隨著城市交通運輸的發展與進步，運輸載具動力也逐漸由燃油動力轉變成電子電力方式驅動，例如：大眾捷運系統；而目前世界上電力驅動之大眾捷運系統有架空線及導電軌二種主要型式。本研究以導電軌型式之捷運系統為主，並選定國內營運超過20年且具代表性的台北捷運(MRT)做為研究對象。
台北捷運(MRT)電聯車以「集電靴」來接觸導電軌收集電力，使其維持運行，而導電軌往往為了配合現地調度因素及供電區間採取不連續設置，使集電靴於不連續處會有暫時脫離導電軌之情形，為了使集電靴能平順出入導電軌不連續處，不致因出入接觸不當影響電力供給效果，故於進出不連續處之二端設置端部組件，然而導電軌長時間與電聯車集電靴接觸運行後，也就造成其接觸表面產生磨損，減弱電力傳輸效果、對行車供電造成嚴重的停滯、其影響甚鉅，而最容易產生磨耗處又以導電軌端部較為頻繁，另外探究其影響磨耗速率因素眾多，不外乎包含行車速度、集電靴與導電軌端部間的接觸磨擦力、軌道線型以及導電軌電弧侵蝕等。
台北捷運(MRT)以往針對導電軌端部磨耗並無相關的研究及預測方式，造成後續備品採購、維修調度以及導電軌壽命及磨耗預測均欠缺憑據，故本研究係以導電軌端部實際磨耗數據及趨勢，運用「算術平均法」及「二次曲線迴歸法」發展出一套正確預測方式，並可進一步推展應用至台北捷運(MRT)路網，以提升導電軌端部採購及更換使用壽命之準確度，俾利妥善安排維修調度能量。
本研究的實驗主要可分為三個部份：首先以專門量測工具及儀器對台北捷運(MRT)導電軌全線磨耗率較高共50處之導電軌端部量取導電軌端部磨耗數據若干筆做為研究，再者將磨耗數據代入「算術平均法」及「二次曲線迴歸法」推估未來特定時間點磨耗值，其中「二次曲線迴歸法」可繪製導電軌端部磨耗趨勢線，藉由趨勢線判讀了解其磨耗速率變化情形。
最後將二種方法推估磨耗值與實際磨耗值相互比較其準確率並實施結果交叉比對；由結果來看，二種方法預測之準確率相當，另外於研究過程中發現影響準確率最大的因素為量測資料筆數，代入之資料筆數越多，準確率越高。
綜上、本研究不但歸納出一套導電軌端部磨耗預測流程外，更可針對全線1408處端部建立專屬的磨耗預估方程式，並實際應用於台北捷運(MRT)各處導電軌端部位置。

The driven power of transportations is gradually changed by fuel power into electrical power in order to encourage and advance the urban transportation developmentsuch as mass transit system.The electrical power-driven mass transit systems can be divided into two types around the world. One is overhead lines (Upper) and the other is with conductive rail (Lower). This research is focusedon the conductive rail type in MRT system, and the representative Taipei MRTis selected because ofits operation experience is more than 20 years.
Taipei MRT trains are all used "current collecting shoe" to contact the conductive rail in order to collect electricity.The conductive railsare often discontinuous because of the site factors, so that the current collectingshoes are not contacted with conductive rail in these discontinuity areas. In order to provide the current collecting shoes running smoothly and keep power supply working in the discontinuity, the high speed rampsare provided at the two ends of every discontinuity. However, it will cause the contact surface wearing, weaken the power transmission, and a serious stagnation after long timeoperation.The most prone to wear is at the high speed ramps of the conductive rails.There are also other factors affect the wearing rate of high speed ramps which are including the train speed, the contact friction between current collectinghoe and conductive rails, track linear and electric corrosion.
There is no relevant research and forecasting methods for the high speed ramps wearing rates of the conductive railsin Taipei MRT (MRT).It might cause the lack of credentials for the spare parts procurement, maintenance scheduling and conductivity track life and wear prediction. In this research, we use the "Arithmetic mean formula" and "Regression quadratic equation" to implement the wearing forecast of the high speed ramps. Moreover, it could be further applied to the Taipei MRT (MRT) network to enhance the procurement, and accuracy wearing rates of high speed ramps in order toprovide the better arrangements for maintenance schedule.
