Search

With the rise in carbon emissions, it has become fashionable to construct a new power system using new energy as the primary raw material for power generation. This, however, might result in a significant rise in the number of system states and frequent changes in the system state. These issues raise the bar for the effectiveness of system dependability evaluation. Traditional evaluation procedures necessitate computing Optimal Power Flow (OPF) for many systems states one by one, which is unquestionably time-consuming. For that purpose, this research proposes a reliability evaluation approach based on two-dimensional convolutional networks. Branch circuit fault messages are analyzed with generator fault messages. Multiple channel state data from power system node admittance matrices, power generation, and energy demand are fed into a two-dimensional convolutional neural network. This is used to determine the regression connection between load curtailments and system status data and to forecast load curtailments. Finally, it is applied to the RTS-79 system, and experimental results demonstrate that the technique has certain advantages in terms of computing speed and accuracy.

Under the global zero-carbon campaign (United Nations Climate Change, 2021), more stochastic renewable generations are being integrated into the system. This trend significantly expands the state space and increases the computational burden for the model-driven reliability assessment. Data-driven approaches are developed to improve efficiency based on artificial intelligence. However, the requirement of large-scale samples limits it in applications. To address that, this paper adopts a self-supervised stage to avoid the high cost of labeling, while ensuring the efficiency and accuracy of reliability assessment. The training process of this method is split into two stages. In the first stage, feature reconstruction and unsupervised learning are used to provide the initial network parameters. Thereafter, the second learning stage can be trained in a task-agnostic way with fewer labels. The results of case study demonstrate the effectiveness of the proposed approach.

This paper presents a new efficient reliability assessment method for multiple energy systems (MES) based on shadow price (SP). The coupling and interaction of multiple energy flows are quantitatively described by the energy hub (EB) model. The shadow price theory can accelerate the calculations of optimal load shedding in the reliability assessment for MES. Large numbers of system states are generated by the fluctuations of renewable energy generation. These numerous system states can be evaluated by the shadow price theory instead of cumbersome optimization computations. The case study shows that the proposed SP approach can greatly improve the efficiency of MES.

With the continuous expansion of renewable energy in the power grid, a massive number of operating states have been added to the power system. The heavy computational task brings new challenges to the efficiency of reliability assessment. In this regard, a stacked-denoising-auto-encoders-based (SDAE) model is proposed to replace the time-consuming optimal power flow (OPF) algorithm in the reliability assessment of power system operation. The system state and the optimal load curtailment are used as training samples, and the corruption process of SDAE is improved to make it more suitable for power systems. Moreover, unsupervised pretraining and supervised fine-tuning are used to get the optimal coding parameters which can establish the nonlinear mapping relationship between the system states and the minimum load curtailment. Case studies are performed on the RTS-79 system considering renewable energy. Results indicate that the proposed method can greatly improve the efficiency of reliability assessment for power system operation.

The optimal load shedding calculations of an enormous number of contingency states are the most challenging impediments for power system reliability assessment, especially in a large-scale system with multiple types of loads. To deal with that, this paper proposes a fast method that accelerates the calculation of optimal load shedding based on the shadow price. This method calculates the optimal load shedding of systems by shadow-price-based linear functions, rather than the time-consuming OPF algorithms. The hundred percentage criterion is used to look for the matching shadow price of states. Case studies are performed on the RTS-79 system. The results show that the proposed methodology can significantly simplify the solving process and reduce the calculation time.

Mobile energy storage (MES) is a typical flexible resource, which can be used to provide an emergency power supply for the distribution system. However, it is inevitable to consider the complicated coupling relations of mobile energy storage, transportation network, and power grid, which can cause issues of complex modeling and low efficiency. To address that, this paper proposes a mobile energy storage dispatch model to minimize the load curtailment. The framework of rolling optimization is established to update the optimal strategy with the real-time factors, including grid, transportation, and MES. To increase solving efficiency, the model is extended to a mixed integer second-order cone programming model. Finally, the simulation results demonstrate the effectiveness of the proposed method in various scenarios.

The increasing penetration of renewable energy resources in power systems inadvertently leads to a surge in the number of random states. The calculation of optimal load shedding for all the proliferated states is rather overwhelming. It has been considered a bottleneck in the reliability evaluation process. To address this issue, a deep-learning-based approach is proposed to evaluate system reliability more efficiently considering the fluctuation of generation and loads. In the proposed power flow model, a stacked denoising auto-encoder (SDAE) serves as a multilayer neural network for the deep-learning process. Its stacked structure and encoding-decoding recursion enable it to extract high-order features even from non-linear equations. The features of contingency states are crucial to directly acquire the minimum load curtailment without the time-consuming optimal power flow (OPF). The after-trained model is applied to states simulated with the Monte-Carlo method in the RTS-79 system. The numerical result reveals the great advantage of the SDAE-based method in computation time while ensuring high accuracy. As an alternative way to classic OPF algorithms, it can be integrated with the impact-increment method and other state selection approaches.

Utilizing energy storage systems have been considered as a feasible pathway to achieve carbon neutrality. However, the common battery type for energy storage systems is the cheap lithium iron phosphate battery, which has low output efficiency and is almost impossible to charge in cold areas. Lithium titanate battery has high output efficiency and charge efficiency in cold areas. Meanwhile, the price of a lithium titanate battery is three times that of a lithium iron phosphate battery with the same capacity. To achieves the complementary advantages of lithium iron phosphate battery and lithium titanate battery, this paper proposes the dual battery framework of energy storage systems. Based on this, the operation model of energy storage systems is formulated. Besides, five cases are designed to demonstrate the effectiveness of the proposed method and the results show that the dual battery framework has high output efficiency with a cheaper cost than lithium titanate battery.