Open Access Analysis of Monitoring System Reliability for Wind Turbine Based on Wireless Sensor Network

Transcription

1 Send Orders for Reprins o reprins@benhamscience.ae 21 The Open Auomaion and Conrol Sysems Journal, 214, 6, Open Access Analysis of Monioring Sysem Reliabiliy for Wind Turbine Based on Wireless Sensor Nework Fu Zhixin 1, Liu Zhen *,1, Ding Guorong 1, Li Hanbing 2, Zhao Min 1, Li Qiaomu 1 and Yuan Yue 1 1 College of Energy and Elecrical Engineering, Hohai Universiy, 8 Focheng Wes Road, Jiangning Disric, Nanjing, Jiangsu, 2111, China 2 Sae Grid Lianyungang Power Supply Company, 1 Xingfu Norh Road, Xinpu Disric, Lianyungang, Jiangsu, 222, China Absrac: In his paper, based on wireless sensor nework echnology, we propose he monioring sysem for wind urbine. In order o ensure ha he monioring sysem can compleely and accuraely obain he condiion monioring informaion, using exponenial disribuion funcion, we esablish he reliabiliy model. Because he opology conrol echniques can improve he monioring sysem reliabiliy, we presen he hierarchical nework opology for monioring sysem. Then, combined wih he presened hierarchical nework opology, we sudy he reliabiliy model. Finally, simulaion shows ha he nework parameers affec he reliabiliy and he working ime of monioring sysem. Keywords: Wind urbine, monioring sysem, reliabiliy, wireless sensor nework, nework parameers. 1. INTRODUCTION As a kind of clean and renewable energy power generaion, wind power can no only effecively solve deerioraion problems of ecological environmen, bu also mee adjusmen demands of energy srucure. Therefore, wind power has developed rapidly [1]. However, wind urbine is normally locaed in remoe area wih harsh environmen and climae variabiliy. Moreover, is srucure is very complicaed. Wind urbine is easy o be damaged, which causes operaion fauls. Thus, i is very necessary o monior reliably wind urbine condiion. Tradiional regular mainenance and faul mainenance lack he mechanism o grasp operaion saus of wind urbine. They have drawbacks of high cos and low faul repair rae [2]. The maurer exising monioring sysem of wind urbine uses wired monioring echnology having problems of high-cos invesmen, difficul equipmen insallaion and so on [3]. The condiion monioring sysem based on wireless sensor nework (WSN) is a kind of low-power, low-cos, self-organizaion, and miniaurizaion monioring sysem. The WSN-based monioring sysem of wind urbine (WSN- MSWT) can solve problems of wired monioring and i can provide real ime condiion monioring for wind urbine [4]. In recen years, wih he capaciy increase of wind urbine, i is more difficul o carry ou he operaion mainenance. As a resul, he WSN-MSWT has received widespread aenion. Nodes deploymen for wind urbine is he firs problem o solve [5]. And how o se up he nework planning of wind urbine monioring based on WSN makes sense [6]. Addiionally, he monioring sraegy should consider he mechanical ransmission sysem of wind urbine [7]. In [7], auhors design he hardware and sofware of monioring sysem, and gives he hardware srucure of nodes. Furhermore, in [8], auhors propose he approach in sensing wind speed for wind energy using WSN echnology. Moreover, according o Spacek [9], monioring mechanical vibraion in wind urbines is very imporan, bu i ignores he real nework opology of WSN. The WSN-MSWT operaes normally in bad environmen. In order o obain compleely and accuraely he condiion monioring informaion of wind urbine, i is necessary o analyze he monioring sysem reliabiliy. We can improve he monioring sysem reliabiliy considering he nework lifeime [1]. Topology conrol echniques can exend lifeime of WSN applied is smar grid [11]. Bu auhors do no deermine wheher opology echniques can influence he monioring sysem reliabiliy. Besides, we should ake ino accoun he sysem average working ime o evaluae he monioring sysem reliabiliy [12]. According o [12], oal amoun of subsysems and nodes influence he monioring sysem reliabiliy. When he WSN-MSWT runs, he nodes in each cluser will separaely collec and process condiion informaion of wind urbine, hen upload he informaion o he cluser head node. The cluser head node can communicaion wih oher cluser head nodes, and ransmi he monioring informaion o base saion. This paper presens he opological srucure of he monioring sysem for wind urbine. Furhermore, we esablish he reliabiliy model and research he monioring sysem reliabiliy. Finally, simulaion analysis shows he influence of nework opology on he monioring sysem reliabiliy / Benham Open

