216 America Cotrol Coferece (ACC) Bosto Marriott Copley Place July 6-8, 216. Bosto, MA, USA Equivalet Cosumptio Miimizatio Strategy for Hybrid All- Electric Tugboats to Optimize Fuel Savigs Liza Chua Wa Yua, Tegoeh Tjahjowidodo, Gerald Seet Gim Lee, Ricky Cha, Alf Kåre Ådaes Abstract The stregtheig of ship emissio cotrols ad expected icrease of fuel oil prices i the future have drive the iterest i fuel efficiet solutios for the maritime idustry. Hybrid all-electric tugboat has show its potetial i improvig fuel ecoomy. However, usig covetioal rule-based/heuristic power maagemet strategies relies largely o the experiece of the desigers ad crew operators, which caot guaratee optimality. Motivated by the optimizatio-based power maagemet strategies for Hybrid Electric Vehicles (HEVs), this paper ivestigates the potetial beefits of Equivalet Cosumptio Miimizatio Strategy (ECMS) i hybrid allelectric tugboats applicatio. The proposed power maagemet strategy promises 17.6% of fuel savigs compared to covetioal rule-based strategy. I. INTRODUCTION The iteratioal shippig idustry cotributes sigificatly to the global ecoomy, as it is resposible for the carriage of about 9% of world trade. As world populatio cotiues to icrease, the world seabore trade carried i toes is expected to icrease expoetially from 1 billio to approximately 17 billio i 23 [1]. With this icrease i shippig demad, ship emissio is expected to icrease. Without emissios cotrol regulatios beig put i place, the shippig emissios ca icrease up to 25% by 25 [2]. I 211, the Iteratioal Maritime Orgaizatio (IMO) itroduced a revised MARPOL Aex VI Prevetio of air pollutio from ships [3], which sets tighter emissio limits o SO x, NO x ad diesel particulate matter (PM), to progressively reduce the global emissios. I additio, Emissio Cotrol Areas (ECAs) are itroduced to further reduce those air pollutats i the desigated areas. Startig from year 215, ships i the Sulphur Emissio Cotrol Areas (SECAs) are required to use light fuel oil (LFO) with maximum.1% sulphur cotet. This icreases the operatio cost cosiderably as LFO with lower sulphur cotet are more expesive [4]. The evirometal regulatios ad the expected icrease i the operatio cost led to a growig iterest i the maritime idustry to search ito fuel efficiet solutios. Tugboats are used to tow takers ad bulk carriers to the harbor, where the waters are too shallow for the vessels to move by themselves. With icrease i shippig demad, the demad for tugboats are expected to icrease. Furthermore, Liza Chua Wa Yua, Tegoeh Tjahjowidodo ad Gerald Seet Gim Lee are with the School of Mechaical ad Aerospace Egieerig, Nayag Techological Uiversity, Sigapore 639798. (e-mail: ch2za@e.tu.edu.sg, ttegoeh@tu.edu.sg, mglseet@tu.edu.sg) Ricky Cha ad Alf Kåre Ådaes are with BU Maries ad Ports, Process Automatio Divisio, ABB Pte. Ltd., Sigapore139935. (e-mail: ricky.cha@sg.abb.com, alfkare.adaes@c.abb.com) tugboats usually operate aroud the harbor, which is close to the urba areas ad hece, are more likely to face stricter emissio cotrols i subsequet years. However, tugboats with covetioal diesel mechaical propulsio system ted to face a greater challege to achieve fuel efficiet operatio, due to large load variatios i the differet modes of operatio. A represetative load profile of a harbor tug uder differet modes of operatio i oe operatio cycle is obtaied from [6], as show i Fig. 1. The load profile is derived based o idustrial iformatio o the typical duratio ad load demad durig differet modes of operatios. The 1% loadig o the vertical axis of the load profile represets the total istalled propulsio power of the harbor tug. The egies are covetioally sized to provide for the maximum propulsio power durig the workig regio. As see from the figure, the workig regio lasted for less tha 25% of the total duratio of the operatio cycle. The egies operate at low loadig coditio for the rest of the operatio for stadby ad trasits, which is iefficiet for the egies. I recet years, the hybrid all-electric propulsio has bee itroduced as the alterative solutio, which ca overcome low egie loadig ad offers other beefits [5]. The system cosists of mai egies ad batteries as the power sources. Shore power