Hindawi Jounal of Advanced Tanspotation Volume 2017, Aticle ID 8092718, 9 pages https://doi.og/10.1155/2017/8092718 Reseach Aticle Dopple Effect-Based Automatic Landing Pocedue fo UAV in Difficult Access Envionments Jan M. Kelne and Cezay ZióBkowski Institute of Telecommunications, Faculty of Electonics, Militay Univesity of Technology, Gen. Sylweste Kaliski St. No. 2, 00-908 Wasaw, Poland Coespondence should be addessed to Jan M. Kelne; jan.kelne@wat.edu.pl Received 23 June 2017; Accepted 14 August 2017; Published 4 Octobe 2017 Academic Edito: Seungjae Lee Copyight 2017 Jan M. Kelne and Cezay Ziółkowski. This is an open access aticle distibuted unde the Ceative Commons Attibution License, which pemits unesticted use, distibution, and epoduction in any medium, povided the oiginal wok is popely cited. Cuently, almost unesticted access to low-lying aeas of aispace ceates an oppotunity to use unmanned aeial vehicles (UAVs), especially those capable of vetical take-off and landing (VTOL), in tanspot sevices. UAVs become inceasingly popula fo tanspoting postal items ove small, medium, and lage distances. It is foecasted that, in the nea futue, VTOL UAVs with a high take-off weight will also delive goods to vey distant and had-to-each locations. Theefoe, UAV navigation plays a vey impotant ole in the pocess of caying out tanspot sevices. At pesent, duing the flight phase, dones make use of the integated global navigation satellite system (GNSS) and the inetial navigation system (INS). Howeve, the inaccuacy of GNSS + INS makes it unsuitable fo landing and take-off, necessitating the guidance of a human UAV opeato duing those phases. Available navigation systems do not povide sufficiently high positioning accuacy fo an UAV. Fo this eason, full automation of the landing appoach is not possible. This pape puts fowad a poposal to solve this poblem. The authos show the stuctue of an autonomous system and a Dopple-based navigation pocedue that allows fo automatic landing appoaches. An accuacy evaluation of the developed solution fo VTOL is made on the basis of simulation studies. 1. Intoduction Militay applications [1 3] wee the beginning of the development of unmanned aicaft systems. The incease in thei availability contibutes to the widespead use of unmanned aeial vehicles (UAVs) in civilian applications. In this case, monitoing lage aeas of land o sea is the main pupose. It concens such aeas of human activity as an agicultue [4, 5], enegetics (i.e., photovoltaic plants [6 8] and high voltage lines [9]), envionment potection [10], seach and escue [11], foesty and fie detection [12, 13], wate aea management [14, 15], and so on. An automation of monitoing pocedues, low costs, and minimization of human esouces in the UAV exploitation ae conducive to the dynamic gowth of thei use in the civilian applications. The UAV monitoing is mainly based on optical sensos, wheeas take-off and landing ae usually done in the same place. Cuently, almost unesticted access to low-lying aeas of aispace ceates an oppotunity to use UAVs in tanspot sevices. In this case, the place of the take-off and landing is fa emoved. This significantly hindes the implementation of navigation pocedues, especially at the landing stage. In this aticle, we pesent a method of automatic landing appoach that can be used especially in UAV tanspot sevices ove long distances. The advantage of this mode of tanspot is its independence of oad infastuctue, taffic volume, and difficult teain conditions. Hence, UAVs ae inceasingly used to tanspot long distances to small postal items and medicines in had-to-each aeas. A pactical example of such a solution, developed at the initiative of the United Nations Childen s Fund (UNICEF), is the use of UAVs to tanspot blood samples in Malawi (Afica) [16]. Nowadays, the tanspotation of blood at close distance between hospitals [17] o othe packages [18] is aleady achieved. In the futue, inceasing the load capacity of UAVs will enable fast tanspot sevices in had-to-each envionments such as islands, mountainous, deset, and pola aeas (i.e., the Actic and the Antactic). In this case, vetical take-off and landing (VTOL) UAVs will play a special ole, as they do not equie landing stips, but only small landing pads. Hence, VTOLs may be used to land at
2 Jounal of Advanced Tanspotation such locations as islands, oil platfoms, vessels, o skyscape oofs. The basic method fo navigating UAVs ove long distances is based on the use of a global navigation satellite system (GNSS) suppoted by the UAV s own inetial navigation system (INS) [19 22]. Howeve, these systems cannot be diectly used in the final stage of the flight, that is, duing landing, due to the low positioning accuacy of moving objects inheent in such systems. A pactical solution to this poblem is to use optical cameas, which allow the opeato to emotely contol the landing pocess. Howeve, this method equies the use of a boadband contol channel and may only be used duing daylight and in good visibility conditions. Unde night conditions and in poo visibility, a themal imaging camea [6, 23 25] o synthetic apetue ada (SAR) [26 28] may be used. Futhemoe, long distance wieless