Automotive Innovation

, Volume 1, Issue 1, pp 62–69 | Cite as

Design of a Regenerative Auxiliary Power System for Service Vehicles

  • Yanjun Huang
  • Milad Khazeraee
  • Hong Wang
  • Soheil M. Fard
  • Tianjun Zhu
  • Amir Khajepour


This paper presents a RAPS, namely the regenerative auxiliary power system, for the vehicle with special accessory power systems. Taking city buses and delivery trucks as examples, service vehicles keep engines idling to power their auxiliary devices (e.g., refrigeration systems and hydraulic pumps). The potential fuel savings brought on by the electrification of these auxiliary systems are first quantitatively analyzed over a typical drive cycle for a delivery truck. The RAPS is then designed, and its components are sized in accordance with the objectives of compactness and cost-effectiveness. By introducing the proposed RAPS into a conventional delivery truck with an internal combustion engine, the powertrain can be treated as a hybrid because of adding an extra battery. As a result, to pursue a high overall efficiency, a holistic controller is presented for determining how and when to recharge the battery while minimizing the auxiliary system’s power consumption. More importantly, the proposed RAPS saves about 7% fuel when compared with consumption by conventional service vehicles.


Anti-idling Component sizing Auxiliary system electrification Power management control 

1 Introduction

Different accessory devices, e.g., air-conditioning or refrigeration (A/C–R), are being widely used in modern service vehicles [1]. These devices usually account for a part of the total engine load as well as the consumed fuel, which cannot be ignored. Taking a refrigerated delivery truck as an instance, the A/C–R device consumes up to one-fourth of the total tank fuel and more for long-haul vehicles [2]. It is often the case that engines have to idle for these auxiliary systems to maintain power for their special functions when the vehicle stops. Diesel engines usually have up to 40% efficiency, when idling however the efficiency drops to 1–11% with more pollutants. That is why, in many countries, the issued bylaws [3] put more strict regulations on engine idling. As a result, it is urgent to take measures by researchers in both automotive academia and industry in anti-idling. The development of energy storage systems (ESSs) in electric vehicles (EVs) and hybrid electric vehicles (HEVs) allows the auxiliary systems to be powered independently [4, 5]. That means the electrification of the auxiliary devices in vehicles is possible when using such ESSs. In this way, the idling due to powering auxiliary devices when the vehicle is stopped can be eliminated such that the performance and efficiency of the automotive system are greatly improved. Even though many types of anti-idling products have been introduced to reduce this kind of idling in recent years, there are still many improvements that need to be made. As a result, the proposed optimum regenerative auxiliary power system (RAPS) not only satisfies the auxiliary power device requirements to realize anti-idling but also makes good use of recovered braking energy to further improve energy efficiency.

In the proposed RAPS indicated in Fig. 1, the ESS (i.e., battery) is integrated into a delivery truck powertrain and it is able to individually power an A/C–R device. The RAPS has the ability to recollect and save one part of the kinematic energy when braking in the battery by using an alternator or an added generator. The alternator is assembled with engine via a serpentine belt, while the generator is bolted to the transmission through a power take-off (PTO). Meanwhile, the battery can be smartly recharged either by the engine or plug-in electricity when required by a holistic power management controller (PMC). Therefore, the main advantage of the battery is it makes use of the regenerated or engine energy in an efficient manner and thus electrifies the A/C–R device. In this way, the RAPS distinguishes from its existing counterparts (e.g., auxiliary battery powered systems (ABPs) or auxiliary power units (APUs)).

The components’ size of the RAPS has to be optimized because it is designed and installed in a conventional vehicle. Therefore, a multidisciplinary design optimization (MDO) strategy is thus used to size the main components. At the same time, after introducing the optimized RAPS to a delivery truck, the new powertrain changes into a parallel-like hybrid owing to adding an extra ESS, whereas the fact that the added battery solely drives the A/C–R system instead of powering the vehicle makes it different from the common parallel hybrid powertrain. However, a PMC is required to determine how and when to recharge the battery. In this study, the auxiliary system is the refrigeration system of a delivery truck. A holistic controller is developed to incorporate the power management strategy and the energy-saving controller of the refrigeration system.
Fig. 1

The schematic of a powertrain with a RAPS

The remaining of this study is structured as follows. A state-of-the-art study on existing anti-idling products is presented in the next section. The potential fuel savings brought on by the electrification of the auxiliary system are then quantitatively analyzed. In addition, modeling and component sizing using MDO are presented and then the PMC is elaborated upon. Furthermore, the benefits introduced by the presented RAPS as well as the designed holistic MPC are demonstrated in a case study. Finally, conclusions and future work relating to the study are discussed.