The experiments of this research areconsisted of three parts: First, we use particular measurement instruments to measure the 50 higher wearing rates of the high speed ramps in Taipei MRT network.Then, we use the "Arithmetic mean formula" and "Regression quadratic equation" to implement the wearing forecast of thesehigh speed ramps. At last, we use "Regression quadratic equation" to produce wear Trend line for these high speed ramps.According to these trend lines, we could predict the wearing rate of each high speed ramp.
Finally, we use the results fromthese two methods to compare with the actual wearingrate of high speed ramps. The results show that the accuracy of the two methods is similar. In the course of the research, the factor that has the most important impact rate is the accumulated of measurement data. When the measurement data is accumulated a great quantity, the forecast accuracy rate is raising indeed.
In summary, this research is not only summed up a set of high speed ramps of the track wear prediction process, but also for the 1408 high speed ramps of conductive rail to establish a wear prediction equation in order to apply for the Taipei MRT’s high speed ramps of the conductive rails.

[1] Ciandella, D. R. 1996. “Human side of L.A. Metro.”Civil Engineering ,v 66(12)：36-38.
[2] Mahlke, D.2005. “Aluminum-steel third rails - Important components of the Berlin City Railway third-rail system.”eb – Elektrische Bahnen ,v103：83-91.
[3] Janetschke, K., Freidhofer, H., Mier, G. 1982. “Testing of a New Third Rail Construction with Aluminum Third-Rail Interconnection in the Berlin Underground. ”eb – Elektrische Bahnenv 80(1)：17-23.
[4] Jinfa, G.,Jiqin, W., Yuan, Z. 2004. “Dynamics analysis of electric shoe gear and conductor rail system.”Journal of Vibro engineering ,v16(4)：1992-2007.
[5] Mahlke, D. 2001. “Wear-resistant stainless-steelthird-railstop” Eisen bahningenieur ,v52：66-67.
[6] Stewart, E. Weston, P. Hillmansen, S. and Roberts, C. 2011. “Using bogie-mounted sensors to understand the dynamics of third rail current collectionsystems”Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit ,v225：219-227.
[7] Yang, L.F., Guo, C., Cheng, Y. and Liu, H. W. 2004“New technologies of steel aluminum composite conductor rail.” Zhongguo Tiedao Kexue/China Railway Science ,v25( 6)：103-108.
[8] Dong, L., Li, F., Chen, G., and Zhou, Z. 2006.“Friction and wear of steel aluminum composite conductor rail/collector shoe with and without direct current.”Lubrication Engineering(6)：36-38.
[9] Deutzer, M. 2009.“Measurement of geometry, wear and current on third rail installations.”Lage-Verschlei-Und Stromübergangs Messungenan Strom schienen ,v107：530-537.
[10] T.K. Sheiretov, H.K.Yoon,and C. Cussno. 1998. “Scuffing under dry sliding condition：Ⅰ Experimental studies,”Tribology Transactions ,v41：435-446.
[11] Fracchia, M., Hill, R.J.,Pozzobon, P. and Sciutto, G. 1994.“Accurate track modelling for fault current studies on third-rail metro railways.”Proceedings of the 1994 ASME/IEEE Joint Railroad Conference：97-102.
[12] Taipei Rapid Transit Corporation. 2017. The high capacity conductive rail facilities maintenance standard operation process ,v26. Taipei City.
[13] Taipei Rapid Transit Corporation. 2017. The high capacity conductive rail facilities emergent combination maintenances standard operation process ,v2. Taipei City.
[14] Bhaumik, D. K., Gibbons, R. D. 2004.“An upper prediction limit for the arithmetic mean of a lognormal random variable.”Technometrics ,v46(2)：239-248.
[15] Kim, H.J. 2007. “A Monte Carlo method for estimating prediction limit for the arithmetic mean of lognormal.”Communications in Statistics - Theory and Methods ,v36(11)：2159-2167.