communications ae chaacteized by significant delays and, in the case of the had-to-each aeas, only satellite communications [29] can be used. In these cases, the navigation system equies a lage and expensive extension, which pevents its commecial use. High costs and, above all, the necessity to change the destination location (landing aea) fo the tanspot also pevent the use of conventional landing navigationsystems[20,21,30,31]suchastheinstument landing system (ILS), micowave landing system (MLS), o local aea augmentation system (LAAS). In this case, the opeational ange fo these systems is limited to the space aound lage aipots. The high pecision equied fo detemining the cuent position of the object and the equied flexibility in tems of landing spots limit the ability to use UAVs fo ai tanspot. Theefoe, the study of a pecise and simple positioning method,whichwouldgivetheuseflexibilityintemsof landing locations and conditions, is essential fo development of this tanspot secto. This pape pesents a poposal fo a solution that uses spatially distinctive featues of the Dopple effect. The pesented navigation pocedue is based on an analytical elationship that descibes the Dopple fequency shift (DFS), as a function of the eceive position [32]. This fomula is the basis of the signal pocessing method called the signal Dopple fequency (SDF) [33 35], which is used in the location systems of emission souces [36] and navigation of objects [37, 38]. Using this method enables complete automation of the UAV landing pocedue and eliminates the need fo a boadband emote contol channel. Contempoay navigational systems often make use of pulse signals. Consequently, these systems equie lage spectum esouces. In contast to these solutions, the developed system is based on naowband signals (hamonics), which minimizes the spectum cost. Simulation studies fo VTOL ae pefomed in ode to detemine the effectiveness of the developed pocedue, that is, the accuacy with which the flight tajectoy is detemined. The obtained esults show the position eos that may occu duing the VTOL landing pocess and thus povide the oppotunity to evaluate the pactical implementation of the developed pocedue. The emainde of the pape is oganized as follows. Section 2 pesents opeation pinciples of the Dopple-based landing appoach system fo UAVs. The authos show a stuctue of this system and a navigation pocedue fo the landing phase. Simulation scenaios ae descibed in Section 3. In simulation studies, the authos assume that the vehicle is VTOL-capable. In Section 4, obtained esults ae pesented. Section 5 contains the summay of the pape. 2. Opeation Pinciple of the Poposed Landing System Assuming that the eceive (i.e., the UAV) moves at constant velocity, V, theelationshipbetweendfs,f D,andthesignal souce coodinates, (x,y,z), is descibed by [32] x Vt f D (x,y,z,t) f Dmax, (x Vt) 2 (1) +y 2 +z 2 whee f Dmax =f 0 V/c is the maximum DFS, f 0 is the caie fequency of the emitted signal, and c is the speed of light. Basedon(1),itfollowsthat,foknown(x, y, z),diection, and values of V, DFS measuement gives the possibility of detemining the coodinates of two possible positions of the eceive. By using a system of two efeence souces that ae located at the distance, we eliminate the ambiguity of the esult. Aveaging the esults obtained fom seveal efeence souces educes the eo of estimation fo the position coodinates of the object. Thus, inceasing the numbe of efeence souces inceases the accuacy of the positioning of the eceive. The stuctue of the naowband automatic landing system based on the SDF method is shown in Figue 1. The basic elements of the system ae the naowband navigation eceive (NR) and GNSS eceive integated with INS, both of which ae installed in the UAV, and fou adio beacons (RBs) that seve as efeence signal souces. Thee of RBs emit hamonic signals at fequencies f 1, f 2,and f 3, espectively. The fouth RB emits a modulated signal that contains infomation about the location of each RB elative to the destination UAV landing site. In addition, the naowband measuement eceive (MR) of this RB measues the cuent fequency of each RB. This infomation is also included with the modulating signal. This minimizes the impact of RB instability on the DFS detemination accuacy [39]. The fame stuctue of the modulating signal is pesented in Figue 2. Tansmission of infomation contained in the modulation signal fame is based on diffeential binay phase shift keying (DPSK). Selecting this modulation type makes it easy to eliminate signal modulation. The opeation of aising the DPSK signal to the second powe povides fo the econstuction of the caie wave whose fequency includes DFS [40]. As shown in Figues 1 and 4, the UAV navigation pocess consists of two essential stages: long-ange navigation and landing. Long-ange navigation uses integated GNSS + INS, because at this flight stage, accuate positioning is not a citical issue. Howeve, the landing stage equies high positioning accuacy. The standad GNSS eceive cannot achieve such accuacy, and thus navigation stage equies the use of dedicated solutions. A genealized algoithm of the pecise UAV positioning at the landing stage is shown in Figue 3.