2 Literature Review

Engine idling is the main contributor to air pollution, noise pollution as well as the health problems. It is therefore urgent to pay attention to eliminate idling. Many studies [6, 7] have quantitatively demonstrated the negative effects of idling, and many related bylaws have been introduced to mitigate these effects [3]. Owing to a pressing demand on lower pollution and higher fuel economy of the transportation sector, a numerous anti-idling technologies have already come up recently [8]. According to the current literature, two classification methods have been adopted to sort these technologies. On the one hand, from a perspective of location, products are referred to the mobile or stationary [9] as shown in Fig. 2; on the other hand, based on their functionality, products are defined as the partially functional and fully functional [10].

Located on vehicles, the mobile type includes APUs, ABPs, automatic engine shutdown devices, and other parts developed to perform specific functions, like the heating of a sleeper compartment and cooling. Among them, ABPs and APUs are the fully functional type because they are able to fully supply the required power, while others refer to the partial functional type because of their limited provisional abilities (e.g., either cooling or heating). Meanwhile, from where truck drivers could acquire services, e.g., electricity, heating/cooling, or the Internet, stationary products also refer to truck stop electrification (TSE) systems and are categorized as onboard and off-board products. The former (e.g., CabAire LLC and Shorepower Technology) requires installing the cooling or heating systems and accessories in the vehicle, whereas the latter (e.g., Envirodock, IdleAire, and AireDock) offers electricity, cooling, and heating via an external device.
Fig. 2

Anti-idling product types

In this section, only products similar to that proposed in this paper are reviewed; for other product reviews, please refer to the literature [11]. APUs, units that include a relatively small engine and a generator totally assembled in a truck’s original HVAC, are the most classic and popular solutions because they fully provide the needed auxiliary power to reduce engine idling. However, the structure of an externally extra engine and alternator results in vehicles becoming more expensive, heavy, noisy, and requires more maintenance. Moreover, the extra engine generates more emissions if designed improperly [12], and the added weight negatively affects the fuel economy. Having emerged recently as a competitive alternative, ABPs replace their counterparts in conventional APUs with a pack of batteries and possess the same features but without noise and emissions. The onboard ESS is recharged by engine when running or by the stationary-type systems, and discharged, while the auxiliary systems should stay on and the vehicle stops. However, ABPs have inherent problems with regard to their batteries, e.g., a short lifecycle and associated costs. Consequently, alternative technologies, such as solar energy system and fuel cells, have been proposed, but they have not been popularized because of the requirements for excessive modifications, expensive materials, and a long start-up time. However, GM launched a hybrid electric military truck in 2003 with a fuel-cell APU and a diesel engine. Mercedes-Benz also proposed hybrid electric trucks 2004, and many auto-companies subsequently launched their own hybrid buses and trucks. Famous for designing and manufacturing commercial vehicles, Eaton is providing both hydraulic and electric hybrid powertrains. The M2 106 hybrid electric trucks of Freightliner utilize a customized electronic PTO (ePTO) to eliminate engine idling, and thus, it makes trucks appropriate for long-idling scenarios, e.g., tree trimming. The total idling can be decreased by almost 87%. Truck owners can optionally add a 5-kW auxiliary power generator (APG) for an extra saving as well. Other motor manufacturers, e.g., Hino Motors, UD Trucks, and Mitsubishi Fuso Truck & Bus Corporation, have designed their own exclusive hybrids, adopting similar technologies to reduce engine idling.

