　The combined use of these techniques allows the graphic representation in an interactive way not only operation parameters of the coal mining and excavation machinery(location, advances, roof support pressure

78、s among others)but also parameters impossible to measure or observe directly(predicted advance, stress condition of the ground).</p><p><b>　　中文翻譯</b></p><p>　　有限元法、模糊邏輯以及虛擬現(xiàn)實模型對長壁采煤法

79、的模擬</p><p>　　摘要：長壁采煤法的動態(tài)模擬不是一項簡單的工作，尤其是對于厚度超過4m的煤層而言。這些關(guān)于滾筒采煤機和強力支架的工作面運作的參數(shù)無論是定義還是量化都很困難。其中應(yīng)用到涉及到多個不同的學(xué)科的技術(shù)，包括：地質(zhì)、巖石力學(xué)、應(yīng)力和變形的計算以及水力學(xué)。</p><p>　　建模工具，模糊邏輯，神經(jīng)網(wǎng)絡(luò)和3D有限元計算被應(yīng)用到其中來建立計算機模型以便預(yù)測不斷變化的工作狀態(tài)下

80、設(shè)備的反應(yīng)。這種反應(yīng)經(jīng)過由進一步的測量所獲取的數(shù)據(jù)檢驗之后最終通過VRML工具展示給系統(tǒng)的使用者。</p><p>　　關(guān)鍵詞：模糊邏輯 長壁采煤法 虛擬現(xiàn)實 強力支護</p><p><b>　　1概述</b></p><p>　　煤礦床的開采是一門具有極度多樣性和不確定性的科學(xué)。采煤方法的選用主要取決于礦床的規(guī)律和產(chǎn)狀。對于傾角小于35度近

81、水平和規(guī)則的煤層，常用的方法是被稱作長壁采煤法的，一種幾乎可以完全采用機械化的高產(chǎn)高效采煤法。這種采煤方法從二十世紀50年代開始被廣泛應(yīng)用于世界各地。</p><p>　　在這種方法中，每次推進采下1000×80m的一塊長條形的煤，稱作“煤壁”。這個80m寬度的面，采用滾筒采煤機開采，滾筒采煤機是一種自移式機械，裝備一對可調(diào)高滾筒，通過金屬截齒切割煤體。煤體的高度被稱為煤厚，長壁采煤法適用于最小厚度為2

82、m的煤體。本文中為說明問題，煤層的厚度規(guī)定為3～4.2m。采下的煤落在一個工作面輸送機上然后被運出采煤工作地點。</p><p>　　工作地點滾筒采煤機上部的巖石被液壓支架支撐，由液壓千斤頂承擔其所受壓力。50多部支架并排放置以保護整個工作面，這些支架在另一種液壓千斤頂?shù)淖饔孟伦詣忧耙聘M滾筒采煤機切割出的空間，當支架緊跟滾筒采煤機前移后，頂板巖層隨之垮落。</p><p>　　盡管割煤、

83、運輸和支護體系全部實現(xiàn)機械化，但是工作面推進帶來的圍巖運動和巖層自身性質(zhì)的變化迫使我們采用一種人工的方法的操作機器和進行管理。更何況，開采的結(jié)果不能被充分的預(yù)知。煤壁的推進受到諸多因素作用，這給預(yù)測下一步開采情況帶來了極端的困難，僅靠經(jīng)驗豐富的工程師和工人無法完成。而且這種預(yù)測還僅限于同一煤層相似的條件下。工作環(huán)境是不斷變化的，地質(zhì)區(qū)域常常由一種完全合適的開采條件迅速且不確定地變得非常不適合開采。</p><p>

84、;　　能有一種工具來模擬上面提到的情況會是很有趣的而且這將開創(chuàng)一種半自動化控制的新局面。不過這種工具的研發(fā)是一項極為復(fù)雜的工作。這項工作將用到不同學(xué)科的技術(shù)，包括地質(zhì)、巖石力學(xué)、應(yīng)力和應(yīng)變計算以及水力學(xué)等。</p><p>　　本文介紹一種用來模擬長壁采煤法在工作環(huán)境改變時的反應(yīng)并給予操作者以可見的明確的處理方法的計算機模型的創(chuàng)建和論證過程。創(chuàng)建過程中應(yīng)用的理論依據(jù)和實驗設(shè)備為：</p><p

