A Two-Stage Multi-Agent EV Charging Coordination Scheme for Maximizing Grid Performance and Customer Satisfaction
Sensors
Abstract and Figures
Advancements in technology and awareness of energy conservation and environmental protection have increased the adoption rate of electric vehicles (EVs). The rapidly increasing adoption of EVs may affect grid operation adversely. However, the increased integration of EVs, if managed appropriately, can positively impact the performance of the electrical network in terms of power losses, voltage deviations and transformer overloads. This paper presents a two-stage multi-agent-based scheme for the coordinated charging scheduling of EVs. The first stage uses particle swarm optimization (PSO) at the distribution network operator (DNO) level to determine the optimal power allocation among the participating EV aggregator agents to minimize power losses and voltage deviations, whereas the second stage at the EV aggregator agents level employs a genetic algorithm (GA) to align the charging activities to achieve customers’ charging satisfaction in terms of minimum charging cost and waiting time. The proposed method is implemented on the IEEE-33 bus network connected with low-voltage nodes. The coordinated charging plan is executed with the time of use (ToU) and real-time pricing (RTP) schemes, considering EVs’ random arrival and departure with two penetration levels. The simulations show promising results in terms of network performance and overall customer charging satisfaction.
![Daily residential and commercial demand profiles [24].](https://www.researchgate.net/publication/369097008/figure/fig5/AS:11431281390469816@1745275561355/Daily-residential-and-commercial-demand-profiles-24_Q320.jpg)
Figures - available from: Sensors
This content is subject to copyright.
... This problem is independent of the time of planning and related works usually addressed the assignment problem together with either day-ahead and real-time planning or both [8]. Some previous studies considered the process after the vehicle-to-charging station assignment, involving the optimizing coordinated charging schemes for different chargers in the station based on objectives such as minimizing charging costs [12], maximizing user charging demand [13] and maximizing grid power stability [14] [15]. Nevertheless, in terms of the power distribution network operation, the literature review focuses on off-line or day-ahead planning and real-time planning. ...
... Off-line or day-ahead planning is typically used for charging events with regular behavior or when there is more available information. For instance, Amin et al. [13] proposed a twostage multi-agent EV charging coordination scheme for maximizing grid performance and customer satisfaction, assuming that the desired state-of-charge (SOC) and departure time were collected to support charging power allocation. However, the designed scenario was suitable for private chargers or chargers with a reservation function aimed at utilizing valley-period electricity prices to reduce charging costs. ...
... According to the process of the MS-CC, Pω,max should follow the constraints from (9) to (11). The decision variable Pω has limits set by (12) and (13) using the rules of QoSW. If the charging process has started, Pω cannot be 0 in order to keep the stability of the information transmission between EVs and chargers. ...
The widespread adoption of electric vehicles (EVs) with fast-charging capability has presented significant challenges to the safe operation of power distribution networks. Issues such as node overvoltage and frequency fluctuations should be addressed to ensure power quality while maximizing charging Quality-of-Service (QoS) for EV users. This challenge is particularly crucial for fast-charging events that have more flexible demands. Charging priority has traditionally been used to determine an EV charging scheme, including waiting time, arrival order and other factors. However, the scheduling potential predicted from the vehicle side has not been fully considered in the priority establishment when determining real-time operation. To address this gap, a bi-level framework is proposed for a coordinated charging scheme at the fast-charging station that factors in demand-based priority. The framework considers available charging power allocation at different priority levels and optimizes the charging scheme under the same priority level based on three real-time QoS objectives. This approach can improve the average charging QoS for EV users compared to traditional algorithms without considering the demand priority. In robustness evaluation, the QoS of scheduled charging events can reach 86.9% (median) of the optimal solution, and the range of fluctuations in different trials is relatively more stable.
... The widespread adoption of Electric Vehicles (EVs) has great potential to significantly reduce the transportation's sector role, especially in urban areas, in contributing to GHGs emissions. Moreover, the use of EVs helps to mitigate reliance on fossil fuels, leading to a cleaner environment [4]. computing technologies, hybrid methodologies have been studied extensively. ...