In summation, the optimized RAPS has major advantages over its counterparts. Firstly, the optimized RAPS is cleaner and quieter than existing APUs because of no additional engine. Secondly, the optimized RAPS is capable of recapturing braking energy despite its similarity to an ABP; thus, a relative small ESS is able to easily satisfy the requirements and only necessities a small engine power to charge, resulting into a lower cost. In addition, a designed holistic PMC will supervise the RAPS working at its maximum efficiency. Furthermore, due to the extremely high costs, hybrid service vehicles are still not being widely used. For example, a 40-foot hybrid bus usually meanly costs $450,000–$550,000 and $280,000–$300,000 for an engine-powered bus [13]. Therefore, engine-powered vehicles would continue to be dominant in the near future. However, current when developing hybrid electric vehicles, their accessory power is usually assumed as a fixed value or even ignored, leading to a suboptimal or even non-optimal solution. The method or procedure used in developing RAPS can be easily transferred or directly used in the development of any type of hybrid vehicle.

3 Potential Fuel Saving by Electrifying the Auxiliary System

This section quantitatively studies benefits in terms of fuel savings brought upon by the electrifying of the auxiliary devices for a conventional truck. In this research, our target vehicle is the GMC Savana 2500, whose specifications are given in the literature [14]. To study the vehicle’s fuel economy, a drive cycle is created by combining several standard drive cycles as shown in Table 1 and Fig. 3.
Table 1

Created driving cycle with each segment


Duration (s)

















This drive cycle can be used to represent daily driving information of a typical delivery truck. At 10:00 am, the truck is loaded for about 15 min at a warehouse and then driven to two retail stores for delivery via a standard highway and city driving. However, to keep the goods (e.g., vegetables and fruit) fresh, the refrigeration system should always be working regardless of whether the truck is being loaded or unloaded or is moving. During loading/unloading periods, the conventional truck is not moving, but the engine should be kept on to power the refrigeration systems. Therefore, compared with the engine idling when waiting for red lights, idling during loading/unloading periods lasts much longer and discharges more pollutants.
Fig. 3

Combined drive cycle

Table 2

Driving power and auxiliary power

Drive cycle


The proposed drive cycle

Average driving (auxiliary) power (kW)

17.15 (2)

21.53 (2)

For the simplicity of the study, the nominal power of the refrigeration system is assumed to be 2 kW. The average driving power and auxiliary power are presented in Table 2, where the average driving powers of the vehicle in a typical urban drive cycle (UDDS) and the proposed drive cycle are compared with the power consumed by the auxiliary device. The comparison shows that the auxiliary power is relatively low. This type of vehicle is thus referred to as a light service vehicle.

The use of the model, where the simulations are performed to calculate the fuel consumption over the proposed drive cycle, will be discussed in the following section. Compared with the conventional service vehicle, the vehicle with the electrified auxiliary system has the potential for 11% fuel consumption savings (Table 3).
Table 3

Fuel consumption comparison over the proposed drive cycle


The conventional service vehicle

The vehicle with the electrified auxiliary system

Fuel consumption (L)



To calculate fuel consumption and the potential savings when the auxiliary system is electrified, a battery is used to store the regenerative braking energy and electrify the auxiliary system. Meanwhile, the fuel consumption heavily depends on the driving information. The results are derived from the scenario with the drive cycle shown in Fig. 3.

For large potential fuel savings, it is beneficial to propose a product that can electrify the auxiliary devices of the service vehicles and reduce unnecessary idling.

4 Modeling and RAPS Component Sizing

The system model will also be used by the optimization algorithm to find the optimal component size and proper power management strategy. Component models should be generic, modular, and flexible such that they can be easily modified by the optimizer. In this study, a backward-looking modeling approach is used to reduce the unnecessary dynamic details of the components and vehicle system, which leads to less computational time. The detailed modeling process of the RAPS components has been presented in the literature [14].