2 Analysis of Monioring Sysem Reliabiliy for Wind Turbine The Open Auomaion and Conrol Sysems Journal, 214, Volume Fig. (1). Hierarchical opology of monioring sysem. Cluser Head Node Node of Base Saion 2. NETWORK TOPOLOGICAL STRUCTURE OF MONITORING SYSTEM 2.1. Selecion of Nework Topology Topological srucure of WSN is divided ino hree kinds: sar, mesh and hybrid ype [13, 14]. Sar opology has he simple single-hop srucure. This opology consumes small energy, bu is communicaion reliabiliy is poor. Mesh opology is complex because of he muli-hop srucure. This opology has srong communicaion robusness while is energy consumpion is very high. Hybrid opology is mixure of sar opology and mesh opology. I is characerized by is small energy consumpion and srong flexibiliy. Among hese opologies, he hierarchical opology is a ypical hybrid opological srucure. In he hierarchical opology, sensor nodes are divided ino several independen clusers, and cluser head node should be seleced by a cerain algorihm. The cluser head node is responsible for receiving monioring informaion from oher nodes in he cluser. Afer being processed, he monioring informaion will be sen o base saion. The opological nework srucure needs o realize independen monioring for differen componens of wind urbine, and differen componens can communicae wih each oher. Thus, he opology of monioring sysem should be hierarchical. Based on above analysis, he opological srucure of WSN-MSWT should adop he mixed nework opology o have he hierarchical monioring (Fig. 1). The nework opology of monioring sysem in his paper is he mixure of sar opology and mesh opology, wherein each inner subne uses he mesh nework, and nework beween he subne and he base saion adops he sar nework. The monioring sysem consiss of m subnes, each subne is composed of n nodes. Each subne operaes individually, uninfluenced by oher subnes. Table 1 corresponds o he disribuion of subnes in various posiions of wind urbine. Table 1. Disribuion of subnes in wind urbine. Posiion Blade TRANSMISSION SYSTEM Engine room subnes amous m 1 m 2 m 3 In Table 1, mm 1 +m 2 +m 3. In acual deploymen, values of m1, m2, m3 and n are chosen depending on he specific shape of each componen in wind urbine Selecion of Monioring Poins We should deermine he hierarchical nework opology of WSN-MSWT depending on specific circumsances of monioring poins in wind urbine. The working environmen of wind urbine is relaively bad, wind urbines easily break down. Among componens in wind urbine, he occurring fauls in ransmission sysem have he highes percenage. Therefore, under he premise of ensuring he reliabiliy of WSN-MSWT, i is imporan o sudy sensor nodes deploymen in he ransmission sysem [15]. Vibraion parameers are characerisic quaniy. They can more accuraely reflec operaion saus of ransmission sysem in wind urbine, hey are also easy o be deeced. Thus, his paper deploys sensor nodes in monioring poins and in heir nearby posiions o monior and o ge vibraion informaion of wind urbine. Then, he WSN-MSWT can diagnose he operaion condiion of wind urbine. The disribuion of monioring poins in wind urbine is illusraed in Fig. (2). We selec six monioring poins (Fig. 2): monioring poin is responsible for monioring spindle bearing condiion. Monioring poins and are respecively responsible for he inpu and oupu vibraion monioring of he spindle bearing. Monioring poin is responsible for monioring planeary gear of gearbox. Monioring poin is responsible for monioring inpu end bearing of engine.

