The measurement accuracy of electric energy meters (EEMs) is crucial for respecting the fairness of the electricity market and the justice of electric energy settlement. However, the basic error (BE) of the measurement suffers from external stress, especially under extreme environments. It is challenging to evaluate the degradation trend of EEMs under multi-stress. To address this issue, this paper proposes modeling and evaluating methods for precise degradation analysis of EEMs under multi-stress. First, an optimized k-nearest neighbor (OKNN) is proposed to correct the outliers, where the stress-related weight and weighting factor are designed. Then, a varied-bias wiener process with hierarchical Bayesian (VWHB) model is introduced to evaluate and explore the impact of stress on the BE. The multi-stress, as well as the BE, are fused to parameterize this impact. Integrating the OKNN and VWHB, a degradation analysis framework is further proposed to model the data and conduct the degradation analysis. Extensive experiments on field data collected from the typical operating environment lab demonstrated that the proposed degradation analysis framework reveals superior performance than some state-of-art approaches. The root of the mean of the square of errors and the mean of absolute value of errors of OKNN-VWHB are the lowest with 0.0356 and 0.0281, respectively, for all the EEMs from three different companies.

Air conditioning load (ACL) is an important flexibility resource in smart grids, and its analysis and evaluation have significant implications for demand response, which depends on the nonintrusive extraction of high-frequency ACL profiles. Existing methods based on thermodynamic models require high parameter accuracy, and high-frequency data-driven methods incur high costs for data collection and storage, which limit their widespread application. Considering that smart meters are widely deployed as low-frequency data sources, in this article, we propose a semi-supervised ACL monitoring method based on a small number of high-frequency ACL feature samples in low-frequency data scenarios. First, we introduce a low-frequency ACL state recognition model based on self-supervised contrastive representation learning, which enhances the smart meter data feature unsupervisedly. Then, by merging the identified ACL state with smart meter data, we present an ACL super-resolution generative adversarial network with a specific aggregated adversarial loss, for the super-resolution extraction of ACL curves. Validation on the Dataport dataset shows that the proposed method improves the accuracy of ACL state recognition under low-resolution smart meter data and can accurately extract and reconstruct high-resolution ACL curves.

The increasing integration of renewable energy and electronic power devices into power systems is expanding the data volume and seriously challenging data interaction. The improvement of data reliability contributes to ensuring power supply reliability through the accurate assessment of the power grid's status and the correct response of equipment. To quantify the impact of data reliability on power supply reliability, a data-driven analysis method is proposed. First, based on low-voltage telemetry data, a set of evaluation indicators for data reliability and high-quality power supply reliability is proposed. Then, a reliability scoring method considering combined weight is developed based on the improved technique for order preference by similarity to an ideal solution and grey relational analysis. Additionally, the analysis method based on convolutional neural networks and support vector machines nonlinear fitting is proposed to indicate the impact of data on power supply reliability and find a reasonable range of data reliability indicators. Finally, the effectiveness of the proposed method is verified by the experiments based on actual distribution network data from the eastern province of China.

The high voltage direct current (HVdc) intertie has been applied to provide ancillary-services for ac grids, utilizing the real-time feedback from phasor measurement units (PMUs). However, PMU data communication is vulnerable to false data injection attacks (FDIA) due to protocol defects, thus the HVdc ancillary control and system stability will be threatened. To address this issue, this article proposes a novel HVdc control strategy based on a hybrid data-driven (HDD) methodology. The HDD methodology is first proposed to detect the types and duration time of multiple frequency attacks. Specifically, the Hilbert Huang transform (HHT) is used to decompose the frequency data, using variational mode decomposition instead of the traditional empirical mode decomposition, to extract data features. Second, a multikernel support vector machine is proposed to classify the attacked data based on the designed distinctive features from HHT. Meanwhile, the attacking duration time is decided using an unsupervised technique. Third, an HDD-based HVdc ancillary control strategy is established to eliminate the effect of FDIAs on the HVdc frequency response. Comprehensive experiments of HDD-based HVdc ancillary controls under different FDIAs suggest that the proposed HDD could fast and accurately classify the FDIAs, and the HDD-based HVdc ancillary control strategy could significantly suppress the impact of the FDIAs.

The increasing deployment of phasor measurement units and the advances of their reporting rates are challenging the present data centers in terms of storing and analyzing large-volume data. Under power system disturbance conditions, it is difficult to retain critical information while compressing the synchrophasor data effectively. This article combines the cross entropy and the singular value decomposition, proposing a novel model to compress the synchrophasor data to an extremely small size yet keep superior accuracy. The proposed model is extensively tested and compared with the state-of-the-art algorithms using the simulated and the FNET/GridEye field-collected data. The result indicates that the proposed algorithm has superior performance in compressing the data while retaining critical information under disturbance conditions.

With the development of information technology, most of the medical institutions have established the web medical platform, and it will have a large number of patients' evaluating textual information. These subjective text messages contain emotional information, such as views, opinions and attitudes of patients. It takes a lot of manpower and time to analyze and evaluate positive and negative evaluations by manual methods. Therefore, this paper presents a method of emotion evaluation of medical reviews based on character-level vector convolution neural network. Aiming at the problem of text input noise caused by word segmentation and polysemy, ignoring the structure information of sentence in traditional convolution neural network model, this paper proposes a segmentation pooling convolution neural network model based on character-level vector. Using the improved skip-gram model to train the character vector, and using different convolution kernels of different sizes to extract the sentence features, this proposed model then use the method of segmentation pooling to preserve the maximum eigenvalues of each part ofthe sentence. The experiments show that the accuracy of the proposed model in the emotional analysis of medical texts is about 12% higher than that ofthe traditional convolutional neural network model. In the actual task of emotion analysis of medical texts, the accuracy of the model is as high as 88.2 %.