As illustrated in Fig. 4, the optimization algorithm updates system models according to vehicle specifications. The algorithm uses the drive cycle, service cycle (i.e., the auxiliary power profile), and design variable candidates (i.e., component size and proper charging strategy variables) as inputs. Using the updated model and the inputs, system operation is simulated to calculate the energy consumption of the vehicle in the form of fuel and electricity consumption. The results of this simulation will be transferred to objective function J to evaluate the total system’s cost. According to changes in the total system’s cost (objective function), the algorithm chooses an updated set of variables in the feasible domain and the whole process is repeated until it reaches the optimal variables.
Fig. 4

Optimization algorithm process

The design challenge lies in balancing the size and cost of the battery and other added parts against the cost of the engine’s fuel and plug-in electricity consumption for running the service devices. The objective function J is based on the total cost of the system while considering the operating costs (e.g., fuel and electricity) and initial costs of added accessories and battery packs. It is defined as
$$\begin{aligned} \begin{array}{l} J=\left( {\begin{array}{l} \mathrm{Fuel}_\mathrm{consumed} \times \mathrm{Fuel}_\mathrm{cost} \\ \quad +\,\mathrm{Plug}\_\mathrm{in}_\mathrm{energy} \times \mathrm{Electricity}_\mathrm{cost} \\ \end{array}} \right) \times \mathrm{Days}_\mathrm{active} \\ \qquad \times \, \mathrm{Years}+\mathrm{Battery}_\mathrm{banks} \times \mathrm{Battery}_\mathrm{cost} + \mathrm{Accessory}_\mathrm{cost} \\ \end{array}\nonumber \\ \end{aligned}$$
Table 4

Description of objective function variables

\(\mathrm{Fuel}_\mathrm{consumed} \)

Total fuel consumption over one complete drive cycle

\(\mathrm{Fuel}_\mathrm{cost} \)

The price of fuel per liter

\(\mathrm{Plug}\_\mathrm{in}_\mathrm{energy} \)

Total electrical energy kWh from plug-in

\(\mathrm{Electricity}_\mathrm{cost} \)

Unit price of purchased kWh electricity

\(\mathrm{Days}_\mathrm{active} \)

The number of working days of service vehicles


Targeted lifespan for the battery pack (5 years)

\(\mathrm{Battery}_\mathrm{banks} \)

The number of battery banks in EES

\(\mathrm{Battery}_\mathrm{cost} \)

The price of each battery bank

\(\mathrm{Accessory}_\mathrm{cost} \)

Total costs of added parts (except EES parts)

All the variables in J and their values in this study are explained in Table 4. In this research, several models of lithium-ion (A123_ALM12V7, A123_ ANR26650, GBS_100 Ah) and lead–acid (EV12_140X, EV12_180X, EV12_8DA_A) batteries are studied for selecting the optimal option for designing RAPS. According to the life cycle of regular battery packs in electric and hybrid vehicles, it is assumed that the battery pack can last for almost 5 years (in working condition for the service vehicle, a 260-day work year) without the need for replacement due to degradation.
Fig. 5

Drive cycles and prediction points

Using the model and an objective function to maximize the return of the RAPS over a 5-year period, the size of the components for the targeted service vehicle, the GMC Savana 2500, is calculated. When considering the input drive cycle in Fig. 3, the optimization algorithm results obtained by a genetic algorithm show that the optimal solution (i.e., the solution having the lowest total system cost or greatest savings) is achieved when two packs of the EV12_8DA_A Discover Dry Cell battery are used. The nominal voltage is 12 V, and the capacity is 520 Ah. The optimal charging strategy also prevents the batteries from reaching 30% depth of discharge. For the optimal RAPS, the initial investment will be returned after 2–3 years, which is half the expected lifetime of the batteries. Further information is provided in the literature [14].

5 Holistic Controller Development for the RAPS

On the basis of the above model and parameters, the MPC-based holistic controller is developed to optimize battery charging and to optimize the energy consumed by the A/C–R system. Different advanced control strategies have been used in vehicle systems in recent studies [15, 16]. However, the slow dynamics of the overall system are suited to the real-time application of the MPC. The overall process of developing the holistic controller is described in the literature [2]. Figure 5 presents the real-world and the nominal cycles of a service vehicle. Owing to the aforementioned factors, the real cycle can still follow the nominal cycle but is not in full agreement with the nominal one. Since the future driving information is unknown, the only accessible knowledge for prediction is the nominal drive cycle. Once the step size for prediction is enlarged during developing MPC, the effects on state prediction exerted by the negative and positive deviations from the nominal cycle will be counteracted, which will render the two-state prediction trajectories similar. In addition, the prediction horizon will be enlarged to bring the solutions and global optimal solutions closer with no extra computational cost incurred. To show the advantages of the presented holistic MPC, the step size for prediction is set at 10 s and its performance will be verified against the prescient MPC [10].