3 212 The Open Auomaion and Conrol Sysems Journal, 214, Volume 6 Zhixin e al Hub Spindle Bearing Gearbox Engine Engine Room Fig. (2). Disribuion of monioring poins in wind urbine. Monioring poin is responsible for monioring free end bearing of he engine. In addiion o he above six monioring poins, a monioring poin should be pu in cabin deck used for vibraion condiion monioring of engine room and conneced o he ower. Moreover, we should pu a monioring poin in blade roo ha is easy o break down. We use his monioring poin o gain blade vibraion informaion. Obviously, amoun of he monioring poins for ransmission sysem should be larger, and monioring differen posiions for ransmission sysem is more imporan. For he purpose of comprehensive informaion acquisiion of monioring poins for ransmission sysem, and of he reliabiliy of communicaion beween subnes and base saion, he ransmission sysem is composed of more han one subne, namely m 2 represens a se of subnes. We will discuss below abou how o combine monioring needs of ransmission sysem and how o guaranee he reliabiliy of he consruced subne sysems. 3. RELIABILITY MODEL OF MONITORING SYSTEM 3.1. Concep and Definiion of Sysem Reliabiliy (1) Sysem Reliabiliy Reliabiliy is a kind of probabiliy: in he specified condiions and wihin he sipulaed ime, he probabiliy of compleing prescribed asks [16]. We usually denoe reliabiliy index funcion as R(), called as reliabiliy funcion. The consan T represens ime for research objec from normal sae o faul sae, namely he objec lifeime. The reliabiliy funcion can be expressed as: ( ) P{T } R > (1) Equaion (1) means he probabiliy for an individual or a sysem of no breaking down beween (, ). (2) Cumulaive Disribuion Funcion The cumulaive disribuion funcions refers o he unreliabiliy ha means he probabiliy of no compleing asks under specified condiions and wihin prescribed ime period for hydraulic producs. The cumulaive funcion F() is generally presened by: ( ) P{T } R (2) Because he compleed even and uncompleed even wihin he prescribed ime are a pair of opposie even, he following equaion can be obained: ( ) 1 R( ) F (3) Equaion (3) describes he probabiliy of faul for an individual or a sysem beween (, 1). (3) Probabiliy Densiy Funcion of Faul The probabiliy densiy funcion of faul f() is a derivaive of unreliabiliy, namely d F ( ) f ( ) d (4) Faul Rae The faul rae λ refers o, in any ime, he probabiliy of faul occurrence of an individual in a uni of ime. Assuming ha he sudying individual number is N, faul number afer is r(). Then faul number afer + Δ is r( + Δ ), so he faul rae can be expressed by: λ λ R ( ) ( ) ( ) ( ) r + Δ r() [ N r( )] Δ When N and Δ, ake limi of Equaion (5): ( ) r + Δ r() r( + Δ ) r() lim lim N Δ N [ N r( )] Δ N N r() Δ Δ N The esimaion for reliabiliy is shown in Equaion (7): N r() N Expression for he probabiliy densiy esimaion is given in Equaion (8): f ( ) λ ( ) ( ) r + Δ r() N Δ Solve equaions from (6) o (8): f() R () (4) (5) (6) (7) (8) (9)

4 Analysis of Monioring Sysem Reliabiliy for Wind Turbine The Open Auomaion and Conrol Sysems Journal, 214, Volume From equaions (3) and (4), Equaion (1) is as follows: d F ( ) d R ( ) f ( ) (1) d d When Equaion (1) is subsiued ino Equaion (9), he following equaion can be obained: 1 d R( ) λ ( ) R () d (11) Take inegral for eiher side of he equals sign in Equaion (11): λ ( ) d lnr( ) + ln R() (12) In he case of knowing λ() and R(), we can solve he normal working ime. (5) Average Normal Working Time (ANWT) In he reliabiliy sudy of an individual or a sysem, he ANWT is a very imporan characerisic quaniy, we can use i direcly o reflec he lifeime of sudy objec. In he experimenal sudy of acual sample, he normal working ime from he firs sample o he n h sample is presened as 1, 2, 3 n, assuming ha here are n samples: n 1 ANWT (13) i n i 1 When he ANWT is generalized o an absrac sysem, i can be presened by he expecaion of normal working ime of monioring sysem: ANWTE( ) R( )d (14) 3.2. Reliabiliy Model of Sensor Nodes To describe he sysem reliabiliy, here are usually models of binomial disribuion, exponenial disribuion, Gamma disribuion, Weil-bull disribuion, normal