ca be utilized to charge the batteries whe the tugboat returs to harbor, which is cheaper ad produce less hazardous emissio. This eergy ca be used as a additioal power source to assist the egies whe the tugboat is i operatio off the shore. Studies have show hybrid allelectric propulsio has large potetial to achieve sigificat amout of fuel savigs for harbor tugs [6]. However, a hybrid eergy source icreases the complexity of the power maagemet problem. Power maagemet strategy usig covetioal rule-based/heuristic method depeds largely o the experiece ad kowledge of the desigers ad crew operators, which caot guaratee the optimal power split betwee the egies ad batteries to meet the varyig load demad while esurig fuel efficiecy [7]. With the icreased flexibility to utilize multiple power geeratio sources, a more itelliget power maagemet strategy is required to coordiate the power allocatio betwee the differet power geeratios sources. Istead, power maagemet cotrol ca be formulated ito a optimizatio problem. Optimizatio-based power maagemet strategies have bee widely researched for ladbased vehicles ad have prove to improve fuel ecoomy as compared to rule-based approaches [8]. I the case of power maagemet cotrol with the objective of reliability ad fuel efficiecy, a cost fuctio ca be formulated to solve the optimal power split for each geeratio source to achieve least fuel cosumptio. 978-1-4673-8681-4/$31. 216 AACC 683
Several researches have looked ito differet optimizatio-based power maagemet strategies for marie hybrid systems. I [9], Model Predictive Cotrol (MPC) is used to determie the power split betwee the batteries ad the ultra-capacitors, to miimize power fluctuatios. I [1-11], a combiatio of dyamic optimizatio techiques is used to mitigate dyamic pulse loads for aval vessel applicatios. I these literatures, less focus is placed o fuel efficiecy. Optimizatio algorithms are commoly used for model parameter estimatio [12-13]. I [14], optimizatio problem is formulated to achieve fuel efficiecy while meetig load demad for hybrid all-electric tugboats. Dyamic optimizatio techiques are used to miimize the cost fuctio, with ideal assumptio of a kow loadig profile of the vessel. However, this is ot realistic as each job is differet i terms of duratio ad total power requiremet. Hece, i [15-16], a predictio scheme is itroduced to predict the load profile. The performace of the power maagemet strategies will largely deped o the accuracy of the predictio. Furthermore, low egie loadig ad frequet switchig betwee egies ad batteries power is also observed from the optimizatio results, which is udesirable i practice. Equivalet Cosumptio Miimizatio Strategy (ECMS) is a static optimizatio-based strategy that miimizes the istataeous cost fuctio. Hece, it does ot require prior kowledge of the load profile, which makes it suitable for real-time implemetatio. Furthermore, ECMS have show to achieve fuel efficiecy that is close to the global optimal solutio from Dyamic Programmig [8]. ECMS has bee largely applied to series ad parallel HEVs [17-2]. More recetly, the applicatio has also bee exteded to hybrid fuel cell vehicles [21], city buses [22] ad more electric aircraft [23]. Motivated by the extesive researches o ECMS for hybrid lad-based vehicles, ECMS is ivestigated for hybrid all-electric tugboat applicatio i this paper. The objective of the power maagemet strategy is to meet the power load demad with miimal fuel cosumptio. Several factors to be take ito cosideratio iclude the egie optimal operatig rage ad power limits, battery chargig/dischargig rates, ad the state-of-charge (SOC) of the batteries. The mai cotributio of the proposed ECMS is the ability to maitai the egie i their optimal operatig rage while effectively maages the istataeous power-split betwee the egies ad the batteries to achieve fuel savigs. The performace of the proposed strategy is evaluated by comparig the fuel cosumptio of a rule-based method o the same system. Figure 1: Harbor tug load profile The rest of the paper is orgaized as follows. Chapter II describes the power system compoets of the hybrid allelectric tugboat. Chapter III itroduces the cocept of ECMS ad formulatio of the optimizatio