... Some of the studies use the metaheuristic-based method to deal with the large-scale real-time scheduling problems. In such scenarios, the genetic algorithm (GA) [7][8][9], artificial bee colony (ABC) [10][11][12] and particle swarm optimization (PSO) [13][14][15] are widely used to search the optimal/suboptimal solution. However, in other studies, the reinforcement learning based algorithms are used to get a better scheduling plan for larger or more complex scenarios. ...
This paper addresses the challenge of large-scale electric vehicle (EV) charging scheduling during peak demand periods, such as holidays or rush hours. The growing EV industry has highlighted the shortcomings of current scheduling plans, which struggle to manage surge large-scale charging demands effectively, thus posing challenges to the EV charging management system. Deep reinforcement learning, known for its effectiveness in solving complex decision-making problems, holds promise for addressing this issue. To this end, we formulate the problem as a Markov decision process (MDP). We propose a deep Q-learning (DQN) based algorithm to improve EV charging service quality as well as minimizing average queueing times for EVs and average idling times for charging devices (CDs). In our proposed methodology, we design two types of states to encompass global scheduling information, and two types of rewards to reflect scheduling performance. Based on this designing, we developed three modules: a fine-grained feature extraction module for effectively extracting state features, an improved noise-based exploration module for thorough exploration of the solution space, and a dueling block for enhancing Q value evaluation. To assess the effectiveness of our proposal, we conduct three case studies within a complex urban scenario featuring 34 charging stations and 899 scheduled EVs. The results of these experiments demonstrate the advantages of our proposal, showcasing its superiority in effectively locating superior solutions compared to current methods in the literature, as well as its efficiency in generating feasible charging scheduling plans for large-scale EVs. The code and data are available by accessing the hyperlink: https://github.com/paperscodeyouneed/A-Noisy-Dueling-Architecture-for-Large-Scale-EV-ChargingScheduling/tree/main/EV%20Charging%20Scheduling.
... For example, Frendo, Gaertner, and Stuckenschmidt (2019) applied mixed integer programming to pre-compute the optimal scheduling for all EVs the day ahead and prioritized the charging during the day. Note that the twostage optimization is applied at the aggregator level, which is different from the two-stage optimization for system-level scheduling, where the first stage optimizes the allocation to different aggregators and the second stage optimizes the charging for EVs at each aggregator, as in Amin et al. (2023). ...
... The implementation of Fuzzy C Mean and Fuzzy K Mean techniques under the RTP framework has further contributed to minimizing electricity charging costs and tempering the peak-power demand curve [79]. In [80], the economic benefits of coordinated charging were investigated using both TOU and RTP. The results show that customers are able to achieve minimum charging costs with delayed charging when using the RTP scheme. ...
High penetration of electric vehicles (EVs) has resulted in an increasing need for charging infrastructure and efficient smart charging coordination of EVs. In addition to residential and public charging, destination charging is another important mode of EV charging, and it has the potential to cover residual public charging demand. Understanding of the factors and charging coordination strategies that ensure a sustainable destination charging business is therefore of utmost importance. Literature on the topic of EV smart charging and coordination is growing exponentially, but the terminology used to describe coordination strategies remained ambiguous. Hence, this paper systematically classifies the literature focused on the destination charging. In doing so, various charging tariffs and business models associated with destination charging are reviewed and a comprehensive discussion on their profitability and availability is provided. Recent EV charging coordination strategies are reviewed and categorized to clarify the commonly used terminology. Destination charging coordination strategies are then classified, and real-world charging coordination initiatives are reviewed and summarized. The analysis contained in this paper shows that further research work on charging coordination for destination charging should account for user behavior and the potential obstacles faced by the intended scale of destination charging to determine a suitable coordination strategy. It is also recognized that implicit charging coordination strategies are understudied and should be investigated because they may circumvent problems encountered by real-world EV charging coordination programs.