85、>　　地壓用神經(jīng)網(wǎng)絡(luò)模擬；</p><p>　　大量的采煤面參數(shù)的實測數(shù)據(jù)已經(jīng)獲得；</p><p>　　頂板支護用Cosmos代碼通過有限元法模擬；</p><p>　　長壁工作面復(fù)雜的反應(yīng)用模糊模型進行模擬，模型通過先前獲得的實測數(shù)據(jù)進行修正和校驗；</p><p>　　整個裝置由VRML工具進行創(chuàng)建，創(chuàng)建完成后所有的實測數(shù)據(jù)可以一

86、次性的全面的顯示出來。</p><p>　　上面提到的幾種方法的結(jié)合創(chuàng)造出一種工作策略能夠模擬出長壁采煤法工作環(huán)境的改變。這種方法提供給我們一種十分有用的工具來對將來的采煤面列表，當應(yīng)用于其它礦床的時候能預(yù)測出裝置將來的一些可預(yù)見的工作狀態(tài)。</p><p>　　這一工具還簡化了對機械的操作，由于它能將圍巖相關(guān)狀態(tài)的直接信息展現(xiàn)給操作者，這樣盡管不能排除畢竟減少了工作面的人力消耗。它同時還

87、為裝備的自動化開辟了一條可能的道路，在裝置模擬中由于缺少控制支架的液壓系統(tǒng)的電氣相關(guān)資料而無法得到。</p><p><b>　　2 掘進和支護設(shè)備</b></p><p>　　開拓方式中，長壁采煤法使用的Glinik-18/41-Poz型支架和雙滾筒采煤機適用于極差的地質(zhì)和開采條件，尤其是頂?shù)装鍘r性都較軟的條件。</p><p>　　長壁開采

88、設(shè)備按照不同層次要求分類并由西班牙生產(chǎn)商提供。</p><p>　　強力支架由53部Glinik-18/41兩柱式支架組成，支架支撐高度1.80~4.10m，架間寬度1.5m（如圖1所示）。采煤機為Famur KGS-324型雙滾筒采煤機，滾筒功率2×132KW，工作面輸送機為Glinik260/724BP1型,最大輸送能力1000t/h，采煤機械的主要性能如表1和表2所示。影響支架選型的幾個關(guān)鍵因素為

89、：</p><p>　?。╝）頂梁不能為剛性，滑動的前探梁可以使支架的長度更及時支撐起直接頂；</p><p>　　（b）護幫板不應(yīng)緊靠煤壁，應(yīng)該最大限度的展開呈水平以便能托起頂梁接觸不到的頂板和邊緣的支護密度適當加大以保證直接頂?shù)姆€(wěn)定</p><p>　　（c）推進裝置需要有輔助的液壓缸以調(diào)高或者降低輸送機的高度和切割水平</p><p>

90、　?。╠）支架的底座傳遞給底板一部分壓力</p><p>　?。╡）支架有液壓缸以調(diào)整傾向的高度防止支架傾倒</p><p>　　圖二所示為掩護式支架，滾筒采煤機和刮板輸送機正在運行。頂梁前部的細節(jié)如圖三所示。</p><p>　　設(shè)備的調(diào)試和運行以往都依靠工程師的經(jīng)驗和詳細的日常觀測數(shù)據(jù)，這些數(shù)據(jù)的獲得通常要在現(xiàn)場經(jīng)過回采兩個半工作面（大約2000m）后從165,

91、000個數(shù)據(jù)中總結(jié)得到。</p><p>　　日常觀測數(shù)據(jù)主要包括：日期、推進度、采高、底板產(chǎn)狀、縱向傾角、橫向傾角，支架載荷密度、煤壁光滑及破碎的程度。</p><p>　　3 有限元法力學(xué)模型</p><p><b>　　3.1力學(xué)FEM</b></p><p>　　有限元法，F(xiàn)EM，包括：不同組件的三維設(shè)計和建模，

92、三維模型和頂板支護部件的組裝，支架各部件在煤層中動作的模擬和頂板支護移動時的4D壓力計算（[5/6]）。以上幾部分的設(shè)計要用到SoildWorks和CosmosM軟件等功能強大的工具，這些工具能夠讓我們完成三維設(shè)計和設(shè)備各部件的組裝以及模型活動和設(shè)備在工作時應(yīng)力和變形的計算。</p><p>　　長壁工作面共有58部支架。由于它們都是按同樣的方式依次緊靠前一部支架向前移動的，因此我們只分析一部支架即可。支架89塊

93、組件之間的每個連接部位都被模擬出來并描述出其物理屬性，最終得到圖4所示的總共236處聯(lián)系。SoildWorks有一個轉(zhuǎn)化組件能就給每一塊模型賦予質(zhì)地效果就像圖中顯示的那樣，圖5顯示了五部支架聯(lián)合工作時的布置形式。</p><p>　　當頂板支護組建完成后，支架運動的過程隨著CosmosMotion軟件對移動的模擬而繼續(xù)下去。通過對支架各組成部分同步電傳感的引入，運用CosmosMotion的“expresions