problem. I Chapter IV, the proposed optimizatio algorithm is evaluated usig oliear costraied optimizatio i MATLAB/Simulik eviromet. The results from simulatio are discussed, followed by cocludig remarks i Chapter V. II. HYBRID ALL-ELECTRIC SYSTEM DESCRIPTION A geeral layout of the power distributio system of the hybrid all-electric tugboat ivestigated i this paper is show i Fig 2, which is similar to the system show i [6]. The hybrid all-electric power system comprises of egies/geerators (gesets) ad batteries as the mai power geeratio sources, power coverters ad electrical AC motors. The loads cosidered icludes the mai propulsio load, service ad hotel loads such as pumps, Heatig, Vetilatio ad Air-coditioig (HVAC), lightigs, etc.. The system cosidered uses a DC distributio etwork istead of the covetioal AC power distributio system. Oe of the mai advatage is that the geerators frequecy eed ot be sychroized ad hece, the use of variable speed egies is possible, which ca achieve greater fuel savigs [5]. Other advatages of usig the DC distributio are highlighted i [24]. Figure 2: Geeral layout of hybrid all-electric tugboats power distributio system To simplify the system, the followig assumptios were made: 1. Dyamics of power coverters i the order of millisecods are assumed egligible. Losses through the power coverters are assumed a costat drivetrai efficiecy value, ηη dddd. 2. Dyamics of the egies ad geerators are ot cosidered. The system assumes a ideal coditio where power demad from the loads ca be met istataeously. 3. Trasiets ad temperature of the batteries are ot cosidered. The power flow from the simplified power system, assumig umbers of gesets, are outlied i Fig. 3. ηη eeeeee, ηη gggggg, ηη bbbbbbbb are the efficiecies of the egie, geerator 684
ad battery respectively. ηη dddd is the average efficiecies of the power coverters. PP eeeeee,1 PP eeeeee, ad PP gggggg,1 PP gggggg, are the power supplied from each egie ad geerator respectively ad PP bbaaaaaa is the chargig (PP bbbbbbbb < ) ad dischargig power (PP bbbbbbbb ) from the battery, expressed i watts. The total power demad from the propulsio, service ad hotel loads, deoted by PP llllllll, must be met by the power supplied from the gesets ad batteries. Takig ito cosideratio the efficiecies of the respective compoets, the followig equatio ca be formed. P load = i=1 P ge,i η dc + P batt η batt η dc (1) betwee the egie ad geerator, the relatioship betwee the egie ad geerator power ca be expressed as P eg = P ge ge (3) B. Batteries The SOC of the battery is the measure of the amout of capacity i the battery. I order to maitai the operatig life spa of the battery, limits o the SOC eeds to be set so as to prevet over-chargig or over-dischargig. The rage of SOC where the battery is permitted to vary is referred to as the Depth-of-Discharge (DOD). I this study, the oly state of iterest is the SOC of the battery. Hece, a simple geeric battery model as show i Fig.5, is used to derive the state equatio for the battery SOC. Figure 3: Simplified power flow of hybrid all-electric power system A. Variable speed egies Specific fuel oil cosumptio (SFOC), i g/kwh, is the measure of the amout of fuel cosumed (g) by the egie to produce a uit of eergy (kwh). The SFOC at differet egie loadig is derived from the egie performace curve, which ca be obtaied from the egie maufacturers. The SFOC curve show i Fig. 4 is a represetative tred of high speed diesel egies while operatig at variable speed betwee 6% - 1%, derived from multiple egie models. The 1% SFOC o the vertical axis represets the SFOC at maximum cotiuous ratig of the egie. Noticed that the SFOC is the lowest whe the egie is loaded at 6%-8% of the egie s rated power. Whe the egie loadig falls below 4%, the SFOC icreases expoetially. Hece, i order to achieve better fuel efficiecy, it is udesirable to operate the egies i the low loadig coditio. SFOC (%) 125 12 115 11 15 1 Variable speed egie 95 2 4 6 8 1 Egie loadig (%) Figure 4: SFOC curve of a variable speed egie The relatioship betwee egie fuel cosumptio ad the egie power ca be expressed as: C eg = SFOC P eg (2) 3.6 1 6 where, CC eeeeee (g/s) is the egie fuel cosumptio, SSSSSSSS (g/kwh) is fuel oil cosumptio