[16] Karaca, Z.,Yilmaz, N. G. and Goktan, R.M. 2008.“Comparison of averaging procedures for point load Testing of rock.”Geotechnical Testing Journal ,v 31(5)：456-460.
[17] Afzal, N.2007.“Friction factor directly from transitional roughness in a turbulent pipe flow.”Journal of Fluids Engineering, Transactions of the ASME ,v129(10)：1255-1267.
[18] Liu, H., Jin, J., Cao, S. and Ge, S. 2011.“The quantitative analysis of surface roughness in the dry friction”CailiaoYanjiuXuebao/Chinese Journal of Materials Research ,v25(5)：483-488.
[19] Marcus, A.H. 1974.“Theoretical prediction of highway noise fluctuations”
Journal of the Acoustical Society of America ,v56(1)：132-136.
[20] Eslami, H.2001.“Equation of state for nonpolar fluid mixtures: Prediction from boiling point constants.”International Journal of Thermophysics ,v22(6)：1781-1793.
[21] Tian,S.,Lu,W., Hui,T.,Cai,N. and Yang,Y. 2010.“Comparing evaluated precision of straightness error among twospot, least square method and minimum envelope zone method.”2nd IITA International Conference on Geoscience and Remote Sensing, IITA-GRS, 67-70,2010.
[22] Sum,J.andHo,K. 2006. “On-line estimation of the final prediction error via recursive least-squares method.”Neurocomputing ,v69：2420-2424.
[23] Luo, C.、Zhang, Z.、Mei.W.andDeng,Y. 2007.“A damped least square robust estimation method for spatial three-dimensional rectangular coordinate transformation.”Geomatics and Information Science of Wuhan University ,v32：707-710.
[24] Zhao, Z.Y.,Sagara,S. and Wada,K.1991.“Bias-compensating least squares method for identification of continuous-time systemsfrom sampled data.”International Journal of Control ,v53：445-461.
[25] Anderson,J. M.M.、Mair, B.A.、Rao, M.and Wu, C.H. 1995.“Weighted least-squares method for PET.”IEEE Nuclear Science Symposium & Medical Imaging Conference ,v2：1292-1296.
[26] Zhang, Y.C. and He, X. 1989.“Analysis of free vibration and buckling problems of beams and plates by discrete least squares method using B5-spline as trial function.”Computers and Structures ,v31：115-119.
[27] Vahabi, M.、Mehdizadeh,E、Kabganian, M. and Barazandeh, F. 2010.“Identification of an IPMC actuator model using incorporating a nonlinear with linear leastsquares method.”ASME 2010 10th Biennial Conference on Engineering Systems Design and Analysis, ESDA2010 ,v5：171-176.
[28] Feng,Y. S., andBen,A. 1996.“Modeling rail fatigue behavior with multiple hazards.”Journal of Infrastructure Systemsv2(2)：73-82.
[29] U. Olofsson and T. Telliskivi. 2003. “Wear plastic deformation and friction of two rail steels-A full-scale test and a laboratory study,”Wear ,v 254：80-93.
[30] Y.Jin,M. Ishida and A. Namura. 2011. “Experimental simulation and prediction of wear of wheel flange and rail gauge corner,”Wear ,v 271：259-267.
[31] Cotton, I. andAylott, P.2008. “A simulation tool to predict the impact of soil topologies on coupling between a light rail system and buried third-party infrastructure.”IEEE Transactions on Vehicular Technology ,v57(3)：1404-1416.
[32] Park, Y.J.andStone, D.H. 1978.“FAILURE ANALYSIS OF RAILS FROM FACILITY FOR ACCELERATED SERVICE TESTING.”Proceedings, Technical Conference-American Railway Engineering Association ,v79：413-446.
[33] Lombaert, G.,Galvín, P., François, S. and Degrande, G. 2014.“Quantification of uncertainty in the prediction of railway induced ground vibration due to the use of statistical track unevenness data.”Journal of Sound and Vibration ,v 333(18)：4232-4253.
[34] P.Rodriguez, B. Suarez, J. A. Chover, J. Terron and J.D. Sanz. 2011. “Pantograph wear assessment in over head conductor rail systems.”Proceedings of the 22nd International Symposium on Dynamics of Vehicles on Roads and Tracks, Manchester, U.K, 14-19.August 2011.