Jounal of Advanced Tanspotation 3 P Z h T RB-3 h L Rx MR RB-4 RB-2 O Y RB-1 NR GNSS + INS L UAV X Figue 1: Stuctue of the automatic landing system. Cuent fequencies of each RB fom MR Fame Index [X1 Y1 Z1] F1 [X2 Y2 Z2] F2 [X3 Y3 Z3] F3 [X4 Y4 Z4] F4 Cyclic Redundancy Check (CRC) RB-1 RB-2 RB-3 RB-4 Figue 2: Fame stuctue of the modulating signal. GNSS + INS navigation and DFS contol of each RBs Detemination of neaest RB diection and coection of flight tajectoy Positioning of UAV(NR) based on teestial system by using (6) LP tajectoy of UAV Landing of UAV by altitude eduction PO tajectoy of UAV Figue 3: Genealized algoithm of the pecise UAV positioning at the landing stage.
4 Jounal of Advanced Tanspotation Z P L h L =h P h T O /2 d X (a) Z L h L P h P h T O /2 X d (b) Y L RB-4 RB-1 d P, O X RB-3 RB-2 (c) Figue 4: Diection of VTOL aival to landing site with espect to the location of RBs. Stictly detemined DFSs, which occu fo a close poximity to paticula RBs, ae the citeion fo the tansition of the UAV navigation system to the landing stage (point L,see Figues 1 o 4). At a elatively lage distance fom the landing site, the azimuth angle, φ, has a dominant impact on f D.In this case, it is θ φ, so f D f Dmax cos φ, (2) whee θ is the angle between the velocity vecto of the eceive (i.e., UAV) and the diection to the signal souce (i.e., RB). Hence, changes of DFSs on the LP (see Figues 1 o 4) line segment ae the basis fo coecting the UAV flight diection at a fixed altitude. Length of this segment is d. Because UAV is moving in the diection of signal souces, the citeion of this coection is a simultaneous maximization of DFSs fo the all RBs signals. In close poximity, the elevation angle begins to play a significant impact. If the smallest DFSs decease to a cetain citeia value, fo example, 0.8f Dmax, then UAV changes the flight diection by θ and begins to appoach towad the neaest RB. Duing movement, the RBs coodinates ae detemined in NR with espect to the system whose O (see Figues 1 o 4) is the oigin and the xaxis coincides with the diection of the new tajectoy. The fomulas that descibe the RBs coodinates ae defined using the SDF method [33] x k = V t 2A k (t 2 ) t 1 A k (t 1 ) t 2 A k (t 2 ) t 1 A k (t 1 ), k=1,...,k y 2 k + z 2 k =[V (t 2 2 t 1 )A k (t 2 )A k (t 1 ) ], A k (t 2 ) A k (t 1 ) k=1,...,k, (3)
Jounal of Advanced Tanspotation 5 whee ( x k, y k, z k ) is the estimated coodinates of the kth RB, A k (t) = 1 F k 2 (t)/ F k (t), F k (t) = [c f Dk (t)]/[vf k (t)], f Dk (t) is DFS fo the kth RB estimated by NR at t time moment, f k (t) is the caie fequency of the kth RB signal measued by MR and obtained fom the last eceived data fame, and t 1 and t 2 ae two diffeent time moments of the DFS measuement. The navigational coodinates of UAV ae detemined by tansfoming the RBs coodinates and the coodinates obtained by the system based on the UAV s flight tajectoy x UAVk =x k +( x k cos α+ y k sin α) cos β+ z k sin β, y UAVk =y k x k sin α+ y k cos α, k=1,...,k k=1,...,k z UAVk =z k ( x k cos α+ y k sin α) sin β+ z k cos β, k=1,...,k, whee ( x k, y k, z k ) ae the eal coodinates of the kth RB included in each data fame and α and β ae the estimated diections of the UAV flight in the azimuth (OXY) and elevation (OXZ) planes detemined elative to the destination landing site (see Figue 4) by using (3); that is, K α =atan ( 1 y k +y k ) K x k +x k k=1 β =atan ( 1 z k +z k ). K x k +x k K k=1 In [33, 35, 37], analysis of the SDF method shows that the tajectoy location elative to the signal souce has a significant influence on the accuacy of the positioning the object. The smallest positioning eo occus when α tends to 90, that is, when DFS conveges to zeo. Theefoe, to minimize the navigation eo, the weighted aveage coodinates elativetotheindividualrbsaeusedtoestimatetheuav coodinates [37] ( x UAV, y UAV, z UAV ) =( 1 W K k=1 K 1 1 w k x UAVk, w W k y UAVk, w W k z UAVk ), k=1 whee w k =1 F k (t) and W= K k=1 w k. RB system simplicity gives us the ability to position the navigation system in any field conditions. Additionally, the minimization of spectal esouces is a significant advantage of the pesented method compaed to the existing solutions. K k=1 3. Scenaios of Simulation Studies The effectiveness of the developed navigation pocedue detemines the accuacy of detemining the cuent UAV coodinates. In this pape, an assessment of the pocedue (4) (5) (6) accuacy is made on the basis of simulation tests, of which scenaios concen the VTOL navigation. In ou studies, the following assumptions and input data ae accepted: (i) landing point is the oigin of the coodinate system; (ii) base of the navigation system is fou RBs whose positions descibe the coodinates, (x k,y k,z k ),whee k = 1,2,...,K, (K= 4); the RBs coodinates and fequencies, f k,oftheemittedsignalsaepesented in Table 1; (iii) the Kth RBs emit the modulated DPSK signal, which contains infomation about the position coodinates and the cuent fequencies of the individual RBs; (iv) bandwidth of the DPSK signal is B T =80kHz; (v) opeating fequency of NR is f R = 2.4 GHz and the eception bandwidth is about B R = 500 khz; (vi) instantaneous DFS is detemined evey 0.5 sonthe basis of the spectal analysis duation of 1.0 s; basic fequency of the spectal analysis is 0.1 Hz; (vii) to analyze the Dopple cuves, the time windows T 1 = 5 sandt 2 =10saeused; (viii) in electomagnetic envionment, additive noise is occued,andtheleveloftheemittedsignalsatthe most distant point of the tajectoy povides SNR = 8 db; (ix) VTOL speed is V =72km/h = 20 m/s and the flight altitude is h L =50m. The eseach focuses on the impact assessment of vaious factos, such as the VTOL tajectoy position and the flight diection elative to the RBs. In the fist stage of simulation studies, two scenaios, Sc. 1 and Sc. 2, ae examined (α =0). Figue 4 shows thei geomety in the elevation (a and b, esp., fo Sc. 1 and Sc. 2) and azimuth plane (c). The second stage of the eseach is also based on Sc. 1 and Sc. 2. In this case, the study focuses on the impact assessment of the aival diection in the azimuth plane, α. We assumed that the UAV is flying using the long-ange navigation (GNSS + INS) method at an aveage altitude of h L.AtthepointL, the aicaft navigation system switches to the landing phase, that is, it begins to use the navigation pocedue descipted in Section 2. In Sc. 1, VTOL flies at a constantaltitudeto the pointp located above the landing site. The destination landing point O is eached by educing the altitude in the vetical diection. In the case of Sc. 2, the flight onthelinesegmentlp is pefomed at the angle β, assuming that, at the point P, UAVisatthealtitudeh P.Thisangleis detemined by β=atan ( h L h P ), (7) d whee d is the length of LP gaphical pojected on the azimuth plane (OXY). In simulation studies, we assume that d = 400 mandh P =30m. The VTOL appoach diection with espect to the position of RBs has also a significant influence on the navigation