Once the optimal trajectory is acquired, only its first element is fed into the actuator to operate the plant for the next time instant. The MPC then utilizes the fresh data to repeat the whole process. The demonstration of the RAPS with the holistic controller is shown in Fig. 6. The proposed holistic controller plays four major roles of guaranteeing sufficient energy in the battery for all engine-off conditions, commanding regenerative braking, minimizing the power consumption of the refrigeration system, and determining whether and when the engine charges the battery to maximize the overall efficiency.
Fig. 6

Structure of the RAPS with a holistic controller

6 Case Study

In order to verify the benefits bought on by the holistic controller and the proposed RAPS, a case study is preceded in this section. To simulate factors of real-world driving conditions, such as traffic and lights, a real driving cycle is created by switching two segments of the nominal drive cycle shown in Table 5 or by adding about 15% white noise.
Table 5

Real drive cycle segments


Duration (s)

















Figure 7 shows the 2.5-h real-world drive cycle, and the amplified vehicle velocity is shown in the bottom subfigure. The service cycle indicates the consumed power of the refrigeration system. For delivery trucks, the refrigeration system as the main auxiliary device consumes much more than the remaining accessory devices. The power of the refrigeration system varies with the operating conditions and ambient. Using the same method as drive cycle, the real service cycle with disturbances is captured as shown in Fig. 8 including the heating load (HL) and ambient temperature. The HL is denoted proportional to the temperature difference. Besides, there is an extra HL owing to the frequent door opening applied onto the existing HL to form the nominal one, which is assumed 0.1 kW.
Fig. 7

Nominal and real drive cycles

Fig. 8

Ambient temperature and HL

In this section, to show the fuel savings of the new vehicle, the proposed RAPS and the holistic controller are compared with a traditional vehicle equipped with an on/off refrigeration system. At the same time, two MPCs are studied and extensively compared for verifying the robust characteristic of the designed holistic MPC. The prescient MPC detonated by MPC\(_{1}\), which is aware of the real-world service and drive cycle beforehand, while the other (MPC\(_{2})\) only has access to the nominal service drive cycle. To prolong battery’s life span and protect the battery, its SOC should be kept within limitations. Besides, the battery supplies the power to the refrigeration system to eliminate idling so the SOC is better to be maintained at a higher level during the whole drive cycle. 0.9 is chosen as a SOC reference, and the weight for the SOC is set at a relatively small value to prevent the SOC from going too far from its reference. The prediction horizon must be a reasonable value to balance the optimality and prediction accuracy. As suggested in the literature [17], in order to enhance the controller’s stability, a terminal weight 10 times larger than the small SOC reference weight is selected. Meanwhile, to maintain engine working in the high-efficiency area, the item indicating engine efficiency is provided to the input weight matrix. An open-source code in milliseconds solves the formulated QP problem to make sure that this method could be used in real time.

The following figures demonstrate the comparison results. Specifically, the SOC of the added battery is shown in Fig. 9, which demonstrated that the battery is employed to drive A/C–R device without recharging from the engine, resulting in eliminating idling, while it is recharged by the regenerated energy during braking. It also shows that the battery is automatically recharged by the engine when it is working in high efficiency and the regenerated braking energy is not enough or the SOC goes too far from its reference. The proposed controller decides when to recharge the battery in accordance with the predicted behaviors. The response of the cargo temperature is provided in Fig. 10. As can be seen, the initial temperature of the cargo is assumed identical to the ambient temperature at 10:00 am when refrigeration system starts to work. Around 15 min later, the temperature arrives at its set point (10 \({^{\circ }}\hbox {C}\)) and then maintains even when additional HL appears.
Fig. 9

SOC trajectories

Fig. 10

The response of cargo temperature

Refrigeration system input actions obtained by controllers are presented in Figs. 11, 12, and 13. As can be seen, the initial temperature is much higher than the setpoint. To cool the cargo, the refrigeration system runs in full-load mode. The input actions start to change in accordance with the outside HL after the temperature arrives at its set temperature. From Fig. 14, we can see that, once the actual SOC deviates too far from its reference, direct charging happens from the engine in periods of high efficiency. Charging does not occur when the vehicle is not moving.
Fig. 11