give a particular egie loadig. Takig ito cosideratio the mechaical loss Figure 5: Simple battery model Accordig to the battery model, effective battery power ca be defied as: P batt =V oc I batt -I batt 2 R batt (4) where VV oooo (V) is the ope circuit voltage of the battery, II bbbbbbbb (A) is the battery curret ad RR bbbbbbbb (Ω) is the battery iteral resistace. II bbbbbbbb 2 RR bbbbbbbb represets the power loss across the battery iteral resistace. For simplicity, the battery efficiecy ηη bbbbbbbb accouts for the average iteral battery losses ad VV oooo is assumed costat. By defiitio, the variatios i battery SOC ca be represeted by the state equatio show i [7] as follows: SOC = I batt (5) Q max where QQ mmmmmm (Ah) is the battery capacity. By itegratig equatio (5) over oe samplig period t ad combiig equatio (4), the followig states equatio of the battery SOC ca be obtaied. Positive P batt idicates battery dischargig ad vice versa. SOC(k+1)=SOC(k)- III. ECMS P batt t V oc Q max 36 (6) Equivalet Cosumptio Miimizatio Strategy was first itroduced i [14], where a equivalet fuel cost is accouted for the battery power used. I this way, the eergy from the egies ad the batteries are made comparable. The cocept is developed based o charge sustaiig, which ca be illustrated i Fig. 6. It is assumed that whe the battery discharges, the amout of eergy that the batteries discharge will be charged by the egies i the future. The amout of fuel required for the egie to charge the battery will be the equivalet fuel cosumptio of the battery. The mai objective of the power maagemet cotrol strategy is to determie the optimal power split betwee the 685
egies ad batteries that miimize the fuel cosumptio, while esurig that the load demad is met. Hece, the cost fuctio ca be formulated as: mi J= C total,eqv = i=1 C eg,i +C batt (7) where C total,eqv (g/s) is the total equivalet cosumptio, ii= CC eeeeee,ii (g/s) is the total fuel cosumptio from egies, ad CC bbbbbbbb (g/s) is the equivalet fuel cosumptio of the batteries. CC tttttttttt,eeeeee ca be greater or lesser tha the actual fuel cosumptio, depedig o whether the batteries are chargig or dischargig. At every time step, give a load demad, the ECMS will solve the optimal power-split betwee the geerator, PP gggggg ad battery PP bbbbbbbb that will result i the miimal total equivalet fuel cosumptio. Figure 6: Equivalet fuel cosumptio of the battery For tugboats, the mass flow rate of the fuel is expressed as a fuctio of the egie power loadig, istead of a fuctio of torque ad speed used i most HEVs literature. Combiig equatio (2) ad (3), the total egie fuel cosumptio ca be calculated as: i=1 C eg,i = SFOC i P ge,i i=1 (8) ge,i 3.6 1 6 The equivalet cosumptio of the battery ca be defied as: C batt = s FC P batt (9) 3.6 1 6 where s is the equivalece factor, FFFF is the fuel cosumptio coversio factor i g/kwh. The results of the ECMS hece largely depeds o the equivalece factor ad the fuel cosumptio coversio factor. Existig literatures has ivestigated differet alteratives to defie the equivalece factor [19]. Oe of the methods is to defie the equivalece factor as a costat. The costat equivalet factor is derived usig the average efficiecies from the fuel to the battery ad vice-versa durig chargig ad dischargig cases. By applyig this method, the equivalet egie power required for the charged/discharged battery power will be derived. I equatio (1), PP bbbbbbbb has bee defied to be the actual power output of the battery, hece the equivalece factor ca be defied as: 1 s = (1) ge,i η batt I this case, the fuel cosumptio coversio factor, FFFF is actig similar to a weightig fuctio. Hece, i this paper, it is proposed that settig the value of FFFF as the lowest SFOC of the egie ca discourage low egie loadig. Whe the load demad falls withi the iefficiet low loadig regio of the egie (eg. below 4%), the value of SFOC for the egie will be higher tha value of FFFF, sice FFFF is fixed at the lowest poit of the SFOC curve. Hece, miimizig the cost fuctio will either lead to the use of batteries etirely, or icreasig the egie loadig to the optimal operatig poit ad uses the excess power to charge the battery, if the battery is available for chargig