6 Jounal of Advanced Tanspotation Table 1: Location coodinates of RBs and fequencies of signals emitted by them. kth RB x k (m) y k (m) z k =h T (m) f k (khz) Notes 1 20 20 2 2,399,800 Hamonic signal 2 20 20 2 2,399,850 Hamonic signal 3 20 20 2 2,399,900 Hamonic signal 4 20 20 2 2,400,100 DPSK; RB + MR accuacy. The esults of these studies show the equied location of RBs elative to the expected diection of VTOL appoach. In simulation studies, Sc. 1 and Sc. 2 ae also used to evaluate the impact of the VTOL appoach diection on the navigation eo. 4. Results The simulation studies involve the implementation of pocedues such as the geneation of RBs hamonic signals, geneation of envionmental noise, and detemination of thevtolcoodinatesbasedontheestimateddfss.each geneated hamonic signal contains DFS that esults fom thevtolpositionelativetocoespondingrb.thisdfsis detemined on the basis of (1). The geneated envionmental noise is a nomal band signal, whose dynamics elative to the hamonic signal dynamics povides SNR =8dB at the ange of the navigation system. The VTOL positioning pocedue is pefomed as descibed in Section 2. To evaluate the positioning eo, ΔR, also called the navigation eo, the following metic is used: ΔR = ( x UAV x UAV ) 2 +( y UAV y UAV ) 2 +( z UAV z UAV ) 2, whee ( x UAV, y UAV, z UAV ) and (x UAV,y UAV,z UAV ) ae the estimated and eal coodinates of UAV, espectively. The assessment of the impact of the landing tajectoy position with espect to RBs is pefomed fo α = 0 and two analyzed tempoal windows. Figues 5 and 6 pesent the navigation eo vesus the distance to the taget landing site fo T 1 and T 2,espectively. As you can see, the aveage navigation eos, ΔR on line section d = 400 mfosc.1,aemoethan1.3 and 1.6 times smalle compaed to Sc. 2 fo T 1 and T 2,espectively.Atthe taget landing point, these eos each T 1 Sc. 1: ΔR = 0.08 m and Sc. 2: ΔR = 0.33 m; on the othe hand T 2 Sc. 1: ΔR = 0.18 mandsc.2:δr = 0.46 m. Fo Sc. 2, the navigation eo has a lage deviation (spead), σ Δ,whichis12 m, and fo Sc. 1, this eo deviation does not exceed 5 m. The diection of the VTOL landing appoach elative to the RBs locations has also a significant impact on ΔR.Based on Sc. 1 fo T 1, simulation studies that show a vaiability in navigation accuacy as a function of distance fom the landing site ae pefomed. The esults of these tests, that is, ΔR vesus the distance fom the landing point, ae shown in Figue 7 fo selected values α=0, α=20,andα=40. ItcanbeseenthatthepopeVTOLappoachdiection with espect to the landing site minimizes navigation (8) Navigation eo ΔR (m) Navigation eo ΔR (m) Navigation eo ΔR (m) 10 2 10 1 10 0 10 1 400 350 300 250 200 150 100 50 0 Distance UAV to point P, l (m) Sc. 1: ΔR(l) Sc. 2: ΔR(l) Sc. 1: ΔR = 7.1 m Sc. 2: ΔR = 9.05 m Figue 5: Navigation eo fo Sc. 1 and Sc. 2 and T 1. 10 2 10 1 10 0 10 1 400 350 300 250 200 150 100 50 0 Distance UAV to point P, l (m) Sc. 1: ΔR(l) Sc. 2: ΔR(l) Sc. 1: ΔR = 3.71 m Sc. 2: ΔR = 6.03 m Figue 6: Navigation eo fo Sc. 1 and Sc. 2 and T 2. 10 2 10 1 10 0 10 1 400 350 300 250 200 150 100 50 0 Distance UAV to point P, l (m) Sc. 1: ΔR(l) fo =0 Sc. 1: ΔR(l) fo =20 Sc. 1: ΔR(l) fo =40 Figue 7: Navigation eo fo α=0, α=20,andα=40 Sc.1.