Speed of the compressor

Fig. 12

Frequency of the evaporator fan VFD

Fig. 13

Frequency of the condenser fan VFD

Fig. 14

Direct charging power

The traditional delivery truck equipped with an on/off controller for its refrigeration system is adopted here as a benchmark for comparing to show savings contributed by our proposed RAPS. The on/off controller design process is given by the literature [18, 19]. The corresponding inputs of the refrigeration system are shown in Figs. 15 and  16. It can be seen that the input actions switch quickly between zero and the maximum if the temperature exceeds the limitation. The switching frequency is higher in the early morning and becomes lower at noon.
Fig. 15

Temperature response of the cargo by using an on/off controller in the traditional delivery truck

Fig. 16

Refrigeration system inputs in the conventional vehicle

Table 6

Comparison of MPCs

Traditional and on/off controller


FC (L)

SAVE (%)






\(\hbox {MPC}^{1}\)




\(\hbox {MPC}^{2}\)




Results of fuel consumption are compared in Table 6. A SOC-correction approach [11, 20, 21] is employed to counteract the influence of fuel consumption causing by the difference of the final SOC. The total consumed fuel (i.e., 19.3 L) of the traditional powertrain, as well as an on/off controller, is treated as a benchmark for comparing. Employing designed RAPS with two MPCs (i.e., the prescient and the proposed one), 7.7 and 7.5% of fuel saving, is obtained, respectively. The MPC\(^{2}\), therefore, performs closely with the prescient MPC\(^{1}\), and MPC\(^{2}\) is robust against large disturbances. The saved fuel also means reduce the corresponding emissions. As a result, the results demonstrate that the electrification of the auxiliary system can provide appreciable benefits to both vehicle owners and the environment.

7 Conclusions and Future Work

This work designed a holistic controller MPC for the proposed regenerative auxiliary power system. This system was assembled in the traditional powertrain of a delivery truck for electrifying its A/C–R system to anti-idling as well as maximizing the energy efficiency of the updated powertrain. The overall structure, as well as the models of RAPS, was briefly presented. An MPC was used in the holistic controller development owing to its ability to predict and consider the constraints. In addition, the fact that most service vehicles (e.g., city buses and delivery trucks) drive along the predefined routes also facilitates using MPC. Due to the aforementioned real-world factors, the nominal drive cycle and real drive cycle will not agree with each other. That explains why the MPC using a large step size for prediction was proposed and studied. Then, the robust feature of the controller was also validated. The results of a case study show the proposed RAPS with the controller can save about 7% of the total fuel for a studied light service vehicle. Moreover, the designed MPC performs similarly with the optimal one (i.e., the prescient one) even in conditions that the predicted drive cycle differs from the real cycle. In the future, we will study the feasibility of the distributed MPC used in the RAPS and do the real test by selecting a delivery truck.



The authors acknowledge financial support from Automotive Partnership Canada (APC) and the Collaborative Innovation and Platform Environment Construction Project of Guangdong Province (2015A050502053).


