This paper proposes an approach for distribution system load forecasting, which aims to provide highly accurate short-term load forecasting with high resolution utilizing a support vector regression (SVR) based forecaster and a two-step hybrid parameters optimization method. Specifically, because the load profiles in distribution systems contain abrupt deviations, a data normalization is designed as the pretreatment for the collected historical load data. Then an SVR model is trained by the load data to forecast the future load. For better performance of SVR, a two-step hybrid optimization algorithm is proposed to determine the best parameters. In the first step of the hybrid optimization algorithm, a designed grid traverse algorithm (GTA) is used to narrow the parameters searching area from a global to local space. In the second step, based on the result of the GTA, particle swarm optimization is used to determine the best parameters in the local parameter space. After the best parameters are determined, the SVR model is used to forecast the short-term load deviation in the distribution system. The performance of the proposed approach is compared to some classic methods in later sections of this paper.

A fault detection, identification, and location approach is proposed and studied in this paper. This approach is based on matching pursuit decomposition (MPD) using Gaussian atom dictionary, hidden Markov model (HMM) of real-time frequency and voltage variation features, and fault contour maps generated by machine learning algorithms in smart grid (SG) systems. Specifically, the time-frequency features are extracted by MPD from the frequency and voltage signals, which are sampled by the frequency disturbance recorders in SG. A hybrid clustering algorithm is then developed and used to cluster the frequency and voltage signal features into various symbols. Using the symbols, two detection HMMs are trained for fault detection to distinguish between normal and abnormal SG operation conditions. Also, several identification HMMs are trained under different system fault scenarios, and if a fault occurs, the trained identification HMMs are used to identify different fault types. In the meantime, if the fault is detected by the detection HMMs, a fault contour map will be generated using the feature extracted by the MPD from the voltage signals and topology information of SG. The numerical results demonstrate the feasibility, effectiveness, and accuracy of the proposed approach for the diagnosis of various types of faults with different measurement signal-to-noise ratios in SG systems.

An approach of big data characterization for smart grids (SGs) and its applications in fault detection, identification, and causal impact analysis is proposed in this paper, which aims to provide substantial data volume reduction while keeping comprehensive information from synchrophasor measurements in spatial and temporal domains. Especially, based on secondary voltage control (SVC) and local SG observation algorithm, a two-layer dynamic optimal synchrophasor measurement devices selection algorithm (OSMDSA) is proposed to determine SVC zones, their corresponding pilot buses, and the optimal synchrophasor measurement devices. Combining the two-layer dynamic OSMDSA and matching pursuit decomposition, the synchrophasor data is completely characterized in the spatial-temporal domain. To demonstrate the effectiveness of the proposed characterization approach, SG situational awareness is investigated based on hidden Markov model based fault detection and identification using the spatial-temporal characteristics generated from the reduced data. To identify the major impact buses, the weighted Granger causality for SGs is proposed to investigate the causal relationship of buses during system disturbance. The IEEE 39-bus system and IEEE 118-bus system are employed to validate and evaluate the proposed approach.

The ever-increasing penetration of Electric Vehicles (EVs) has emerged as an evident challenge within the Power Distribution System (PDS). Existing Distribution System State Estimation (DSSE) models are constrained by the low-observability and often neglect the spatio-temporal correlations inherent in the PDS. To address this, our study proposes a robust learning architecture named meta physics-informed graph TimesNet that can effectively learn the evolving topology of the PDS, as well as capture the spatio-temporal correlations of the PDS state, even under the condition of limited data availability. The model leverages a combination of a graph convolutional network and TimesNet as the foundational modules for constructing an encoder-decoder framework. Furthermore, to capture the spatio-temporal heterogeneity of the PDS state, our study inserts a topology learning module driven by a meta physics-informed graph bank into the encoder-decoder framework. Subsequently, power flow calculations are performed on the predicted power to reduce error accumulation. The IEEE 8500-node test feeder and UTD19 are utilized as the datasets for the PDS and urban transportation system, respectively. By using 42 electric vehicle charging stations as coupling points, a novel large-scale power-transportation coupled system dataset is constructed through multi-source information fusion. The performance of our proposed model on this dataset has shown to surpass existing DSSE methods in terms of accuracy (MAPE, etc.) and effectiveness, particularly considering the increasing prevalence of EVs. Codes are available at https://github.com/lishijie15/MPGTN-for-DSSE .

Wind energy is highly location-dependent. Many desirable wind resources in North America are located in rural areas without direct access to the transmission grid. By connecting megawatt-scale wind turbines to the distribution system, the cost of building transmission facilities can be avoided and wind power supplied to consumers can be greatly increased; however, integrating megawatt-scale wind turbines on distribution feeders will impact the distribution feeder stability, especially voltage stability. Distributed wind turbine generators (WTGs) have the capability to aid in grid stability if equipped with appropriate controllers, but few investigations are focusing on this. This paper investigates the potential of using real-time measurements from distribution phasor measurement units for a new WTG control algorithm to stabilize the voltage deviation of a distribution feeder. This paper proposes a novel auxiliary coordinated-control approach based on a support vector machine (SVM) predictor and a multiple-input and multiple-output model predictive control on linear time-invariant and linear time-variant systems. The voltage condition of the distribution system is predicted by the SVM classifier using synchrophasor measurement data. The controllers equipped on WTGs are triggered by the prediction results. The IEEE 13-bus distribution system with WTGs is used to validate and evaluate the proposed auxiliary control approach.

Spatio-temporal forecasting plays a crucial role in the dynamic perception of smart cities, such as traffic flow prediction, renewable energy forecasting, and load prediction. Its objective is to understand the patterns of spatio-temporal changes under the interaction of various factors. Accurate spatio-temporal forecasting relies on sufficient high-quality data and powerful models. However, in reality, data is often sparse. In such cases, while adaptive graphs and large language models (LLMs) can maintain performance, they face issues of spatial spurious associations and hallucinations, respectively. These issues hinder the ability of the model to learn and infer cross spatio-temporal and cross-scale features effectively. To address this, we propose a novel model termed spatio-temporal causal intervention large language model (STCInterLLM). This model employs a newly designed causal intervention encoder to update spatial spurious correlations in the spatio-temporal adaptive graph. Subsequently, the novel chain-of-action prompting text is utilized to enforce the decomposition of the prediction process, thereby enhancing the causal representation of features while mitigating hallucinations in LLMs. Finally, a lightweight marker alignment module ensures the consistency between the encoder, prompting text, and LLM, enabling accurate forecasting of distinct scale spatio-temporal evolution patterns. Extensive experiments conducted on power distribution systems integrated with renewable energy sources and transportation systems encompassing diverse types of data, demonstrate that the proposed STCInterLLM consistently achieves state-of-the-art performance across significantly varied scenarios. Codes are available at https://github.com/lishijie15/STCInterLLM.