Jounal of Advanced Tanspotation 7 ±180 150 150 120 90 1 0.5 1.5 120 90 60 Sc. 1: ΔR(0),T 1 Sc. 1: ΔR(0),T 2 60 30 30 0 espectively. These cases occu when the appoach diection ovelap with the diections set by pais of RBs. In the case of Sc. 2 fo T 2, we can see that a distibution of the final eos has a moe unifom chaacte. This is also evidenced by the eo deviations, which fo Sc. 1 ae lage (0.22 m) than fo Sc. 2 (0.08 m). Fo T 1, the eo deviations ae simila, that is, 0.14 mand0.16 mfosc.1andsc.2,espectively. The esults of the simulation tests show the high positioning accuacy of the aicaft in the landing phase. This indicates the easonability of the pactical implementation of the developed pocedue. Based on the esults, it can be concluded that the smalle navigation eos wee obtained fo Sc. 1 and the longe analysis time, that is, fo T 1.Howeve, given the natue of the VTOL flight, the smalle time window may be moe pactical. In this case, the location of the flight tajectoy shown in Sc. 2 may be bette, due to the independence of the navigation eo fom the diection of landing appoach. Figue 8: Final navigation eo vesus aival diection fo Sc. 1. ±180 150 150 120 90 1 0.5 1.5 120 90 60 Sc. 2: ΔR(0),T 1 Sc. 2: ΔR(0),T 2 60 30 30 Figue 9: Final navigation eo vesus aival diection fo Sc. 2. eos. This fact esults fom the compaison of mean eos, which ae 7.10 m, 2.66 m, and 3.10 m, espectively, fo α = {0,20,40 }. Howeve, the navigation eo that occus at the destination landing point is most significant. Hence, the influence assessment of the appoach diection in the azimuth plane on the final eo of the VTOL positioning is made on the basis of simulation eseach. Fo Sc. 1 and Sc. 2, the obtained esults ae pesented in Figues 8 and 9, espectively. Fo Sc. 1, the obtained esults show that the aveage eos of the final position fo all diections ae smalle than 0.5 m fo T 1 and T 2. In the case of Sc. 2, these eos ae 0.52 m and 0.63 mfot 1 and T 2, espectively. In Figue 8, we see that thee ae fou cucial sectos of the VTOL aival fo which ΔR can each up to 0.90 mand1.3 mfot 1 and T 2, 0 5. Conclusions This pape povides the navigation pocedue that enables the automatic landing appoach of VTOL. The developed pocedue is based on the Dopple effect and can be made using a simple shot-ange navigation system that is mounted anywhee at the taget UAV landing. Aound this place, RBs ae deployed, which tansmit the hamonic signals and the naowband signal containing the infomation about them positions. In addition, the tansmission of fequency coections ensues that the influence of fequency instability of signal souces is minimized. This navigation system can wok completely independently of GNSS and equies small spectum esouces. In the pape, the authos evaluated the impact of the UAV tajectoy, the diection of landing appoach elative to RBs, and the tempoal window of the signal analysis on the accuacy of the developed pocedue. This assessment wasmadeonthebasisofsimulationstudies.theesults show the high pecision of the VTOL positioning. Pope selection of paametes shows that the navigation eo nea the destination landing point is less than 1 m. The executed eseach has shown that the developed pocedue can contibute to the full automation of the UAV landing pocess, which may be impotant fo thei use in tanspot. Conflicts of Inteest The authos declae that they have no conflicts of inteest. Refeences [1] M. Totonesi, C. Stefanelli, E. Benvegnu, K. Fod, N. Sui, and M. Lindeman, Multiple-UAV coodination and communications in tactical edge netwoks, IEEE Communications Magazine, vol. 50, no. 10, pp. 48 55, 2012. [2] M. A. Ma Sum, M. K. Aofi, G. Jati et al., Simulation of intelligent Unmanned Aeial Vehicle (UAV) fo militay suveillance, in Poceedings of the 2013 5th Intenational Confeence on