... For example, Frendo, Gaertner, and Stuckenschmidt (2019) applied mixed integer programming to pre-compute the optimal scheduling for all EVs the day ahead and prioritized the charging during the day. Note that the twostage optimization is applied at the aggregator level, which is different from the two-stage optimization for system-level scheduling, where the first stage optimizes the allocation to different aggregators and the second stage optimizes the charging for EVs at each aggregator, as in Amin et al. (2023). ...
Existing studies have applied a two-stage approach for large-scale smart EV charging to reduce computation time. The first stage approximates the optimal overall load, and the second prioritizes charging. However, validation and analysis are missing to address whether and why the two-stage approach is suitable. Besides, the existing studies lack exploring different methods to prioritize charging. This work summarizes the two-stage approach and identifies potential concerns. Further, this work validates the approach regarding maximizing the PV-EV temporal load matching at a workplace. Simulation results show that the two-stage approach achieves only 1.7% lower performance than the optimal solution if the optimal overall load is unavailable. Otherwise, lower approximation errors lead to higher final scheduling performance. This work also applies an aggregated model to approximate the overall load. The final performance is only 0.7% lower than when the optimal load is available. Furthermore, this work explores different methods to prioritize charging. Simulation results show that the Least Laxity First performs slightly better, while the Earliest Deadline First can significantly reduce the switching frequency (29.6%). Based on those simulation results, this paper concludes that the two-stage approach can be suitable for aiming to match large-scale EV charging with PV generation.
This chapter aims to examine the effect of perceived relative advantage (PRA), perceived compatibility (PC), perceived complexity (PCX) and perceived usefulness (PU) on the Battery Electric Vehicles' (BEV) Direct Current charging adoption (BDCA) among BEV users in Malaysia. The sample of this study is obtained from individual residing in Malaysia with BEV usage experience in the past 6 months. Data is analysed using quantitative method by applying Structural Equation Modelling. Due to the potential growth in BEVs' DC charging in Malaysia, it is important for EV-related businesses and governments to understand the key factors influencing consumers' adoption in BEVs' DC charging. The 4 key variables (PC, PCX, PRA, PU) will provide an insight for business parties involved to identify consumers' adopting behaviour and differentiate themselves from other competitors with necessary innovations
Fog computing is an emerging research domain to provide computational services such as data transmission, application processing and storage mechanism. Fog computing consists of a set of fog server machines used to communicate with the mobile user in the edge network. Fog is introduced in cloud computing to meet data and communication needs for Internet of Things (IoT) devices. However, the vital challenges in this system are job scheduling, which is solved by examining the makespan, minimizing energy depletion and proper resource allocation. In this paper, we introduced a reinforced strategy Dynamic Opposition Learning based Social Spider Optimization (DOLSSO) Algorithm to enhance individual superiority and schedule workflow in Fog computing. The extensive experiments were conducted using the FogSim simulator to generate the dataset and an energy-efficient open-source tool utilized to model and simulate resource management in fog computing. The performance of the formulated model is ratified using two test cases. The proposed algorithm attained the optimized schedule with minimized cost function concerning the CPU processing period and assigned memory. Our simulation outcomes show the efficacy of the introduced technique in handling job scheduling issues, and the results are contrasted with five existing metaheuris-tic techniques. The results show that the proposed method achieves 10%-15% better CPU utilization and 5%-10% less energy consumption than the other techniques.
Car exhaust is one of the most common causes of ozone hole aggravation, electrical vehicles (EVs) represent a promising solution to avoid this problem. Despite the benefits of EVs, their random charging behavior causes some difficulties regarding the electric network performance, such as increased energy losses and voltage deviations. This paper aims to achieve the proper scheduling of the EVs charging process, avoid its negative impacts on the network, and satisfy the EVs users’ requirements. The EVs charging process is formulated as an optimization problem and solved using particle swarm optimization. The optimization problem formulation considers the EV arrival and departure times and the state of charge required by the user. Different strategies such as separated, accumulated, and ranked strategies with continuous or interrupted fixed charging have been applied to solve the uncoordinated EVs charging problem. These strategies are extensively tested on the modified IEEE 31 bus system (499-node network), using the combination of both Open DSS and MATLAB m-files. The simulation results confirm the effectiveness of the proposed accumulated ranked strategy with interrupted fixed charging in improving the overall power system performance. The achieved improvements include minimizing: the peak power consumed, the peak power losses, and the voltage drop. Moreover, the cost of the EVs charging in most of the feeders has been decreased to a satisfying value. A comparison between the proposed strategy and some previously reported strategies has been performed to ensure the technical and economic enhancement of the proposed strategy.