disribuion and so on. Among hem, he exponenial disribuion is a ype of single parameer disribuion, and i has good adapabiliy. Therefore, o esablish he reliabiliy model of sensor nodes, we apply he exponenial disribuion model, namely nodes lifeime is subjec o exponenial disribuion wih faul rae λ. Then he reliabiliy model can be expressed as: λ R () 1 F () e, > (15) When Equaion (15) is subsiued ino Equaion (14), he following equaion can be obained: ANWT λ e d 1 λ (16) Usually he value of ANWT is aken as 1 6 hours. When i is subsiued ino Equaion (16), he reliabiliy funcion of sensor nodes in he monioring sysem for wind urbine is: 6 1 R () e, > (17) 3.3. Reliabiliy Model of Monioring Sysem We can decide he reliabiliy of he WSN-MSWT by he reliabiliy of each subne of monioring sysem. Considering impac of each subne on he reliabiliy of he whole monioring sysem, he sysem can be divided ino series sysem, parallel sysem, hybrid sysem and voing sysem. Among hese four kinds of sysems, in he series sysem, if faul occurs in any subne sysem, he whole sysem will malfuncion. In he parallel sysem, if faul occurs in all subne sysems, he whole sysem will malfuncion. Hybrid sysem is a mixure of series sysem and parallel sysem. In he voing sysem, as long as he number of subne sysems is no less han a cerain consan, he whole sysem will no have fauls. Each subne sysem in he WSN-MSWT works alone. Therefore, i necessary o sudy reliabiliy of independen subne sysems before sudying reliabiliy evaluaion model of whole monioring sysem of wind urbine. This paper uses mesh srucure for each subne sysem. For a mesh nework conaining n sensor nodes, assuming ha he nework can olerae f max nodes, so he voing sysem can be used o sudy he reliabiliy of subnes. The reliabiliy of differen operaion saus of nework is shown in he following: (1) n nodes work normally. The nework reliabiliy is: ( ) ( ) R [ R ] n (18) 1 (2) 1 node has fauls and n-1 nodes work normally. The nework reliabiliy is: n 1 n 1 ( ) ( ) ( ) R2 C n R [1 R ] (19) (3) 2 nodes have fauls and n-2 nodes work normally. The nework reliabiliy is: n n 2 ( ) ( ) ( ) R C R [1 R ] n (2) (4) f max nodes have fauls and n-f max nodes work normally. The nework reliabiliy is: n fmax max ( ) ( ) ( ) Rj C R [1 R ] n f fmax n (21) (5) f max +1 nodes have fauls, he nework canno work normally. The nework reliabiliy is: R j+1. Therefore, when he nework can olerae f max nodes, he reliabiliy of each subne is: j i i 1 i n fmax n i ( ) R ( ) i Rc ( ) R ( ) Cn R [1 ] n i (22) For he WSN-MSWT composed of m subnes, subnes form he sar opology nework. To le he enire nework operae normally, each subne ha works alone should operae wihou fauls. This paper uses he voing sysem o sudy he reliabiliy model. The reliabiliy model of he voing sysem can be presened in he following:

5 214 The Open Auomaion and Conrol Sysems Journal, 214, Volume 6 Zhixin e al. Table 2. Deploymen of sensors in WSN-MSWT. Nework parameers Subnes m Subne nodes n Faul-oleraed nodes f max Toal nodes k Mode Mode Mode Mode Table 3. Time of differen reliabiliy in differen mode. (uni: 1 4 h). Reliabiliy Mode Mode Mode Mode R( ) [R c ()] m (23) 4. SIMULATION AND ANALYSIS In his paper, amouns of subnes, nodes of subnes, sysem faul-oleraed nodes and sysem nodes are considered o be he nework parameers. Addiionally, he sysem reliabiliy and he sysem working ime are regarded as evaluaion indexes Reliabiliy of WSN-MSWT The nework parameers of WSN-MSWT are shown in Table 2. In Table 2, here are four kinds of nodes deploymens. The reliabiliy funcion R() can be obained by equaions (22) and (23)( Fig. (3)). Meanwhile, he ime can be calculaed as shown in Table 3. In Fig. (3), he curve of reliabiliy funcion of he monioring sysem have hree secions. In he firs secion, he reliabiliy is very high wih R() >.9. R() decreases slowly as he ime increases. In he second secion, R() decreases rapidly as he ime increases wih.1 < R() <.9. In he hird secion, he reliabiliy is very slow wih R() <.1. R() is approaching and he speed of decreasing becomes slower again. Comparing mode 1 wih mode 2, when he amouns of nodes in each subne and faul-oleraed