Elsevier

Energy

Volume 69, 1 May 2014, Pages 319-335
Energy

Adaptive intelligent energy management system of plug-in hybrid electric vehicle

https://doi.org/10.1016/j.energy.2014.03.020Get rights and content

Highlights

  • The AIEMS can learn while it is running and makes proper adjustments during its action.

  • The fuzzy engine controller is made adaptive by using a hybrid ANFIS with genetic algorithm optimization.

  • Models of vehicle engine, air conditioning, powertrain, and hybrid electric drive system are developed.

Abstract

Efficient energy management in hybrid vehicles is the key for reducing fuel consumption and emissions. To capitalize on the benefits of using PHEVs (Plug-in Hybrid Electric Vehicles), an intelligent energy management system is developed and evaluated in this paper. Models of vehicle engine, air conditioning, powertrain, and hybrid electric drive system are first developed. The effect of road parameters such as bend direction and road slope angle as well as environmental factors such as wind (direction and speed) and thermal conditions are also modeled. Due to the nonlinear and complex nature of the interactions between PHEV–Environment–Driver components, a soft computing based intelligent management system is developed using three fuzzy logic controllers. The crucial fuzzy engine controller within the intelligent energy management system is made adaptive by using a hybrid multi-layer adaptive neuro-fuzzy inference system with genetic algorithm optimization. For adaptive learning, a number of datasets were created for different road conditions and a hybrid learning algorithm based on the least squared error estimate using the gradient descent method was proposed. The proposed adaptive intelligent energy management system can learn while it is running and makes proper adjustments during its operation. It is shown that the proposed intelligent energy management system is improving the performance of other existing systems.

Introduction

PHEV (Plug-in Hybrid Electric Vehicle) provides a platform to reduce fuel consumption and energy waste through control strategies. A PHEV is a hybrid vehicle with rechargeable batteries. The system allows the batteries to be fully restored by plugging into an electric power source [1], [2], [3]. PHEVs are mainly classified into series, parallel (the type used in this study), or series–parallel types [4]. Recent works have shown that vehicle electrification can reduce the energy waste and minimize fuel consumption [5], [6], [7], [8]. The operation optimization of PHEV however requires complete system analysis of the PHEV electric and electronic subsystems [9]. Also the above studies showed that a powertrain control strategy, which is an algorithm used for issuing a sequence of instructions from the vehicle controller to operate the powertrain of the vehicle, is needed to optimize the energy consumption. The control strategy should be able to adjust itself to changes in driving conditions (e.g. road slope or wind behavior) and deliver an optimized energy management. Also the studies show that the powertrain control strategy for series power source has less complexity compared to others since the ICE (Internal Combustion Engine) only drives an alternator to generate electricity for electric motor without having mechanical connection with the wheel.

Assuming that the PHEV has a fully charged ESS (Energy Storage System), indicated by the maximum SOC (State-of-Charge), three options can be considered for its energy management strategy [10]. Those options are: (i) AER (All Electric-Range) or AER-focused strategy, (ii) an engine-dominant blended strategy, and (iii) an electric-dominant blended strategy. The AER-focused strategy requires larger and more expensive electric components compared to other options while the control strategy due to be all-electric is relatively straightforward. The engine-dominant and electric-dominant blended strategies allow the usage of smaller and less expensive electric components. For short driving distances, the engine-dominant blended strategy has a significant fuel use penalty due to underutilization of the electrical recharge energy. Hence, the engine-dominant blended strategy is useful for short and large distance driving scenarios that would benefit from the efficiency maximization approach. Since our aim is to maximize the energy efficiency, we focus on optimizing the engine-dominant blended strategy for practical situations.

