This Is An Electronic Reprint Of The Original Article .

4m ago
24 Views
0 Downloads
3.51 MB
15 Pages
Transcription

This is an electronic reprint of the original article.This reprint may differ from the original in pagination and typographic detail.Cerna, Fernando V.; Pourakbari Kasmaei, Mahdi; Lehtonen, Matti; Contreras, JavierEfficient Automation of an HEV Heterogeneous Fleet using a Two-Stage MethodologyPublished in:IEEE Transactions on Vehicular TechnologyDOI:10.1109/TVT.2019.2937452Published: 01/01/2019Document VersionPeer reviewed versionPlease cite the original version:Cerna, F. V., Pourakbari Kasmaei, M., Lehtonen, M., & Contreras, J. (2019). Efficient Automation of an HEVHeterogeneous Fleet using a Two-Stage Methodology. IEEE Transactions on Vehicular Technology, 68(10),9494-9506. https://doi.org/10.1109/TVT.2019.2937452This material is protected by copyright and other intellectual property rights, and duplication or sale of all orpart of any of the repository collections is not permitted, except that material may be duplicated by you foryour research use or educational purposes in electronic or print form. You must obtain permission for anyother use. Electronic or print copies may not be offered, whether for sale or otherwise to anyone who is notan authorised user.Powered by TCPDF (www.tcpdf.org)

2019 IEEE. This is the author’s version of an article that has been published by IEEE.Personal use of this material is permitted. Permission from IEEE must be obtained for all otheruses, in any current or future media, including reprinting/republishing this material foradvertising or promotional purposes, creating new collective works, for resale or redistributionto servers or lists, or reuse of any copyrighted component of this work in other works.

1Efficient Automation of an HEV HeterogeneousFleet using a Two-Stage MethodologyFernando V. Cerna, Mahdi Pourakbari-Kasmaei, Member, IEEE, Matti Lehtonen, and JavierContreras, Fellow, IEEENOMENCLATUREA. Sets and indexesk/i,d,e,u Indexes stand for intersections i/k, delivery, HEVtechnology, and density value, respectively.op/ps Indexed of operation, and operating strategy,Set of urban roads ki and intersections i of roads.Set of deliveries points d and HEV technologies e.Set of operations s and operating strategies v.Set of density values u.B. ParametersAutonomy value of HEV e (km).Linear battery degradation cost-intercept parameter.Battery capacity of the HEV e (kWh).Battery lifetime in years.Capacity of HEV e (quantity of goods that can betransported by HEV e).Battery cost ( /kWh).Capacity fade at end of life.Length of road ki (km), and the corresponding trafficdensity (veh/km).Value of the traffic density of element u (veh/km).Saturation density value (veh/km).Optimal density value related to(veh/km).CO2 emissions of HEV e (gCO2/km).Maximum traffic flow value (veh/h).Maximum traffic flow value on main roads (veh/h).Maximum traffic flow value on secondary roads(veh/h).Fuel capacity (L).Linear battery degradation cost-slope parameter.Big value used as part of the constraints linearized., , , Hyper-matrix related to the deliveries d in operatingstrategy v of the operation s to be performed by HEV e.Hyper-matrixrelated to the objective function values of, ,operating strategy v for operation s of HEV e.Total number of points to be visited (delivery points andwarehouse).Charging rate of HEV e (kW)./𝑃Probability values related toat main/secondaryroads, respectively./𝑃Accumulated probability value related toor.Type of intersection i (1: if the intersections has acharging station, otherwise 0).Initial SoC of the battery of HEV e (kWh).Indicates the type of intersection i at delivery d.,Indicates the type of road ki (1: main; 0: secondary).Quantity of goods to be demanded in operation s.Maximum average speed value of road ki (km/h).Free flow speed value (km/h).Optimal speed value related to(km/h).𝑣 /𝑣 Optimal speed related toor(km/h).𝛿 , 𝛿 Weighted weights related to total navigation time, andbattery lifetime.Extra units cost ( /units).Parameter for defining the nodes status of , (-1:starting node; 0: intermediate node; 1: arrival node).Fuel consumption rate (km/L).Percentage of CO2 emissions reduction.F. V. Cerna is with the Department of Electrical Engineering, FederalUniversity of Roraima (UFRR), Av. Cap. Ene Garcês, n 2413 - Aeroporto,Boa Vista - RR, 69310-000 (e-mail: [email protected]).Mahdi Pourakbari-Kasmaei, and Matti Lehtonen are with the Department ofElectrical Engineering and Automation, Aalto University, Maarintie 8, .fi,Matti.Lehtonen}@aalto.fi).Javier Contreras is with the E.T.S. de Ingenieros Industriales, University :[email protected]). Abstract—An influential factor in enhancing the attendanceservices, mainly in commercial and emergency sectors, is thevehicular technology used to transport people, goods, orequipment. Although hybrid electric vehicles (HEVs) represent asustainable transport alternative, the existing technical limitationssuch as battery and fuel capacities, and autonomy, among others,highlight the provision of an efficient automation tool. The tool canserve to enhance the operational performance of the HEV byselecting the proper driving mode (on fuel or electricity), and thenavigation strategies to the delivery and charging points in urbanareas. This paper proposes a two-stage methodology that allowsthe HEVs operators to automate the operational performance of aheterogeneous HEV fleet on a city map. Each stage is handled byits corresponding optimization model. In the first stage, the totalnavigation time and the battery lifetime of the fleet during theoperation are optimized. In this stage, constraints related tocharge-sustaining/charge-depleting modes, state of charge (SoC)of the HEVs battery, and deliveries schedules are taken intoaccount. To this end, operating strategies related to theperformance of different types of existing HEV technologies areanonymously considered. In the second stage, the best operatingstrategy among all the operating strategies is selected whileconsidering the capacity of HEVs to deliver a given quantity ofgoods. Moreover, uncertainties during the HEV navigation aresimulated considering the change in traffic density of the urbanroads as a function of the levels of service (LOS). Results show thatthe proposed methodology establishes an efficient operationalscheme for a HEVs fleet, ensuring a significant reduction of energyusage as well as mitigating the CO2 emissions.Index Terms—Attendance services, Battery charging, HEVheterogeneous fleet, operating strategies.,

