Abstract: Machine learning systems and methods for image splicing detection and localization are provided. The system receives an image (e.g., a still digital image, an image frame from a video file, etc.) and divides the image into a set of patches using a patch partitioning algorithm. The system then processes the patches a point set in a high-dimensional feature space, and extracts features from the patches. The system then performs deep learning on the point sets by performing image-level manipulation classification and localization.

Abstract: A method for using adaptive alarm thresholds for rate of change in dissolved gas concentrations for power transformer fault detection may include receiving first dissolved gas data of a power transformer; determining a first rate of change (ROC) of a first gas concentration of the first dissolved gas data; generating, based on the first ROC, a first adaptive alarm threshold for ROC with which to detect a fault in the power transformer; receiving second dissolved gas data of the power transformer; determining a second ROC of a second gas concentration of the second dissolved gas data; comparing the second gas concentration to a static gas concentration threshold; comparing, based on the comparison of the second gas concentration to the static gas concentration threshold, the second ROC to the first adaptive alarm threshold for ROC; detecting the fault based the comparisons; and generating an alert indicative of the fault.

Abstract: A method for using forecasting power transformer faults may include receiving dissolved gas data of power transformers; receiving features of the power transformers; generating, based on the features and the dissolved gas data, clusters of transformers, wherein each respective cluster of the clusters comprises transformers exhibiting similarities to each other; selecting, for each respective cluster and respective gas of the first dissolved gas data, from among machine learning models trained to minimize a difference between a forecast gas concentration and an actual gas concentration, a machine learning model; generating, using a selected machine learning model, a forecasted concentration of the respective gas of the respective cluster; estimating, using forecasted gas concentration, a ROC of the respective gas of the respective cluster; predicting, based on a comparison of the ROC to an ROC alarm threshold, a future fault of a transformer; and generating an alert indicating the transformer maintenance is re

Abstract: The present application provides a method for performing data analytics in hybrid systems. The method may involve: determining a quantum of historical data associated with a machine learning model; determining an order associated with the machine learning model; determining whether a latency associated with the machine learning model is critical; selecting a server from a plurality of servers based at least in part on the quantum of historical data, the order, and the latency; and training the machine learning model using the server. The method may further involve: determining, using the machine learning model, a condition deterioration associated with a power transformer system.

Abstract: Systems and methods for detecting and localizing manipulations of portable document format (PDF) files are provided. The system receives one or more PDF files and processes each PDF file using a plurality of check processes. Each check process includes a defined or learned weight that determines that process's level of influence on any document's overall suspicion score, as well as an opinion score for each PDF file based on unique logic and a plurality of factors related to the particular PDF file. The system can generate an overall suspicion score for the PDF file based on the results of the check processes, which can be calculated by aggregating the weight and an opinion score associated with each check process and file. The system can then determine whether the PDF file is a suspicious file using the overall suspicion score.

Abstract: A power inverter system includes a plurality of power inverters such as solar inverters receiving power from at least one energy source. Each power inverter includes algorithmic maximum peak power tracking (MPPT) software or integrated MPPT firmware to calculate its own potential real power. A controller is in electrical communication with selected power inverters that are operating in curtailed modes of operation and commands each selected power inverter to shift its power production duties and to perform MPPT sweeps. Each selected power inverter then calculates its own potential real power capability in response to corresponding MPPT sweep data.

Abstract: A solar power conversion system is provided. The system includes photovoltaic modules for generating direct current (DC) power, The system also includes power converters for converting the DC power to alternating current (AC) power wherein each of the power converters comprises a local controller and at least some of the local controllers are individually operable as a central controller for providing central control signals to the remaining local controllers and exactly one of the local controllers from the at least some local controllers is operable as the central controller at a given point of time.

Abstract: A power inverter system includes a plurality of power inverters such as solar inverters receiving power from at least one energy source. Each power inverter includes algorithmic maximum peak power tracking (MPPT) software or integrated MPPT firmware to calculate its own potential real power. A controller is in electrical communication with selected power inverters that are operating in curtailed modes of operation and commands each selected power inverter to shift its power production duties and to perform MPPT sweeps. Each selected power inverter then calculates its own potential real power capability in response to corresponding MPPT sweep data.