below the maximum SOC. With this proposed method, a higher pealty is applied to operatig the egies at ay poit other tha the optimal operatig poit with the lowest SFOC, hece discouragig low egie loadig solutios. A similar cocept has also bee applied i [25]. The optimizatio problem is subjected to a set of equality ad iequality costraits. The equality costraits are set accordig to equatio (1) to esure that the load demad is met. The iequality costraits that govers the power limits of the geerators, the maximum charge ad discharge rate of the battery that limits the battery power, ad the SOC limit of the battery are as follows: P ge,i_mi P ge,i P ge,i_max P batt_mi P batt P batt_max SOC mi SOC SOC max (11) The decisio variables are P ge,i ad P batt. The optimizatio problem is o-covex, due to the oliear relatioship betwee P ge ad the egie fuel cosumptio. Hece, oliear programmig techique is used to solve the optimizatio problem. I this paper, the costraied oliear solver fmico i MATLAB/Simulik optimizatio toolbox is used. IV. CASE STUDIES A. System parameters I this chapter, the proposed ECMS is validated for hybrid all-electric tugboat operatio. The system cosidered i this case study is based o a 6 65 BPT harbor tugboats, which cosists of two gesets ad batteries. Note that for this case study, the total istalled geerator power represets less 7% of the maximum propulsio power to achieve full bollard pull. The size of the batteries is calculated to esure that it is sufficiet to supply 15 miutes of cotiuous power to boost the egie powers to meet the maximum load demad. I order to assess the performace of the ECMS i terms of its fuel savigs ability, the same system parameters are used accordig to [6], where the rule-based cotrol is ivestigated for the hybrid all-electric tugboat. The parameters are listed i Table 1. The load profile of a harbor tug for oe operatio cycle as show i Fig. 1, is used to evaluate the performace of the power maagemet strategy. The modes of operatio ca be broadly classified ito stadby/trasit ad workig. At the start of the operatio cycle, the tugboat typically trasits from the harbor to the workig regio, followed by a stadby period ear the workig regio while waitig for the work to 686
start. The workig duratio usually last about 3mis, with the highest amout of power throughout the cycle. The tugboat leaves the workig regio upo completio of work, trasitig either to aother workig regio, or back to the harbor. A costat auxiliary ad hotel load of 5kW is assumed preset throughout the operatio cycle. TABLE I. SYSTEM PARAMETERS Maximum propulsio power P load,max 38 kw Geerator rated power P ge,1_max, P ge,2_max 12 kw SFOC @ 1% egie power SFOC 211 g/kwh Ope circuit voltage V oc 1 V Battery capacity Q 5 Ah Maximum chargig P batt_mi 2C Maximum discharge P batt_max 3C Depth-of-discharge DOD.6 Geerator efficiecy ge,1, ge,2 96.5 % Battery efficiecy η batt 98 % Drivetrai efficiecy η dc 94.5 % B. Evaluatio of simulatio results The optimizatio is performed i the MATLAB/Simulik eviromet, usig iterior-poit method i the fmico algorithm. A sample time of 1s is used ad the results are show i Fig. 7. The simulated results are as expected. From the simulatio results i Fig. 7, it is see that durig low loadig coditios where the tug is trasitig, geerator 1 operate i its optimal operatig poit of about 8% loadig, while geerator 2 is switched off. Depedig o the loadig coditio, the batteries either supply additioal eergy to supplemet the total load demad, or absorb the excess eergy from the geerator. Geerator 1 is well maitaied at its optimal operatig poit throughout the operatio cycle. Geerator 2 is switched o durig the peak loadig coditio where the power from the geerator 1 ad the batteries are ot sufficiet for the load demad. Both geerators 1 ad 2 are loaded to its optimal operatig poit ad the batteries provide the excess power required. The SOC of the battery is maitaied betwee the.3 to.9 limit that is set, ad is close to the miimum SOC limit towards the ed of the operatio. After oe operatio cycle, the tugboats geerally have a idle period, where the batteries ca be charged before the ext operatio. C. Performace compariso with rule-based strategy The fuel