2Percentage of battery capacity, used to calculate theSoC of HEV e returning to the warehouse.Rate of energy spend by HEV e (kWh/km).𝜏 , 𝜏 Charging time variation interval (h).𝜂 /𝜂 Charging/discharging efficiencies values.C. VariablesDecides the travel of HEV e in road ki for delivery d., ,/Length of road ki that HEV e travels in CSM/CDM for, ,delivery d., , SoC of the battery of HEV e at the intersection , atdelivery d (kWh)., , SoC of the battery of HEV e when arrives at theintersection i at delivery d (kWh).Energy for charging of HEV e at intersection i for, ,delivery d (kWh).Denotes the charging time of HEV e at charging station, ,located in the intersection i during delivery d (h).Stands for the charging status of HEV e at intersection, ,i during delivery d.and , , at the, ,, , Represents the product oflinearized constraints.Represents the product ofand, ,, , at the, ,linearized constraints.Represents the product ofand, ,, ,, , at thelinearized constraints.Decides which operation s is made by HEV e via the, ,operating strategy v.Total extra quantity of goods.I. INTRODUCTIONHE electricity demand by the transportation sector willincrease sharply after 2020 due to the projected increase insales of new electric vehicles (EVs) and hybrid electric vehicles(HEVs) [1]. In this context, service sectors, mainly commercial(e.g., postal, merchandise, home care, etc.) and emergency(e.g., ambulances, fire engines, police vehicles, energycompany cars, etc.), can have a positive influence, since thequality of their attendance depends largely on the vehiculartechnologies used to transport people, goods and/or equipmentin urban areas [2]. These services present several challengessuch as minimizing delay times, managing fuel consumption,selecting the most appropriate route, etc., thereby affecting thevehicle fleet performance. These challenges highlight thecrucial role of electric vehicles and the operational andsustainable aspects to be considered within an efficientautomation strategy during their operation [3], [4].Studies show that among EV-based technologies, the HEV isa promising alternative in the transportation sector, mostly dueto socio-environmental factors [5], [6]. Basically, HEVs presenttwo driving modes during navigation such as 1) ChargeSustaining Mode (CSM), and 2) Charge-Depleting Mode(CDM). In CSM, the internal combustion engine drives theHEV, whereas, in CDM, the electric energy of the battery isused for driving purposes [7], [8]. If a fleet of HEVs is used totransport goods, their capacity and performance aredetermining factors that must also be considered in theautomation of this particular type of service. Therefore,depending on the quantity of goods, location of supply pointsand scheduling of operations, proper technology should beassigned. This reveals the significant role of smart tools thatTensure the efficient automation of an HEV heterogeneous fleetconsidering the optimal CSM/CDM selection, optimal batterymanagement, optimal deliveries scheduling, and environmentalissues. Also, the uncertainties due to the variation in trafficdensity for each road on the urban map, as well as speed limitson roads and sustainable performance of HEVs should beconsidered in the automation tool [9], [10].In the literature, most of the existing works related to HEVfleets merely consider the optimal charging or fuel consumptionmanagement. In [6], to fulfill the economic criteria ofautonomous EVs charging in a city with a predefined spatiallimit, a Monte Carlo simulation-based approach was proposed.A mixed-integer linear programming (MILP) model wasdeveloped in [11] to coordinate the charging and power storagein the battery of the HEV fleet. In