A hierarchical energy management strategy for hybrid energy storage via vehicle-to-cloud connectivity
Introduction
Nowadays, a low-carbon, green and sustainable development becomes a key to the energy transition in the world [1]. Global warming as an international issue requires a decrease in fuel consumption and greenhouse gas emission in all types of transportation [2]. To reduce the environmental impact of transportation, advanced algorithms relevant to power and propulsion systems, optimization, efficiency improvement and the use of energy storages play an important role [3]. Energy storage devices (ESDs) used in transportation include batteries [4], ultracapacitors (UCs) [5], flywheels [6], and fuel cells [7]. Hybrid energy storage system (HESS), namely the combination of different ESDs, is able to provide complementary characteristics, such as high power density, high energy density, long cycle life, high operation efficiency, and desired system performance [8]. Due to the complexity of HESS, the system performance highly depends on the energy management strategies (EMSs) [9].
The EMS can be classified into two types: the predictive EMS and the non-predictive EMS [10]. For the predictive EMS, it can also be categorized as “full knowledge”, “partial knowledge” and “zero knowledge”. Most of the studies focus on the first two categories, and the predictive information includes terrain preview, speed limitation, traffic congestion level, road surface condition, future weather, traffic light state, traffic light time, and front vehicle data (e.g., speed and acceleration) [11]. Optimization-based approaches are typically used in predictive EMSs, such as dynamic programming (DP), equivalent consumption minimization strategy, and model predictive control (MPC) [12]. In the HESS applications with battery and UC, DP-based and MPC-based approaches are mostly explored. DP is a global optimization algorithm that can address system nonlinearities and constraints, and therefore providing global optimal solutions for all initial states [13]. MPC is an optimal control strategy and usually includes the following steps [14]: (1) solve the optimization problem within the predictive horizon; (2) only implement the first optimal control law in the optimal control sequence; (3) update the initial states and predictive information, and then perform the optimization again (step 1). In [15], a Pareto-front analysis of the objective function is used for the MPC to reduce fuel consumption and battery aging. The impact of battery state of charge estimation on an MPC-based EMS for a hybrid electric vehicle is investigated in [16]. A Pontryagin’s Minimum Principle based MPC is developed as the EMS for a plug-in hybrid electric bus to reduce the total cost and improve computational efficiency [17]. Note that dynamic programming could also be the optimization solver for MPC, but due to its high computational cost, few existing studies exploit DP for MPC in the real-time applications. The real-time feasibility of DP and MPC are always concerns, especially for DP. For the non-predictive EMS, the rule-based approach, the filter-based approach, and the fuzzy-logic-based approach have been widely explored [10]. Among them, the rule-based approach has been reported as one of the most effective approaches for real-time applications. The “rule” is typically designed using global optimal solutions achieved offline. However, if the driving pattern differs from the designed one, then the performance of the rule-based approach cannot be guaranteed.
Along with the rapid development of machine learning, cloud computing and intelligent transportation system, the full predictive knowledge and advanced optimization-based approaches are becoming achievable [18]. The information from connected and automated vehicles, and the connectivity of vehicle-to-vehicle (V2V), vehicle-to-cloud (V2C) and vehicle-to-infrastructure (V2I) can be used to estimate the future traffic information [11]. A cloud-based battery management algorithm is discussed in [19]. Based on cyber-physical system approaches, vehicle state estimation [20], powertrain hybrid state observation [21], design optimization [22], and energy management [23] have been explored recently. The V2C-based algorithms have been explored, such as the cloud-based resource allocation [24]. The V2C connectivity and cloud computing offer a great opportunity for computationally expensive but effective algorithms, such as dynamic programming, to be implemented in real-time.
This paper focuses on the energy management strategy for electric buses with HESS. To the best of the authors’ knowledge, there are few literatures reported on the optimal EMS for electric buses with battery/UC HESS considering the V2C connectivity. In this paper, a novel hierarchical optimal EMS is proposed to take advantage of V2C connectivity and cloud computing, as shown in Fig. 1. At the cloud level, the traffic information is collected from the infrastructures and other vehicles. With that information, the future speed profile is predicted and converted as a power demand information for EMS. Global optimization is performed in the cloud platform, where the optimal solution can be obtained as a reference for the vehicle-level control. The uncertainties, such as predictive errors, disturbances, and network latency, are addressed in the vehicle-level control.
In this paper, a global optimization approach, i.e., dynamic programming, is used at the cloud level to deal with a highly nonlinear cost function and provide an optimal solution as the reference for the vehicle level control. The highly nonlinear cost function includes HESS electric cost and battery-cycle-life-degradation cost. Due to uncertainties in the driving cycles and potential network latency, this optimal solution might not be a true optimal solution. In order to deal with this issue, an MPC-based approach is developed at the vehicle level. The real-time capability of the proposed MPC is validated in a hardware-in-the-loop experiment. Furthermore, a rule-based EMS is used as a baseline strategy to evaluate the effectiveness of the proposed EMS. The China bus driving cycle (CBDC) and six other real bus cycles in China are used to validate the robustness of the proposed method. The main contributions of the paper are summarized: (1) a novel energy management is developed for electric buses via V2C connectivity; (2) multi-level cost functions are formulated to optimize the electric cost and the battery capacity degradation cost.