In this paper, the impact of EV uncontrolled charging with four levels of EV penetration in overall 21 real low voltage distribution grids in two seasons are analyzed. The employed real grid data is provided by distribution system operators from three European countries: Austria, Germany and the Netherlands. At least six grids in each country were considered and they are categorised into three types, namely rural grids, suburban grids and urban grids. The EV charging data used in this study is based on real measurements or surveys. The seasonal and the weekday-weekend factors are also considered in the EV charging impact research. Three key congestion indicators, the transformer loading, line loading and node voltage as well as several other evaluation indexes are studied. The results reveal that the majority of the simulated grids had no or minor moments of mild overloading while a few rest grids have critical issues. Among all the grids, suburban grids are most vulnerable to massive EV integration. Out of the evaluated grids, those who are located in Germany have the highest redundancy for high EV penetration accommodation. Overall, the impact of uncontrolled EV charging depends on the combination of EV charging demand as well as the grid inherent features.
Energy storage is essential for balancing the generation and load in power systems. Building a battery energy storage system (BESS) with retired battery packs from electric vehicles (EVs) or plug-in hybrid electric vehicles (PHEVs) is one possible way to subsidize the price of EV/PHEV batteries, and at the same time mitigating forecast error introduced by load and renewable energy sources in power systems. This paper proposes a detailed framework to evaluate end-of-life (EOL) EV/PHEV batteries in BESS application. The framework consists of three parts. A generalized model for battery degradation is first introduced. It is followed by modeling the battery retirement process in its first life. Two vehicle types—EV and PHEV—as well as two retirement modes—nominal and realistic modes—are considered. Finally, the application of the second-life BESS in power systems is modeled in a detailed economic dispatch (ED) problem. This is how second-life BESS’s performance translates into cost savings on power generation. An optimization problem is formulated to maximize total cost savings in power generation over the battery’s second life. This is done by striking a balance between short-term benefit (daily cost savings) and long-term benefit (cost savings through service years). Numerical results validate the effectiveness of the proposed framework/models. They show that battery usage and retirement criterion in its first life directly affect the performance in its second life application. In our case study, EV battery packs possess larger EOL energy capacities and consequently generate more cost savings in the second life. However, the BESS built from retired PHEV batteries has higher cost savings per MWh. It is because, with the proposed degradation model, battery health is better preserved in PHEV applications. Compared to nominal retirement mode, realistic retirement mode results in extra cost savings due to the reduced first-life service years.
In the modern world, the systems getting smarter leads to a rapid increase in the usage of electricity, thereby increasing the load on the grids. The utilities are forced to meet the demand and are under stress during the peak hours due to the shortfall in power generation. The abovesaid deficit signifies the explicit need for a strategy that reduces the peak demand by rescheduling the load pattern, as well as reduces the stress on grids. Demand-side management (DSM) uses several algorithms for proper reallocation of loads, collectively known as demand response (DR). DR strategies effectively culminate in monetary benefits for customers and the utilities using dynamic pricing (DP) and incentive-based procedures. This study attempts to analyze the DP schemes of DR such as time-of-use (TOU) and real-time pricing (RTP) for different load scenarios in a smart grid (SG). Centralized and distributed algorithms are used to analyze the price-based DR problem using RTP. A techno-economic analysis was performed by using particle swarm optimization (PSO) and the strawberry (SBY) optimization algorithms used in handling the DP strategies with 109, 1992, and 7807 controllable industrial, commercial, and residential loads. A better optimization algorithm to go along with the pricing scheme to reduce the peak-to-average ratio (PAR) was identified. The results demonstrate that centralized RTP using the SBY optimization algorithm helped to achieve 14.80%, 21.7%, and 21.84% in cost reduction and outperformed the PSO.