94、”功能（由FORTRAN語言編寫）可以得到支架移動的整個過程，因此也更好地解釋了每一塊的移動之間的關(guān)系。</p><p>　　這一語言可以用如下結(jié)構(gòu)表示</p><p>　　if(time？x1：y1,y1,if(time？x2：y2,y2,if(time？x3,y3,y3，0))).</p><p>　　3.2 應(yīng)力-位移計算</p><p&g

95、t;　　頂板的支護分解到三維方向上之后，認為此時的荷載取決于長壁工作面自身的特性，就可以計算出極限狀態(tài)的應(yīng)力和應(yīng)變。頂板荷載是諸多極限位置中最難計算的量之一，它由支護能力、直接頂重量和來壓的頻率決定。這時我們應(yīng)用到程和平制作的或者我們自己利用“SPSS Easy NM”語言制作一個神經(jīng)網(wǎng)絡(luò)，這兩個網(wǎng)絡(luò)都被用于我們的開拓情況。</p><p>　　應(yīng)用程和平制作的神經(jīng)網(wǎng)絡(luò)得到的對應(yīng)于他們所用的長壁采煤設(shè)備的數(shù)據(jù)是：

96、1）頂板的需用支護密度Sd，86.18d/m2；2）直接頂載荷分布的間距L，18.42m；3）過載荷的頻率fy，43%。我們應(yīng)用“SPSS Easy NM”語言，分析Cheng and Peng先生發(fā)明他所用的方法時所用的美國煤層所測得的數(shù)據(jù)也得到了相似的結(jié)果（如圖6所示）。</p><p>　　分析結(jié)果Sd，90.6d/m2；L，19.8m；fy，44.2%也未偏離模擬結(jié)論。將神經(jīng)網(wǎng)絡(luò)應(yīng)用于我們的大采深、軟煤層

97、和不同應(yīng)力分布的巖層的實驗條件下之后，得到了與經(jīng)驗值和實測數(shù)據(jù)相似的結(jié)果：Sd=48t/m2，頂板過載荷頻率fy=34%，直接頂載荷分布的間距L=21m。當荷載確定后，有限元法就可以應(yīng)用了。上面提到的全部的模擬工作現(xiàn)在都可以用來得到有效得到數(shù)據(jù)來驗證對頂板支護所做的設(shè)計。圖7示所示為空載狀態(tài)圖8所示為煤壁前方6cm處支架的最大受力狀態(tài)。圖9所示為支架的移動變形，圖10所示為由Von Mises標準說確定的安全極限。</p>

98、<p>　　4．模糊邏輯模型開發(fā)</p><p>　　4.1 模糊邏輯模型的開發(fā)</p><p>　　上面介紹的有限元計算給我們提供了一種模式，能讓我們很好的理解自移式支架在煤層中運動的機械運動。而另一類模擬參數(shù)則對我們整體性地理解采煤體系起著決定性的作用。</p><p>　　更準確地說，采煤系統(tǒng)設(shè)備的移動和煤層地質(zhì)形態(tài)的變化兩者的相互作用造成了一種以

99、前從沒出現(xiàn)過的復(fù)雜體系，這種體系無論使用數(shù)學(xué)擬合還是普通的計算機算法都無法得到。</p><p>　　這種相互作用使日進尺具備了一種多元函數(shù)的特性，這是一種很重要的“5D”超前面，它的一般形式可以表示為如下形式：</p><p>　　S=f（x1，x2，x3，x4，x5），</p><p><b>　　式中S——日進尺</b></p>

100、;<p><b>　　x1——煤厚</b></p><p>　　x2——推進方向上的傾角</p><p><b>　　x3——縱向傾角</b></p><p><b>　　x4——來壓步距</b></p><p><b>　　x5——煤層硬度</b

101、></p><p>　　如果我們考慮到正確判斷采煤系統(tǒng)的推進情況可以給我們提供用來預(yù)測未來煤層的產(chǎn)狀的一個基本參數(shù)這一點的話，那么它的重要性是很容易理解的。在采煤行業(yè)中，一個完善的計劃常常意味著成功，尤其是當我們面對的競爭不光來自同行業(yè)的其他企業(yè)更有其他能源產(chǎn)業(yè)的時候。</p><p>　　基于模糊理論的計算機軟件已經(jīng)證明了自己在構(gòu)建復(fù)雜的地質(zhì)現(xiàn)象模型時的強大功能（如圖7所示）。這一