nodes are unchanged, he sysem reliabiliy decreases obviously as he amoun of subnes increases. Therefore, he increasing amoun of subnes in he monioring sysem reduces he monioring sysem reliabiliy. Comparing mode 1 wih mode 3, when he amoun of subnes keeps unchanged and he amouns of nodes in subnes and faul-oleraed nodes increase, he sysem reliabiliy improves. However, he reliabiliy decreasing speed acceleraes. Therefore, if he subne amoun is unchanged for he ransmission sysem, increasing he nodes of subne will improve he monioring sysem reliabiliy. Monioring Sysem Reliabiliy Working Time /1 6 h Fig. (3). Reliabiliy funcion R() in differen mode. Mode 1 Mode 2 Mode 3 Mode 4 Comparing mode 1 wih mode 4, when he amouns of he sysem nodes and faul-oleraed nodes are unchanged, reducing appropriaely subne amoun and increasing nodes amoun of each subne can significanly improve he sysem reliabiliy Working Time of Monioring Nework According o he reliabiliy evaluaion model of monioring sysem, hen normal working ime of sysem can be calculaed. Take λ() λ 1-6 and R() 1 combining wih Equaion (23), Equaion (24) is obained as follows:

6 Analysis of Monioring Sysem Reliabiliy for Wind Turbine The Open Auomaion and Conrol Sysems Journal, 214, Volume Sar Inpu x, N k 1 Calculae f( x ) x x 1 f( x ) x1 x ' f ( x ) x 5 1 x < 1? Yes k +1 No No k N? Yes Inpu x1 Ineraion Failure Fig. (4). Flow char of newon ieraive mehod. Table 4. Sysem normal working hours of differen reliabiliy in differen mode. End Nework Parameers Mode 1 Mode 2 Mode 3 Mode 4 Sysem normal working ime (hour) Sysem normal working ime (year) ln[ R( )] m (24) Since Equaion (24) is a ranscendenal equaion, use Newon ieraive mehod o solve Equaion (24). Le f () 1!6 + ln[r()] m, he Newon ieraion mehod is shown in Fig. (4). As shown in Table 4, he normal working ime of differen mode has he same rend wih he ime in Fig. (3). For furher sudying of influence of sysem nework parameers on sysem normal working hours, use mode 1 as an example. There are wo kinds of siuaions: (1) k and f max keep consan, and n changes The curve of he normal working ime wih changing ime is as shown in Fig. (5). I can be seen ha sysem working ime increases as amoun of subne nodes increases. Tha is because he nework of monioring sysem adops he mixure of sar and mesh opology, wherein he each inner subne uses mesh opology while he nework beween subnes

7 216 The Open Auomaion and Conrol Sysems Journal, 214, Volume 6 Zhixin e al. Sysem Normal Woring Time/hour Fig. (5). Sysem normal working hours changing wih n. Amoun of Subne Nodes Sysem Normal Woring Time/hour Amoun of Enire Sysem Nodes Fig. (6). Sysem normal working hours changing wih k. and base saion uses sar opology. As nodes amoun of sysem does no change, subne amoun decreases while amoun of each subne nodes increases. Then he mesh opology srucure wih srong reliabiliy increases, he sar opology srucure wih weak reliabiliy decreases. Thus, he sysem normal working ime increases as n increases. In he exreme case, when n48 and m1, he sysem normal working ime reaches he maximum because he enire sysem nework has mesh opology. (2) n and amoun of faul-oleraed nodes in subne keep consan, and k changes The curve of he normal working ime wih changing k is as shown in Fig. (6). In conclusion, he sysem working ime decreases as amoun of sysem nodes increases. The amouns of subne nodes and faul-oleraed nodes keep unchanged, so he reliabiliy of each subne does no change. However, increasing k means increasing subnes amoun. As a resul, he sysem reliabiliy becomes weaker and sysem normal working ime decreases as more subnes in he enire sysem form he sar opology srucure. CONCLUSION This paper proposes he hierarchical nework opology of WSN-based condiion monioring sysem for wind urbine, and sudies he reliabiliy modeling and analysis. The simulaion resuls show he following conclusion:

8 Analysis of Monioring Sysem Reliabiliy for Wind Turbine The Open Auomaion and Conrol Sysems Journal, 214, Volume ) When he amouns of each subne nodes and fauloleraed nodes are unchanged, increasing he subne amoun reduces he reliabiliy of monioring sysem. 2) For monioring he imporan ransmission sysem, if he subne amoun keeps unchanged, he nodes amoun of subne will increase. Then, he sysem reliabiliy can be improved. 3) When he sysem nodes amoun and he faul-oleraed nodes keep unchanged, reducing appropriaely he subnes amoun means improving he sysem reliabiliy. CONFLICT OF INTEREST The auhors confirm ha his aricle conen has no conflics of ineres. ACKNOWLEDGEMENTS This work is suppored by he Naional Naural Science Foundaion of China (no ), he Naural Science Foundaion of Jiangsu Province of China (no.bk21249), Six Talen Peaks Projec of Jiangsu Province of China(no.214-XNY-8) and he Projec of China Three Gorges New Energy Corporaion. REFERENCES [1] C.X. Wang, Z.X. Lu and Y. Qiao, A Consideraion of he Wind Power Benefis in Day-Ahead Scheduling of Wind-Coal Inensive Power Sysems, IEEE Transacions on Power Sysems, vol. 28, no. 1, pp , 213. [2] Z.X. Fu, and Y Yue, Saus and Prospecs on Condiion Monioring Technologies of Offshore Wind Turbine, Auomaion of Elecric Power Sysem, vol. 36, no. 21, pp , 212. [3] W.X. Yang, P.J. Tavner, and e al, Cos-Effecive Condiion Monioring for Wind Turbines, IEEE Transacions on Indusrial Elecronics, vol. 57, no. 1, pp , 21. [4] J.A.R. Azevedo, and F.E.S. Sanos, Energy Harvesing from Wind and Waer for Auonomous Wireless Sensor Nodes, IET Circuis, Devices & Sysems, vol. 6, no. 6, pp , 212. [5] Z.X. Fu, M. Zhao,Y. Yuan, and e al, Consrucion Sraegies for Offshore Wind Turbine Condiion Monioring Sysem Based on Wireless Sensor Neworks, Auomaion of Elecric Power Sysem, vol. 38, no. 7, pp , 214. [6] C. Popeanga, R. Dobrescu, and N. Crisove, Smar Monioring and Conrolling of Wind Turbines Farms Based on Wireless Sensors Neworks, 212 1s Inernaional Conference on Sysems and Compuer Science (ICSCS), pp. 1-6, 212. [7] C.L. Hus, and T.Y. Hsu, The Pracical Design of Consrucing Daa Transiion Inerface wih ZigBee WSN-Example wih Small Scaled Wind-Power Elecriciy Generaor Sysem, Journal of Sofware, vol. 3, no. 8, pp , 28. [8] Y.K. Tan, and S.K. Panda, Self-auonomous Wireless Sensor Nodes wih Wind Energy Harvesing for Remoe Sensing of Wind-driven Wildfire Spread, IEEE Transacions on Insrumenaion and Measuremen, vol. 6, no. 4, pp , 211. [9] A.D. Spacek, O.H. Ando Junior, J.M. Neo, and e al, Managemen of Mechanical Vibraion and Temperaure in Small Wind Turbines Using Zigbee Wireless Nework, IEEE Lain America Transacions, vol. 11, no. 1, pp , 213. [1] C. Liu, S.Y. Zhang, H.B. Feng, and e al, The Research on Wireless Sensors Nework Reliabiliy, Proceedings of 28 Inernaional conference on Wireless Communicaions, Neworking and Mobile Compuing, pp. 1-4, 28. [11] A.A. Aziz, Y.A. Sekercioglu, P. Fizparick, and e al, A Survey on Disribued Topology Conrol Techniques for Exending The Lifeime of Baery Powered Wireless Sensor Neworks, IEEE Communicaions Surveys & Tuorials, vol. 15, no. 1, pp , 213. [12] C. Vasar, O. Prosean, I. Filip, and e al, A Reliabiliy Analysis for Wireless Sensor Neworks in a Wind Farm, XXII Inernaional Symposium on Informaion, Communicaion and Auomaion Technologies, pp. 1-5, 29. [13] M. Zorzi, and R.R. Rao. Geographic Random Forwarding (GeRaF) for Ad Hoc and Sensor Neworks: Muli-Hop Performance, IEEE Transacions on Mobile Compuing, vol. 2, no. 3. [14] I.F. Akyildiz, W. Su, Y. Sankarasubramaniam, and e al, Wireless Sensor Nework: A Survey, Compuer Neworks, vol. 38, pp , 22. [15] P. Caseliz, and J. Giebharg, Roor Condiion Monioring for Improved Operaional Safey of Offshore Wind Energy Converers, ASME Transacions, Journal of Solar Energy Engineering, vol. 127, pp , 25. [16] P.E. Labeau, and C. Smids, Dynamic Reliabiliy: Towards an Inegraed Plaform for Probabilisic Risk Assessmen, Reliabiliy Engineering and Sysem Safey, vol. 68, pp , 2. Received: Sepember 16, 214 Revised: December 23, 214 Acceped: December 31, 214 Zhixin e al.; Licensee Benham Open. This is an open access aricle licensed under he erms of he Creaive Commons Aribuion Non-Commercial License (hp://creaivecommons.org/licenses/by-nc/3./) which permis unresriced, non-commercial use, disribuion and reproducion in any medium, provided he work is properly cied.