The paper is organized as follows. In Section 2, the system modeling and HESS sizing are presented. The optimization formulation and EMS development are presented in Section 3. In Section 4, the comparison study is performed to evaluate the effectiveness of the proposed method. Section 5 concludes the paper.
Section snippets
System modeling and configuration
In this section, the battery/UC HESS dynamic model, the battery degradation model and the electric bus model are presented. The CBDC is used to design the HESS configuration, namely HESS sizing, and the vehicle model converts the speed profile to a power demand profile. As a global optimization approach, dynamic programming is used to evaluate all the candidates of HESS configuration and provide an optimal size of HESS. The cost function takes both electric cost and battery degradation cost
Hierarchical energy management strategy
Along with the development of cloud computing and 5G communication, advanced methods, especially the ones requiring extensive computation and memory, become feasible for real-time applications. In this section, a novel hierarchical energy management strategy is developed based on V2C connectivity and cloud computing. The global optimization is implemented in the cloud platform, and the MPC with a short predictive horizon is implemented in the vehicle control unit. The traffic and trip
Comparison study and performance evaluation
In order to evaluate the performance of the proposed method, a comparison study is performed in this section. In the beginning, a rule-based energy management strategy is developed based on the optimal solution from dynamic programming under the benchmark driving cycle, namely the CBDC. Moreover, the rule-based EMS and the proposed cloud-based hierarchical EMS are evaluated in the CBDC with uncertainties. To be more realistic, additional six real-world bus driving cycles in China are used to
Conclusion
This paper takes advantage of the vehicle-to-cloud connectivity and proposes a novel hierarchical optimal energy management strategy for electric buses with battery/ultracapacitor hybrid energy storage system. A global optimization using dynamic programming, which minimizes the battery degradation cost and the HESS electric cost, is developed at the cloud level. At the vehicle level, a model predictive control is developed to address uncertainties and deal with the system constraints. To
References (41)
- et al.
Distributed energy and microgrids (DEM)
Appl Energy
(2018) - et al.
A novel diagnostic methodology for fuel cell stack health: performance, consistency and uniformity
Energy Convers Manage
(2019) - et al.
Novel technologies and strategies for clean transport systems
Appl Energy
(2015) - et al.
A novel method on estimating the degradation and state of charge of lithium-ion batteries used for electrical vehicles
Appl Energy
(2017) - et al.
Adaptive model predictive control with propulsion load estimation and prediction for all-electric ship energy management
Energy
(2018) - et al.
Control development and performance evaluation for battery/flywheel hybrid energy storage solutions to mitigate load fluctuations in all-electric ship propulsion systems
Appl Energy
(2018) - et al.
A reconstructed fuel cell life-prediction model for a fuel cell hybrid city bus
Energy Convers Manage
(2018) - et al.
Energy storage systems for a low carbon future-in need of an integrated approach
Appl Energy
(2015) - et al.
Energy management strategies comparison for electric vehicles with hybrid energy storage system
Appl Energy
(2014) - et al.
A survey on driving prediction techniques for predictive energy management of plug-in hybrid electric vehicles
J Power Sources
(2019)
Energy management strategies of connected HEVs and PHEVs: recent progress and outlook
Prog Energy Combust Sci
Adaptive energy management of a plug-in hybrid electric vehicle based on driving pattern recognition and dynamic programming
Appl Energy
Lexicographic MPC with multiple economic criteria for constrained nonlinear systems
J Franklin Inst
Pontryagin’s minimum principle based model predictive control of energy management for a plug-in hybrid electric bus
Appl Energy
Battery durability and longevity based power management for plug-in hybrid electric vehicle with hybrid energy storage system
Appl Energy
Cloud resource allocation for cloud-based automotive applications
Mechatronics
Multi-objective optimization of a semi-active battery/supercapacitor energy storage system for electric vehicles
Appl Energy
Parameter identification of lithium-ion battery pack for different applications based on Cramer-Rao bound analysis and experimental study
Appl Energy
The influence of driving cycle characteristics on the integrated optimization of hybrid energy storage system for electric city buses
Energy
Remaining useful life prediction for lithium-ion batteries based on a hybrid model combining the long short-term memory and Elman neural networks
J Energy Storage
Cited by (94)
Adaptive power allocation strategy for hybrid energy storage system based on driving pattern recognition
2024, Journal of Energy StorageHybrid battery energy storage for light electric vehicle — From lab to real life operation tests
2024, Journal of Energy StorageData-driven energy-efficient speed planning and adaptive car-following control for commuter plug-in hybrid electric vehicles
2024, Journal of the Franklin Institute