8 Jounal of Advanced Tanspotation Advanced Compute Science and Infomation Systems, ICACSIS 2013, pp. 161 166, Septembe 2013. [3] D.Ofanus,E.P.DeFeitas,andF.Eliassen, Self-Oganization as a Suppoting Paadigm fo Militay UAV Relay Netwoks, IEEE Communications Lettes,vol.20,no.4,pp.804 807,2016. [4] J. A. J. Beni, P. J. Zaco-Tejada, L. Suáez, and E. Feees, Themal and naowband multispectal emote sensing fo vegetation monitoing fom an unmanned aeial vehicle, IEEE Tansactions on Geoscience and Remote Sensing, vol.47,no.3, pp. 722 738, 2009. [5] J. Wijitdechakul, S. Sasaki, Y. Kiyoki, and C. Koopipat, UAVbased multispectal image analysis system with semantic computing fo agicultual health conditions monitoing and ealtime management, in Poceedings of the 2016 Intenational Electonics Symposium (IES), pp. 459 464, Denpasa, Indonesia, Septembe 2016. [6] P. B. Quate, F. Gimaccia, S. Leva, M. Mussetta, and M. Aghaei, Light Unmanned Aeial Vehicles (UAVs) fo coopeative inspection of PV plants, IEEE Jounal of Photovoltaics, vol. 4, no. 4, pp. 1107 1113, 2014. [7] F. Gimaccia, M. Aghaei, M. Mussetta, S. Leva, and P. B. Quate, Planning fo PV plant pefomance monitoing by means of unmanned aeial systems (UAS), Intenational Jounal of Enegy and Envionmental Engineeing, vol.6,no.1,pp.47 54, 2015. [8] S. Daliento, A. Choude, P. Gueieo et al., Monitoing, diagnosis, and powe foecasting fo photovoltaic fields: a eview, Intenational Jounal of Photoenegy, vol. 2017, Aticle ID 1356851, 13 pages, 2017. [9]S.Babak,M.Myslovych,andR.Sysak, Modulestuctue of UAV-based computeized systems fo emote envionment monitoing of enegy facilities, in Poceedings of the 17th Intenational Confeence Computational Poblems of Electical Engineeing, CPEE 2016, Septembe 2016. [10] Y. Lu, D. Macias, Z. S. Dean, N. R. Kege, and P. K. Wong, A UAV-Mounted Whole Cell Biosenso System fo Envionmental Monitoing Applications, IEEE Tansactions on Nanobioscience,vol.14,no.8,pp.811 817,2015. [11] D. Edos, A. Edos, and S. E. Watkins, An expeimental UAV system fo seach and escue challenge, IEEE Aeospace and Electonic Systems Magazine,vol.28,no.5,pp.32 37,2013. [12] K. A. Ghamy, M. A. Kamel, and Y. Zhang, Coopeative foest monitoing and fie detection using a team of UAVs-UGVs, in Poceedings of the 2016 Intenational Confeence on Unmanned Aicaft Systems, ICUAS 2016, pp. 1206 1211, June 2016. [13] C. Yuan, Z. Liu, and Y. Zhang, Vision-based foest fie detection in aeial images fo fiefighting using UAVs, in Poceedings of the 2016 Intenational Confeence on Unmanned Aicaft Systems, ICUAS 2016, pp. 1200 1205, June 2016. [14] S. Rathinam, P. Almeida, Z. Kim et al., Autonomous seaching and tacking of a ive using an UAV, in Poceedings of the 2007 Ameican Contol Confeence, ACC, pp. 359 364, July 2007. [15] D. Popescu, L. Ichim, and T. Caamihale, Flood aeas detection based on UAV suveillance system, in Poceedings of the 19th Intenational Confeence on System Theoy, Contol and Computing, ICSTCC 2015 - Joint Confeence SINTES 19, SACCS 15, SIMSIS 19, pp. 753 758, Octobe 2015. [16] UNICEF website, https://www.unicef.og/. [17] AiVein, 11-Aug-], http://www.aivein.pl/. [18] G.Hoaeau,J.J.Liebenbeg,J.G.Musial,andT.R.Whitman, Package tanspot by unmanned aeial vehicles, US0068265 A1,2016. [19] P. D. Goves, Pinciples of GNSS, inetial, and multisenso integated navigation systems,atechhouse,boston,ma,usa, 2nd edition, 2013. [20] M.S.Gewal,A.P.Andews,andC.G.BatoneWiley-Intescience, Hoboken, 3d edition, 2013. [21] E. D. Kaplan, Undestanding GPS: Pinciples and applications, Atech House, Boston, MA, USA, 2nd edition, 2005. [22] F. van Diggelen, A-GPS: Assisted GPS, GNSS, and SBAS,Atech House, 2009. [23] S. Wad, J. Hensle, B. Alsalam, and L. F. Gonzalez, Autonomous UAVs wildlife detection using themal imaging, pedictive navigation and compute vision, in Poceedings of the 2016 IEEE Aeospace Confeence, AERO 2016,Mach2016. [24] F. S. Leia, T. A. Johansen, and T. I. Fossen, Automatic detection, classification and tacking of objects in the ocean suface fom UAVs using a themal camea, in Poceedings of the 2015 IEEE Aeospace Confeence, AERO 