Electric vehicles’ (EVs) technology is currently emerging as an alternative of traditional Internal Combustion Engine (ICE) vehicles. EVs have been treated as an efficient way for decreasing the production of harmful greenhouse gasses and saving the depleting natural oil reserve. The modern power system tends to be more sustainable with the support of electric vehicles (EVs). However, there have been serious concerns about the network’s safe and reliable operation due to the increasing penetration of EVs into the electric grid. Random or uncoordinated charging activities cause performance degradations and overloading of the network asset. This paper proposes an Optimal Charging Starting Time (OCST)-based coordinated charging algorithm for unplanned EVs’ arrival in a low voltage residential distribution network to minimize the network power losses. A time-of-use (ToU) tariff scheme is used to make the charging course more cost effective. The concept of OCST takes the departure time of EVs into account and schedules the overnight charging event in such a way that minimum network losses are obtained, and EV customers take more advantages of cost-effective tariff zones of ToU scheme. An optimal solution is obtained by employing Binary Evolutionary Programming (BEP). The proposed algorithm is tested on IEEE-31 bus distribution system connected to numerous low voltage residential feeders populated with different EVs’ penetration levels. The results obtained from the coordinated EV charging without OCST are compared with those employing the concept of OCST. The results verify that incorporation of OCST can significantly reduce network power losses, improve system voltage profile and can give more benefits to the EV customers by accommodating them into low-tariff zones.
Growing electric vehicle (EV) dissemination will increase charging infrastructure installation at home. Similar daily routines are associated with high peak loads due to simultaneous EV charging. However, predominantly residential power transmission is not designed for such high loads, yielding charging bottlenecks and restricting future charging at home. Addressing such bottleneck situations and including the EV driver perspective, we introduce a power allocation mechanism that considers the total travel time of the upcoming trip, consisting of actual driving time and time required for charging externally (including the detour to public charging facilities). Assuming that travel time generally negatively correlates with EV driver utility, our optimization model maximizes the resulting utility of EV drivers. Avoiding unnecessary external charging stops due to an insufficient state of charge at the time of departure, our approach generates travel time savings that increase overall EV driver utility. We illustrate our approach using exemplary cases.
Electric vehicle (EV) departure time is an important variable in current coordinated charging studies. The predicted value of EV departure time is more reliable than the user‐set departure time. In this study, a long short‐term memory prediction model is used to accurately predict the departure time, and a two‐layer optimal charging strategy is proposed. Total user satisfaction is set as the objective function, with the constraint of a safe distribution network operation. The proposed strategy is tested using a set of real EV travel data in an old residential area, and its performance is comprehensively compared with two alternative charging strategies, namely, uncontrolled charging and two‐layer optimal charging with a set departure time. The proposed strategy outperforms the rival strategies by improving total user satisfaction, while ensuring safe distribution network operation in the old residential areas. In this paper, we built a long short‐term memory prediction model for user departure time and proposed the index of total user satisfaction. In order to achieve higher total user satisfaction, the total electricity distribution and charging schedule for connected electric vehicles at each time point are optimized successively in the two‐layer optimal charging strategy.
In this paper, a coordination method of multiple electric vehicle (EV) aggregators has been devised to flatten the system load profile. The proposed scheme tends to reduce the peak demand by discharging EVs and fills the valley gap through EV charging in the off-peak period. Upper level fair proportional power distribution to the EV aggregators is exercised by the system operator which provides coordination among the aggregators based on their aggregated energy demand or capacity. The lower level min max objective function is implemented at each aggregator to distribute power to the EVs. Each aggregator ensures that the EV customers’ driving requirements are not relinquished in spite of their employment to support the grid. The scheme has been tested on IEEE 13-node distribution system and an actual distribution system situated in Seoul, Republic of Korea whilst utilizing actual EV mobility data. The results show that the system load profile is smoothed by the coordination of aggregators under peak shaving and valley filling goals. Also, the EVs are fully charged before departure while maintaining a minimum energy for emergency travel.