order to optimize the chargingprofile in the predefined charging points, an energy managementsystem for EV fleet operators was proposed in [12]. In [13], asensitivity analysis has been done to find the optimal chargingprocess of the EV fleet and the flow exchange between thepower grid and EVs. In [14], optimal recharging of EVsconsidering the different habits of owners was a part of a smartmicrogrid project in order to manage energy usage, fuel costs,and carbon dioxide emissions, while, as a storage device,addressing the fluctuations of renewable energy output. In [15],a predictive management system was proposed for optimalcharging of the HEVs fleets while the charging station wasequipped with PVs taken into account the distribution networkrestrictions. Decentralized control strategies and mathematicalprogramming models for the charging of an EV fleet weredeveloped in [16] and [17]. In [18], an adjustable robustoptimization model considering a set of charging stations, travelcosts, and battery capacities was investigated. In [19], [20],aiming at minimizing the charging costs, proper policies andcharging strategies to determine the optimal schedulingconsidering historical data were developed. To fairly addressthe optimal scheduling of EVs, the economic charging andbattery degradation were co-optimized in [21] via the Paretofront technique. The EV charging rates were modeled viapartial differential equations under three conditions: 1) EVsreceive energy from the grid, 2) EVs are connected to the gridbut not charging, and 3) EVs deliver energy to the grid, wasconsidered in [22]. Optimal charging scheduling of taxi fleetswas presented in [23] while taking into account the chargingstation locations, unpredictability and balancing issues.Although the aforementioned works highlight the importanceof EV charging scheduling, energy management of both EVfleets and charging points and their impact on the electricitygrid, both the route as well as the uncertainties involved duringnavigation of each EV were not part of their focus, while thenavigation modes, emissions reduction during navigation, andthe EV type (heterogeneous fleet) were also neglected.In order to fill the aforementioned existing gaps, theperformance of EVs and HEVs has been investigated in someworks. The integration of thermal power generation units as asupport system for PHEV charging has been studied in [24]aiming at minimizing the emissions, especially during peakdemand periods. A comparative study on the fuel consumptionand emission output of conventional HEVs and plug-in HEVs

3were conducted in [25] while taking into account the interactionbetween the energy storage system, electric machine, powercontrol unit and internal combustion engine. An energymanagement and vehicle control models were proposed in [26]to simultaneously reduce the fuel consumption and pollutingemissions of the HEVs by taking into account the informationrelated to a city road network. In [27], a neural network-baseddynamic online programming strategy was used to minimizethe fuel consumption of PHEVs where the real-timeinformation, i.e., traffic and control signal, was interchangedbetween the vehicles and the control center. A managementstrategy based on the performance of Toyota Prius wasinvestigated in [28] to control the pollutant effects of CO2emissions. Internal power flow and its efficient management inthe HEV powertrain were analyzed in [29] and [30]. In [29], thelimits of the power grid and the waiting and charging times byHEV were considered, while in [30], travel speed, energy level,and stop-and-go frequency were addressed. However, still,there is a lack of smart tools for an effective route