High penetration of renewable energy may cause considerable system frequency deviations. Electric vehicles (EVs) offer alternative potentials for frequency regulation with rapid responding speed. However, for many EVs under centralized control, existing modeling methods may achieve accurate control results at the cost of high computational complexity, a high real-time communication requirement due to many individual control signals, and possible disturbances to charging preferences. This paper focuses on solving these crucial issues. At the system level, we use the state-space method to further simplify the model for an EV aggregator (EVA) with limited data measurements from EVs. During frequency regulation, the reduced EVA model predicts the EVA's regulation capacity and generates a simplified global control signal for the probabilistic control of individual EVs. In cooperation with the ramp rate of conventional generation, the EVA implements a progressive state recovery strategy to reduce the disturbance of regulation service to EVs' charging preferences. The EVA decentralizes partial control authorities to individual EVs by applying response functions at the user side. Based on operating states and laxities, an individual EV adjusts the response process to ensure its charging preference under probabilistic control. Comparative simulations validate the effectiveness of this EVA modeling and control method.

Existing models featuring numerous electric vehicles (EVs) for centralized frequency regulation achieved high accuracy at the expense of heavy computational workloads and a high real-time communication requirement. This paper develops a state space model (SSM) that provides a probability to realize the real-time power control of aggregated EVs with high accuracy and computational efficiency but a low real-time communication requirement. The SSM, a reduced model based on the state space method, accurately describes aggregated EVs with different connecting states and various state-of-charge (SOC) states. Considering heterogeneous charging characteristics and random traveling behaviors of EVs, the SSM realizes the state transition prediction and the regulation capacity estimation with the Markov state transition method, which has a much higher computational efficiency than the existing models. The SSM is used for the frequency regulation, and the SOC adaptive coefficient is implemented to derive the identical control signal and improve the prediction accuracy. The SSM lowers the real-time communication requirement by replacing some real-time processes with offline processes. Meanwhile, the identical control signal is more suitable for real-time dispatching because it broadcasts the control signal to individual EVs globally. Simulation results indicate that the SSM achieves the high prediction accuracy with much higher computational efficiency. Comparison results are conducted to validate the effectiveness of SSM for real-time frequency regulation.

Aggregated electric vehicles (EVs) are emerging as promising resources for the frequency regulation of the electrical grid, but the aggregation dynamics caused by various aggregation delays make unknown threats against the frequency stability. To address this stability uncertainty, a general aggregation-delay model is established for the electrical grid-electric vehicle system, which integrates heterogeneous-delay effect and reveals the sequential impact process of aggregation delays during the frequency regulation. To precisely visualize the aggregation delay influence, the maximum aggregation delay interval is discretized by the Chebyshev nodes, which formulates a delay distorted matrix reducing the infinite operator dimension. With this matrix, the damping ratio and the sensitivity indexes are then extracted constructing a complete workflow to characterize the asymptotic stability for EV aggregation delays. By the proposed approach, the distorted stability space and uniform target eigenvalue are accurately revealed in a typical electrical grid-electric vehicle system. It is also demonstrated that the aggregation delays infect the frequency stability through three unstable modes in a low-frequency oscillation form.

The existing bidding models for retailers managing electric vehicles (EVs) in distribution-level day-ahead (DA) electricity markets have not fully addressed EVs’ temporal distribution, charging and discharging management, or bidding curves. To address these challenges, this paper proposes a DA optimal bidding model for a retailer with flexible participation of EVs. First, the investigating period for DA optimal bidding is modeled to cover the connecting periods of all EVs that may potentially impact the DA power demand prediction. Then, incentive mechanisms for charging and discharging are proposed to enable retailers to manage EVs under uncertain market prices and EV parameters while also considering the battery degradation cost and the preferences of EV users based on a social survey. The optimal charging model based on incentive mechanisms for charging-discharging helps EVs minimize their energy purchase costs under different price scenarios. Based on the minimum EV energy purchase cost, the optimal bidding model of a retailer aims to achieve the maximum bidding profit considering the conditional value at risk (CVaR), with the step bidding curves and the temporal differences for different time intervals considered. Simulations validate the proposed model under flexible incentive mechanism for charging and incentive mechanism for discharging participation of EVs.

With the increasing integration of wind farms and electric vehicles (EVs) in power systems, voltage stability is becoming more and more serious. Based on vehicle-to-grid (V2G), an efficient power plant model of EVs (E-EPP) was developed to estimate EV charging load with available corresponding response capacity under different charging strategies. A preventive control strategy based on E-EPP was proposed to maintain the static voltage stability margin (VSM) of power system above a predefined security level. Two control modes were used including the disconnection of EV charging load (‘V1G’ mode) and the discharge of stored battery energy back to power grid (‘V2G’ mode). A modified IEEE 14-bus system with high penetration of wind power and EVs was used to verify the effectiveness of preventive control strategy. Simulation results showed that the proposed strategy can not only improve the static voltage stability of power system with considerable wind generation, but also guarantee the travelling comfort for EV owners.

Nowadays, the increasing penetration of renewable energies often leads to system frequency variation due to the intermittent outputs, which limits the utilization of renewable electricity and brings higher requirements to frequency regulation. Aggregated electric vehicles (EVs) are suggested to be used as good regulation resources because of the vehicle-to-grid (V2G) capability and quick response characteristic. However, the participation of EVs inevitably leads to cyber delays due to their underlying communication infrastructure and scheduling, which may cause instability. To guarantee the system stability, this paper investigates the stability of a grid frequency regulation system with electric vehicle (EV) aggregators embedded by cyber delays in the cyber system. A comprehensive stability assessment method is proposed in this paper based on the discrete spectrum iteration. A stability assessment method flow is designed with only five stages: 1) system initialization and modeling; 2) spectrum transformation; 3) Chebyshev discretization; 4) Krylov subspace projection dimension reduction; 5) Newton correction. Comprehensive case studies are performed on the cyber regulation systems with single, double, and three EV aggregators to extract the stability region and to validate the proposed approach. It is revealed that the controller gain, the ratio between the participation ratios of each aggregator, and that between EV and grid are important factors for determining the stability of the regulation system at a given controller gain. Besides, the criterion of the stability region changing with the three factors is proposed. It is found that the mass utilization of electric vehicles in the frequency control disrupts system stability due to the aggregation delay and thus the stability region can be divided into several intervals and presents periodicity. It is expected that the proposed criteria can help to guide the determination of delay requirements for EV aggregators participating in frequency regulation service.