The engine-dominant blended PHEV has been examined from a number of different control strategy perspectives [10]. One of the most important challenges for the development of an efficient PHEV is the optimum use of available mechanical and electrical energy sources. Strategies for energy management control in a PHEV are classified based on their information and computational requirements [11]. For instance, dynamic programming, which is a numerical optimization method to compute the optimal solution by breaking a complex problem into simpler smaller problems, requires the availability of a high level of information and computational resources. In case of optimizing the driving of a PHEV, it is practically impossible to apply this method since it requires complete driving cycle information (including future data) [12]. Dynamic programming also requires a large amount of memory to perform the optimization [13], [14]. Unlike dynamic programming, other algorithms such as, ECMS (Equivalent Consumption Minimization Strategy) [15], stochastic dynamic programming [14], [16], [17] and convex optimization [18] can be implemented on PHEVs. ECMS is based on the fact that in general, a PHEV energy consumption from its battery is replenished by running the engine or the energy coming from the power grid [19]. Therefore, battery discharging at any time is equivalent to a quantity of fuel consumption in the future. The ECMS searches the best combination between the engine and motor power, which minimizes the effective fuel consumption. Recently, convex optimization powertrain methods [18] have also been used to improve the efficiencies of charge-depleting, charge-sustaining and blended strategies as well. The above algorithms in comparison with dynamic programming have fewer information requirements for optimization or tuning of the parameters.

It is important to note that there is a tradeoff between algorithm performance and information requirements. In real driving scenarios, the vehicle performance is affected by road and environmental conditions and driver behavior. Therefore, the vehicle performance would be improved if the energy management strategy were able to adapt to different road and environmental conditions and driver behaviors. This means that information about the road and weather is very important for optimal performance of the PHEV energy management. Furthermore, intelligent control systems have proven to be effective for controlling highly complex and nonlinear processes that are subjected to frequent disturbances. In particular, fuzzy logic is a popular method for controlling the operation of hybrid electric vehicles [20], [21] and has been shown to require fewer computations, less memory and achieves better performance compared to nonlinear control schemes [22], [23]. One of the key contributions of this work is the development of an adaptive fuzzy controller to achieve near optimal performance for a PHEV. The proposed solution is practical and can be implemented on an actual vehicle as it only relies on information that are readily available, e.g. road geometry (slope, bend), wind (direction, speed) and environmental (ambient temperature, solar radiation and humidity) parameters. The study in Ref. [24] demonstrated that achieving thermal comfort in an energy efficient way is a difficult task, requiring good coordination between vehicle engine and the AC (Air Conditioning) system. They developed a coordinated energy management system to reduce the energy consumption of the vehicle AC system.

In our study, the engine-dominant blended strategy for PHEV is selected for the development of an AIEMS (Adaptive Intelligent Energy Management System). The rationale for the development of the AIEMS is that the operation range of conventional engines should be adjusted based on the loading of the vehicle in a PHEV. This is especially beneficial if the vehicle could make reliable predictions about the Look-Ahead of the road. The AIEMS searches for the best combination between the engine, electric motor power and AC system, which minimizes the effective fuel consumption. This allows the engine of the PHEVs to be utilized in its best operating range. Without an AIEMS, high energy efficiency is unlikely to be reached under normal driving conditions.

Main contributions of this paper are as follows. The optimal operation of vehicle air conditioning system is for the first time studied as part of PHEV power management. Furthermore, a novel adaptive energy management system for combined dynamic loads is proposed. This method considers the operation of engine, electric motor/generator, air conditioning and vehicle speed, simultaneously. The proposed adaptive energy management system is intelligent in the sense that during the vehicle operation, the proposed system is able to learn and make proper adjustments to the way different power sources operate under different road geometry and wind and environment thermal conditions.

This paper is organized as follows. Section 2 describes the proposed energy management modeling of plug-in hybrid electric vehicles. A new energy management system to improve the vehicle efficiency using a backwards-forwards simulation similar to the technique employed by ADVISOR [25], [26] is presented. The backwards method means that “assuming the vehicle met the required trace, how must each component perform” and forward approach include a driver model, which considers the required speed and the present speed to develop appropriate throttle and brake commands. Section 3 presents the intelligent control system for the management of energy consumption. Three fuzzy controllers employed within the intelligent energy management systems are: (i) a fuzzy cruise controller to adapt vehicle cruise speed via prediction of the road ahead using a look-ahead system, (ii) a fuzzy air conditioning controller to produce desirable temperature and air quality inside vehicle cabin room using road information system, and (iii) a fuzzy engine controller to generate the required engine and electric motor/generator torque to move the vehicle smoothly on the road. Section 4 describes the proposed adaptive intelligent system for online control and energy management. The simulation results and associated discussions are presented in Section 5. Concluding remarks are given in Section 6.