2015,Mach2015. [25] H. Sheng, H. Chao, C. Coopmans, J. Han, M. McKee, and Y. Chen, Low-cost UAV-based themal infaed emote sensing: Platfom, calibation and applications, in Poceedings of the 2010 IEEE/ASME Intenational Confeence on Mechatonic and EmbeddedSystemsandApplications,MESA2010,pp.38 43,July 2010. [26] C. Leśnik, P. Seafin, and A. Kawalec, Azimuth ambiguity suppession in SAR images using Dopple-sensitive signals, Bulletin of the Polish Academy of Sciences: Technical Sciences,vol. 63, no. 1, pp. 221 227, 2015. [27] B. Dawidowicz, A. Gadoś, A. Gozelańczyk et al., Fist polish SAR tials, IEE Poceedings: Rada, Sona and Navigation, vol. 153, no. 2, pp. 135 143, 2006. [28] M. Cais, S. Stanko, M. Malanowski et al., Mm-Wave SAR demonstato as a test bed fo advanced solutions in micowave imaging, IEEE Aeospace and Electonic Systems Magazine,vol. 29, no. 7, pp. 8 15, 2014. [29] L. J. Ippolito, Satellite Communications Systems Engineeing: Atmospheic Effects, Satellite Link Design and System Pefomance, Satellite Communications Systems Engineeing: Atmospheic Effects, Satellite Link Design and System Pefomance,pp. 1 376, 2008. [30] Instument flying handbook, CeateSpace Independent Publishing Platfom, 2013. [31] M. Kayton and W. R. Fied, Avionics Navigation Systems,John Wiley & Sons, Inc., Hoboken, NJ, USA, 1997. [32] J. Rafa and C. Ziółkowski, Influence of tansmitte motion on eceived signal paametes Analysis of the Dopple effect, Wave Motion,vol.45,no.3,pp.178 190,2008. [33] J. M. Kelne, Analizadoppleowskiej methodlokalizacjizódelemisjifaladiowych, ozpawadoktoska (Analysis of the Dopple location method of the adio waves emission souces, Militay Univesity of Technology, 2010. [34] P. Gajewski, C. Ziółkowski, and J. M. Kelne, Using SDF method fo simultaneous location of multiple adio tansmittes, in Poceedings of the 2012 19th Intenational Confeence on Micowaves, Rada and Wieless Communications, MIKON 2012, pp. 634 637, May 2012. [35] A. T. de Hoop, Comments on: J. Rafa, C. Ziólkowski, Influence of tansmitte motion on eceived signal paametes an analysis of the Dopple effect, Wave Motion 45 (2008) 178 190 [MR2449241], Wave Motion. An Intenational Jounal Repoting Reseach on Wave Phenomena, vol.46,no.1,pp.89 91, 2009.
Jounal of Advanced Tanspotation 9 [36] N. A. Lehtomaki, T. C. Poling, and D. A. Schuyle, Single platfom Dopple geolocation, Aticle ID 130794, pp. 03 2015, 2015. [37] J. M. Kelne, C. Ziółkowski, and L. Nowosielski, Local navigation system fo VTOLs used on the vessels, in Poceedings of the IEEE/ION Position, Location and Navigation Symposium, PLANS 2016, pp. 415 421, Apil 2016. [38] J. M. Kelne, C. Ziółkowski,L.Nowosielski,andM.Wnuk, Reseve navigation system fo ships based on coastal adio beacons, in Poceedings of the IEEE/ION Position, Location and Navigation Symposium, PLANS 2016, pp. 393 402, Apil 2016. [39] J. M. Kelne, C. Ziółkowski, and P. Maszalek, Influence of the fequency stability on the emitte position in SDF method, in Poceedings of the 2016 Intenational Confeence on Militay Communications and Infomation Systems, ICMCIS 2016, May 2016. [40] J. M. Kelne and C. Ziółkowski, The use of SDF technology to BPSK and QPSK emission souces location, Pzeglad Elektotechniczny,vol.91,no.3,pp.61 65,2015.
http://www.hindawi.com Volume 201 Intenational Jounal of Rotating Machiney Jounal of http://www.hindawi.com Volume 201 The Scientific Wold Jounal Jounal of Sensos Intenational Jounal of Distibuted Senso Netwoks Jounal of Contol Science and Engineeing Advances in Civil Engineeing Submit you manuscipts at https://www.hindawi.com Jounal of Robotics Jounal of Electical and Compute Engineeing Advances in OptoElectonics http://www.hindawi.com Volume 2014 VLSI Design Intenational Jounal of Navigation and Obsevation Modelling & Simulation in Engineeing Intenational Jounal of Intenational Jounal of Antennas and Chemical Engineeing Popagation Active and Passive Electonic Components Shock and Vibation Advances in Acoustics and Vibation