cosumptio of the egies for the rule-based cotrol is compared agaist the fuel cosumptio for the proposed ECMS i Fig. 8. Comparig the total egie fuel cosumptio of the rule-based method ad the ECMS for oe operatio cycle, the ECMS achieved a fuel savigs of about 17.6% compared to usig rule-based method. The ECMS is able to recogize that the usig battery cosumes lower equivalet fuel, based o the cost fuctio defied. Hece, ECMS utilizes the battery much more wheever possible as show i Fig.9, resultig i lower total egie fuel cosumptio. Loadig (kw) Loadig (%) Loadig (%) SOC (%) 2 Load demad 1 2 3 4 5 Geerator 1 power 6 7 8 1 5 1 2 3 4 5 Geerator 2 power 6 7 8 1 5 1 2 3 4 SOC 5 6 7 8 1.5 1 2 3 4 5 6 7 8 Time(sec) Total egie fuel cosumptio (kg) 3 25 2 15 1 5 Figure 7: Simulatio results RB ECMS 2 4 6 8 Time (sec) Figure 8: Total fuel cosumptio of RB vs ECMS SOC 1.9.8.7.6.5.4 RB ECMS.3 2 4 6 8 Time (sec) Figure 9: SOC of RB vs ECMS At the ed of the operatio cycle, the SOC of the battery is charged to the origial SOC for rule-based cotrol, while the SOC from the ECMS is close to the miimum limit. Assumig that the egie is chargig the battery to its origial SOC at the optimal loadig, this will result i a additioal 25% of fuel cosumptio. The fuel cosumptio of the ECMS will be comparable to the rule-based strategy ad savigs i fuel cosumptio will ot be as sigificat. However, this amout of eergy ca be supplied by the shore power whe the vessel is berthed. Oe of the mai 687
beefits of havig a hybrid all-electric system for the tugboat is the possibility to utilize shore power that is cheaper ad does ot produce emissios. I this case, the use of shore power is maximized, which will result i reductio of operatioal cost. Aother advatage of the ECMS is that detailed operatioal kowledge ad experiece from operators is o loger required to achieve the optimal fuel cosumptio. It is worthy to ote that a possible limitatio of the proposed power maagemet strategy is that the battery might be discharged close to the miimum SOC limit before the peak load demad. This will lead to a isufficiet power for the peak load demad, as the egies are ot sized to supply for the full peak load demad. Possible solutios iclude icreasig the size of the egies such that egies ca provide the full peak load i the case where batteries are discharged to its miimum limit. For a existig system, a alterative is to charge the batteries durig the stadby period before the peak load demad of the operatio. This will compromise the fuel ecoomy, but esure the reliability of the power system. Future work ca look ito idetifyig the operatio mode ad adjustig the battery SOC accordigly to esure sufficiet power durig peak demad. V. CONCLUSION Power maagemet strategies for the hybrid all-electric vessel poses a greater challege ad complexity as compared to lad-based vehicles. I this paper, ECMS is proposed for hybrid all-electric tugboat applicatio. The proposed ECMS carefully cosiders the fuel cosumptio coversio factor i order to avoid low egie loadig. Case study has show potetial of up to 17.6% fuel savigs as compared to covetioal rule-based strategy with the utilizatio of shore power. I additio, the geerators are maitaied ear optimal operatig poit throughout the operatio. Future work icludes lookig ito methods to esure sufficiet power durig the peak load demad. I additio, a detailed model of the battery that cosiders the trasiets ad temperature effects, as well as the dyamics of the power geeratio sources ca be further ivestigated. ACKNOWLEDGMENT The authors would like to ackowledge ABB Pte. Ltd. for their strog support i providig professioal idustrial kowledge ad techical assistace. REFERENCES [1] Iteratioal chamber of shippig. (214). Shippig, world trade ad the reductio of CO 2 emissios - Uited atios framework covetio o climate chage (UNFCC). Retrieved from: http://www.ics-shippig.org/docs/defaultsource/resources/evirometal-protectio/shippig-world-trade-adthe-reductio-of-co2-emissios.pdf?sfvrs=6 [2] Iteratioal Maritime Orgaizatio. (214). Reductio of GHG emissios from ships. Third IMO GHG study 214 Fial report. Retrieved from: http://www.iadc.org/wp-cotet/uploads/214/2/mepc-67-6-inf3-214-fial-report-complete.pdf [3] Iteratioal Maritime Orgaizatio. (211). Prevetio of air pollutio from ships. Retrieved from: http://www.imo.org/e/ourwork/eviromet/pollutioprevetio/ai rpollutio/pages/air-pollutio.aspx [4] World Buker Idex. (215, Dec). [5] Keedy L, Soog J, & Lidtjør J O. (213). All-electric Hybrid Propulsio for the Next Geeratio of Tugboats, Tugology, Hamburg [6] Cha, R.R., Keedy, L, Odegaard, S, (215). Aalysis of Fuel Efficiet All-Electric Tugboats. Tugology [7] Su, J. (215). Optimizatio-based cotrol for electrified vehicles: challeges ad opportuities. Joural of Cotrol ad Decisio, 2(1), 46-63. [8] Li, C. C., Peg, H., Grizzle, J. W., & Kag, J. M. (23). Power maagemet strategy for a parallel hybrid electric truck. Cotrol Systems Techology, IEEE Trasactios o, 11(6), 839-849. [9] Hou, J., Su, J., & Hofma, H. (214, Jue). Mitigatig power fluctuatios i electrical ship propulsio usig model predictive cotrol with hybrid eergy storage system. I America Cotrol Coferece (ACC), 214 (pp. 4366-4371). IEEE. [1] Opila, D. F., & Solomo, L. (212, July). Optimal cotrol of dyamic pulse power loads i aval power systems usig the Potryagi miimum priciple ad Dyamic Programmig. I Power ad Eergy Society Geeral Meetig, 212 IEEE (pp. 1-7). IEEE. [11] Seeumai, G. (21). Real-time power maagemet of hybrid power systems i all electric ship applicatios (Doctoral dissertatio, The Uiversity of Michiga). [12] Mih, T.V., Tjahjowidodo, T., Ramo, H., & Va Brussel, H. Cotrol of a Peumatic Artificial Muscle (PAM) with Model-Based Hysteresis Compesatio, I It. Cof. o Advaced Itelliget Mechatroics, Sigapore, 29 (pp. 182-187). IEEE/ASME. [13] Do, T.N., Tjahjowidodo, T., Lau, W.S & Phee, S.J. (214). A Ivestigatio of Frictio-Based Tedo Sheath Model Appropriate for Cotrol Purposes, Mechaical Systems ad Sigal Processig, 42(1), 97-114. [14] T. L. Vu, J. S. Dhupia, A. A. Ayu, L. Keedy, ad A. K. Adaes, Cotrol Optimizatio for Power Distributio System of Electric Tugboats: Determiistic Load Demad, Iteratioal Symposium o Flexible Automatio, Awaji-Islad, Hyogo, Japa, 14-16 July 214. [15] Vu, T. L., Dhupia, J. S., Ayu, A. A., Keedy, L., & Adaes, A. K. (214, Jue). Optimal power maagemet for electric tugboats with ukow load demad. I America Cotrol Coferece (ACC), 214 (pp. 1578-1583). IEEE. [16] Vu, T.L.; Ayu, A.A.; Dhupia, J.S.; Keedy, L.; Adaes, A.K., "Power Maagemet for Electric Tugboats Through Operatig Load Estimatio," Cotrol Systems Techology, IEEE Trasactios o, vol.pp, o.99, pp.1,1 [17] Pagaelli, G., Delprat, S., Guerra, T. M., Rimaux, J., & Sati, J. J. (22). Equivalet cosumptio miimizatio strategy for parallel hybrid powertrais. IVehicular Techology Coferece, 22. VTC Sprig 22. IEEE 55th (Vol. 4, pp. 276-281). IEEE. [18] Sciarretta, A., Back, M., & Guzzella, L. (24). Optimal cotrol of parallel hybrid electric vehicles.cotrol Systems Techology, IEEE Trasactios o,12(3), 352-363. [19] Serrao, L., Oori, S., & Rizzoi, G. (211). A comparative aalysis of eergy maagemet strategies for hybrid electric vehicles. Joural of Dyamic Systems, Measuremet, ad Cotrol,133(3), 3112. [2] Khodabakhshia, M., Feg, L., & Wikader, J. (213, Jue). Improvig fuel ecoomy ad robustess of a improved ECMS method. I Cotrol ad Automatio (ICCA), 213 1th IEEE Iteratioal Coferece o(pp. 598-63). IEEE. [21] Rodatz, P., Pagaelli, G., Sciarretta, A., & Guzzella, L. (25). Optimal power maagemet of a experimetal fuel cell/supercapacitor-powered hybrid vehicle. Cotrol Egieerig Practice, 13(1), 41-53. [22] Xu, L., Yag, F., Cao, G., Li, J., Lu, L., & Ouyag, M. (21). Equivalet cosumptio miimizatio strategies of series hybrid city buses. INTECH Ope Access Publisher. [23] Motapo, S. N., Dessait, L., & Al-Haddad, K. (214). A comparative study of eergy maagemet schemes for a fuel-cell hybrid emergecy power system of more-electric aircraft. Idustrial Electroics, IEEE Trasactios o, 61(3), 132-1334. [24] Zahedi, B., & Norum, L. E. (213, September). Voltage regulatio ad power sharig cotrol i ship LVDC power distributio systems. I Power Electroics ad Applicatios (EPE), 213 15th Europea Coferece o (pp. 1-8). IEEE. [25] Dedes, E (213) Ivestigatio of hybrid systems for diesel powered ships. Uiversity of Southampto, Faculty of Egieerig ad the Eviromet, Doctoral Thesis, 323pp. 688