  1. 1.
    Chen, H., Kessels, J.T.B.A., Weiland, S.: Vehicle energy management for on/off controlled auxiliaries: fuel economy vs. switching frequency. In: IFAC-Papers On Line, vol. 48, no. 15, pp. 217–223 (2015)Google Scholar
  2. 2.
    Huang, Y.: Anti-idling systems for service vehicles with A/CR units: modeling, holistic control, and experiments. Department of Mechanical and Mechatronics Engineering, University of Waterloo, Canada (2016)Google Scholar
  3. 3.
    Penny, J.: Situational analysis: the status of anti-idling by-laws in Canada (2005)Google Scholar
  4. 4.
    Huang, Y., Wang, H., Khajepour, A., et al.: Model predictive control power management strategies for HEVs: a review. J. Power Sources 341, 91–106 (2017)CrossRefGoogle Scholar
  5. 5.
    Tang, X., Hu, X., Yang, W., Yu, H.: Novel Torsional vibration modeling and assessment of a power-split hybrid electric vehicle equipped with a dual mass flywheel. IEEE Trans. Veh. Technol. 1 (2017)Google Scholar
  6. 6.
    Lust, E.E.: System-level analysis and comparison of long-haul truck idle-reduction technologies. ProQuest (2008)Google Scholar
  7. 7.
    Wang, Y., Byrne, J., Rickerson, W.: Investigating the cost, liability, and reliability of anti-idling equipment for trucks: final report to delaware department of transportation. University of Delaware, Center for Energy and Environmental Policy, Newark (2007)Google Scholar
  8. 8.
    Shancita, I., Masjuki, H.H., Kalam, M.A., et al.: A review on idling reduction strategies to improve fuel economy and reduce exhaust emissions of transport vehicles. Energy Convers. Manag. 88, 794–807 (2014)CrossRefGoogle Scholar
  9. 9.
    Kuo, P.Y.: Evaluation of freight truck anti-idling strategies for reduction of greenhouse gas emissions. ProQuest (2008)Google Scholar
  10. 10.
    Huang, Y., Khajepour, A., Zhu, T., et al.: A supervisory energy-saving controller for a novel anti-idling system of service vehicles. IEEE/ASME Trans. Mechatron 22, 1–10 (2016)Google Scholar
  11. 11.
    Huang, Y., Khajepour, A., Wang, H.: A predictive power management controller for service vehicle anti-idling systems without a priori information. Appl. Energy 182, 548–557 (2016)CrossRefGoogle Scholar
  12. 12.
    An introduction to AirEDock. Accessed 03 Aug 2016
  13. 13.
    Assessing the Costs for Hybrid Versus Regular Transit Buses. Accessed 03 Dec 2016
  14. 14.
    Khazraee, M., Huang, Y., Khajepour, A.: Anti-idling systems for service vehicles: modeling and experiments. Proc. Inst. Mech. Eng. Part K J. Multibody Dyn. (2017, in press)Google Scholar
  15. 15.
    Zhao, F., Dong, M., Gu, L., Guan, J.: Adaptive neural-sliding mode control of active suspension system for camera stabilization. Shock Vib. 2015, 1–8 (2015)Google Scholar
  16. 16.
    Qin, Y., Langari, R., Gu, L., Guan, J.: Adaptive hybrid control of vehicle semiactive suspension based on road profile estimation. Shock Vib. 2015, 1–13 (2015)CrossRefGoogle Scholar
  17. 17.
    Fagiano, L., Teel, A.R.: Generalized terminal state constraint for model predictive control. Automatica 49(9), 2622–2631 (2013)MathSciNetCrossRefzbMATHGoogle Scholar
  18. 18.
    Huang, Y., Khajepour, A., Bagheri, F., et al.: Modelling and optimal energy-saving control of automotive air-conditioning and refrigeration systems. Proc. Inst. Mech. Eng. Part D J. Automob. Eng. 231(3), 291–309 (2016)CrossRefGoogle Scholar
  19. 19.
    Huang, Y., Khajepour, A., Bagheri, F., et al.: A comparative study of the energy-saving controllers for automotive air-conditioning/refrigeration systems. J. Dyn. Syst. Meas. Control 139(1), 014504 (2017)CrossRefGoogle Scholar
  20. 20.
    Huang, Y., Khajepour, A., Ding, H., Bagheri, F., Bahrami, M.: An energy-saving set-point optimizer with a sliding mode controller for automotive air-conditioning/refrigeration systems. Appl Energy 188, 576–585 (2017)CrossRefGoogle Scholar
  21. 21.
    Liu, J., Peng, H.: Modeling and control of a power-split hybrid vehicle. IEEE Trans. Control Syst. Technol. 16(6), 1242–1251 (2008)CrossRefGoogle Scholar

Copyright information

© Society of Automotive Engineers of China (SAE-China) 2018

Authors and Affiliations

  • Yanjun Huang
    • 1
  • Milad Khazeraee
    • 1
  • Hong Wang
    • 1
  • Soheil M. Fard
    • 1
  • Tianjun Zhu
    • 2
  • Amir Khajepour
    • 1
  1. 1.Department of Mechanical and Mechatronics EngineeringUniversity of WaterlooWaterlooCanada
  2. 2.Department of Electronic Information and Electrical EngineeringZhaoqing UniversityZhaoqingChina

Personalised recommendations