[35] C.Laurent, J. P. Massat, T.M.L. Nguyen-Tajan, J.P. Bianchi. andE. Balmes. 2013. “Pantograph Catenary Dynamic Optimisation Based on Advanced Multibody And Finite Element Co-simulation Tools”International Journal of Vehicle Mechanics and Mobility Volume 52, 2014 - Issue sup1: IAVSD Proceeding Supplement.
[36] L. Dong, G. X. Chen,M. H. Zhu, and Z.R. Zhou.2007. “Wear mechanism of -stainless steel composite conductor rail sliding against collector shoe with electric current.”Wear 263：598–603.
[37] T. Ding, W. Xuan, Q.He, H. Wu, and W.Xiong.2014.”Study on Friction and Wear Properties of Pantograph Strip/Copper Contant Wire for High-Speed Ttain.”The Open Mechanical Engineering Journal 8(1):125-128· April 2014.
[38] F. J. González,J. A. Chover,B. Suarez, M. Vázquez.2008. “Dynamic Analysis using Finite Elements to Calcúlate the Critical Wear Section of the Contact Wire in Suburban Railway Overhead Conductor Rails. “Journal of Rail and Rapad Transit, 2008 ,v222：145-157.
[39] L. Dong, Y. Zhu, H. Wang, H. Jiang, and F. Zheng. 2011. “The Effect of Normal Force on the Coupled Temperature Field of Metal Impregnation Carbon/Stainless Steel under the Friction and Wear with Electric Current.”Procedia Engineering 16(2011)：414-417.
[40] Guan,J. and Zhong, Y.2015. “Testing and simulation of the conductor rail dynamic system”Proceedings of SPIE - The International Society for Optical Engineering ,v 9：794.
[41] Dong, L., Chen, G.X., Zhu, M.H. and Zhou, Z.R. 2007.“Tribological characteristics between third rail and collector shoe under electric”MocaxueXuebao/Tribology ,v27(3)：274-278.
[42] Shih, W.L., Chen, N. andTzeng, Y.S.2008.“Electrical transients analysis for conductor rail gaps of Taipei rapid transit system”Proceedings of the ASME/IEEE/ASCE Joint Rail Conference, JRC：343-350.
[43] J.A. Chover, B. Suarez and P. Rodriguez, 2011. “Simulation Techniques for design of overhead Conductor rail lines for speeds over 140 Km/h”.Proceedings of the 22nd International Symposium on Dynamics of Vehicles on Roads and Tracks, Manchester, U.K, 14-19.August2011.
[44] A. M. Zarembski. 2005.The Art and Science of Rail Grinding. New York：Simmons-Boardman Books Company.
[45] J.Smulders. 2003. “Management and Research tackle rolling contact fatigue,”Railway Gazette International：433.
[46] Railtrack PLC. 2001.Rolling Contact Fatigue in Rail; A Guide to Current Understanding and Practice. Britain：Railtrack PLC paper .
[47] G.Thelen,andM.Lovette. 1996. “A parametric study of the lubrication transport mechanism at the railwheel interface,”Wear ,v191：113-120.
[48] D. Rippeth, J. Kalousek, and J. Simmons. 1996. “A case study of the effect of
lubrication and profile grinding on low rail roll-over derailments at CSX transportation,” Wear ,v191：252-255.
[49] M. Ishidaa,T. Ban, K. Iida, H. Ishida and F. Aoki. 2008. “Effect of moderating friction of wheel/rail interface on vehicle/track dynamic behavior,”Wear ,v265
：1497-1503, 2008.
[50] T.M.Beagley,. 1976. “Severe Wear of Rolling Sliding Contacts,”Wear ,v36：317-315.
[51] H.Yoon, T.Sheiretov, and C. Cusano.2000. “Scuffing behavior of 390 aluminum steel under starved lubrication conditions.”Wear, vol. 237：163-175.
[52] H. Yukeng, D. Chen, and Linqing Zheng.1985. “Effect of surface topography of scraped machine tool guideways on their tribological behavior.”Tribology International ,v18, No.2：125-129.