Section snippets

Energy management modeling of plug-in hybrid electric vehicles

A thorough energy management approach in PHEVs needs to consider a number of factors including: environmental conditions, driver behavior and vehicle specifications. In order to improve overall efficiency, it is necessary to develop an accurate model of factors affecting vehicle energy consumption. To fulfill this requirement, models for several major influential factors have been constructed: road geometry (slope, bend), wind condition (speed, direction), thermal condition (humidity, solar

Intelligent energy management system for PHEVs

To be able to produce the desired Adaptive Intelligent Energy Management System (AIEMS), we first need to design a basic energy management system. This system requires applying a control strategy for energy management. Due to randomness, and uncertainties involved in the dynamics of energy consumption and production, such a control strategy needs to be robustly effective and in a working state for various situations. Therefore, we have chosen to devise and use fuzzy controllers for

Adaptive intelligent system

The operation of a fuzzy controller is affected by the size of the membership functions of the fuzzy sets, the position of the membership functions, the rule weights and the link values. These values are usually calculated empirically and it is consequently ideal for a narrow range of overall system operating parameters. To allow the controller to operate efficiently under a broader range of these values, the design of the key component of the proposed system, the fuzzy engine controller, was

Simulation results and discussions

This section analyzes the properties of two proposed systems for PHEV: the intelligent energy management system (IEMS) and the adaptive intelligent energy management system (AIEMS). To benchmark the performances of the proposed systems, we compared the fuel consumption of those methods with a popular baseline control policy implemented as part of the ADVISOR [3], [14], [26]. Our implementation is based on PHEV definition provided by the IEEE-USA Energy Policy Committee [44]. Our implementations

Conclusion

A novel intelligent energy management system for plug-in hybrid electric vehicle was proposed and evaluated for several scenarios using real-world parameters related to vehicle specifications, environmental conditions, and driver behavior. Realistic models for vehicle's engine, air conditioning, powertrain, and hybrid electric drive system were developed. We also developed models of road by taking into account bend direction and road slope angle, and environmental conditions by including wind

References (44)

  • H. Khayyam et al.

    Coordinated energy management of vehicle air conditioning system

    Appl Therm Eng

    (2011)
  • R.T. Doucette et al.

    Modeling the prospects of plug-in hybrid electric vehicles to reduce CO2 emissions

    Appl Energy

    (2011)
  • X. Hu et al.

    A comparative study of equivalent circuit models for Li-ion batteries

    J Power Sources

    (2012)
  • H. Khayyam et al.

    Adaptive cruise control look-ahead system for energy management of vehicles

    Expert Syst Appl

    (2012)
  • R.K. Ahluwalia et al.

    Fuel economy of hydrogen fuel cell vehicles

    J Power Sources

    (2004)
  • M. Buragohain et al.

    A novel approach for ANFIS modelling based on full factorial design

    Appl Soft Comput

    (2008)
  • M.A. Shoorehdeli et al.

    Identification using ANFIS with intelligent hybrid stable learning algorithm approaches and stability analysis of training methods

    Appl Soft Comput

    (2009)
  • S.S. Gill et al.

    An adaptive neuro-fuzzy inference system modelling for material removal rate in stationary ultrasonic drilling of sillimanite ceramic

    Expert Syst Appl

    (2010)
  • C. Zhang et al.

    Intrusion detection using hierarchical neural networks

    Pattern Recogn Lett

    (2005)
  • B.S.M.C. Borba et al.

    Plug-in hybrid electric vehicles as a way to maximize the integration of variable renewable energy in power systems: the case of wind generation in northeastern Brazil

    Energy

    (2012)
  • M. Ehsani et al.

    Modern electric, hybrid electric, and fuel cell vehicles: fundamentals, theory, and design

    (2005)
  • D. Madzharov et al.

    Integrating electric vehicles as flexible load in unit commitment modeling

    Energy

    (2014)
  • Cited by (157)

    View all citing articles on Scopus
    View full text