[53] P. C. Maria, and C.Cris. 2000. “Running-in aluminum/steel contacts under starved lubrication PartⅡ：Effects on scuffing ”Wear, v242：133-139.
[54] I. Etsion, G. Halperin, and E. Becker. 2006. “The effect of various surface treatments on piston pin scuffing resistance” Wear, v261：785-791.
[55] O.D. Tasbaz, R.J.K. Wood, M. Browne, H.E.G. Powrie , G. Denuault ,1999, “Electrostatic monitoring of oil lubricated sliding point contacts for early detection of scuffing” Wear, v230：86-97.
[56] M. W. Bailey, A. Cameron. 2008. “The Effect of Temperature and Metal Pairs on Scuffing” ASLETransactions, v16(2)：121-131.
[57] S.Wu, H.S. Cheng. 1993.”Sliding wear calculation in spur gears,” J. Tribol115(3)： 493-500.
[58] A. Flodin, S. Andersson. 1997. “Simulation of mild in spur gears,” Wear ,v 207：16-23.
[59] E. Rabinowicz. 1995. Friction and wear of materials. New York：Wiley.
[60] K. C. Ludema. 1996. Friction, wear, lubrication. Boca Raton：CRC Press.
[61] F. P. Bowden, D. Tabor. 1954.Friction and lubrication of solids, Part I. Oxford：Clarendon Press.
[62] N. Ohmae, and E. Rabinowicz. 1980. ”The wear of the noble metals,”ASLE Trans：86-92.
[63] G. A. William. 1992. Materials for tribology, , New York：Elsevier Science.
[64] T.K. Sheiretov, H.K.Yoon,and C. Cussno. 1998. “Scuffing under dry sliding condition：ⅡTheoretical studies,”Tribology Transactions v41：447-458.
[65] Talotte, C., van der Stap, P., Ringheim, M.,Dittrich, M., Zhang, X. and Stiebel, D. 2006. “Railway source models for integration in the new European noise prediction method proposed in Harmonoise,”Journal of Sound and Vibration v 293(3)-(5)：975-985.
[66] Guigou-Carter, C.,Villot, M., Guillerme, B. and Petit, C.2006. “Analytical and experimental study of sleeper SAT S 312 in slab track Sateba system,”Journal of Sound and Vibrationv293(3)-(5)：878-887.
[67] Rathod, C., Shabana, A.A. and Nonlinear Dyn. 2007. “Geometry and differentiability requirements in multibody railroad vehicle dynamic formulations,”Nonlinear Dynamicsv 47(1)-(3)：249-261.
[68] Chang, K.C. and Payne, U.J. 1992. “Numerical treatment of diffusion coefficients at interfaces,”Numerical Heat Transfer; Part A: Applications v 21(3)：363-376.
[69] A.Farsakh, Murad Y.T. and Hani H. 2004. “Assessment of direct cone penetration test methods for predicting the ultimate capacity of friction driven piles,”Journal of Geotechnical and Geoenvironmental Engineering v 130(9)：935-944.
[70] Kakosyan, A.V. 1979.“Theory of estimation of parameters in a linear regression scheme,”Journal of Soviet Mathematicsv12(2)：227-237.
[71] Cressie, N. 1978. “Estimation of the integral of a stochastic process.”BULL. AUSTRAL.MATH.SOC.v18：83-93.
[72] Piegorsch, W.W. 1986. “Confidence bands for polynomial regression with fixed intercepts.”Technometrics v28(3)：241-246.
[73] Kabe,D.G.1986. “Direct proof of a result of strandand west water for the numerical solution of a FREDHOLM integral equation of the firstkind.”Industrial Mathematics v36(2)：181-187.
[74] Bain, J.A. 1973.“Dynamics of high speed sliding power collection systems for electrically propelled vehicles.”American Society of Mechanical Engineers：9.
[75] Amin,Md.F.,Savitha,R.,Amin,M. I., and Murase,K. 2011.“Complex-valued functional link network design by orthogonal least squares method for function approximation problems.”Proc Int Jt Conf Neural Network：1489-1496.