An example method comprises receiving first historical meso-scale numerical weather predictions (NWP) and power flow information for a geographic distribution area, correcting for overfitting of the historical NWP predictions, reducing parameters in the first historical NWP predictions, training first power flow models using the first reduced, corrected historical NWP predictions and the historical power flow information for all or parts of the first geographic distribution area, receiving current NWP predictions for the first geographic distribution area, applying any number of first power flow models to the current NWP predictions to generate any number of power flow predictions, comparing one or more of the any number of power flow predictions to one or more first thresholds to determine significance of reverse power flows, and generating a first report including at least one prediction of the reverse power flow and identifying the first geographic distribution area.

A power grid reactive voltage control model training method. The method comprises: establishing a power grid simulation model; establishing a reactive voltage optimization model, according to a power grid reactive voltage control target; building interactive training environment based on Adversarial Markov Decision Process, in combination with the power grid simulation model and the reactive voltage optimization model; training the power grid reactive voltage control model through a joint adversarial training algorithm; and transferring the trained power grid reactive voltage control model to an online system. The power grid reactive voltage control model trained by using the method according to the present disclosure has transferability as compared with the traditional method, and may be directly used for online power grid reactive voltage control.

A system including a power plant having thermal generating units that operate according to multiple possible operating modes, which are differentiated by a unique operational or maintenance schedule. The system further includes a hardware processor and machine readable storage medium on which is stored instructions that cause the hardware processor to execute a process related to optimizing the operational or maintenance schedule during a selected operating period. The process may include: receiving the selected operating period; selecting competing operating modes for the power plant during the selected operating period according to a selection criteria; simulating the operation of the power plant during the selected operating period for each of the competing operating modes and deriving simulation results therefrom; evaluating each of the simulation results pursuant to a cost function and, based thereupon, designating at least one of the competing operating modes as an optimized operating mode.

A system for integrating electric loads into the utility electric grid in a cost-effective, emissions-minimizing way. The system generates and implements a computer-readable series of instructions for one or more of starting and stopping electrical device charging or operation powering of an electrical device's duty cycle. The system includes a planning server including an optimization engine, one or more processors, and a memory device storing instructions thereon that when executed by the one or more processors, causes one or more of the one or more processors to: receive a charge request for charging an electrical device or operation powering of the electrical device's duty cycle, wherein the charge request includes time charge block sorting factors, generate a charging and powering schedule that includes multiple sets of disjointed time charge blocks to start and stop drawing power from a power grid to charge the electrical device or power the electrical device's duty cycle, calculate greenhouse gas emissions created during the scheduled time blocks, wherein each time charge block represents a block of time, a charge price, and a charge emissions value; and sort the time charge blocks in the charging and powering schedule using sorting factors that include price, emissions, solar availability, and combinations therein. Applied to a large number of devices, this system enables integration of devices like smart hot water heaters and electric vehicles into a utility's electric grid with reduced capacity need, avoiding capital expenses and lower operating expenses, while also minimizing carbon emissions and other air pollution.

The disclosure relates to a two-side stochastic dispatching method for a power grid. By analyzing historical data of wind power, the Gaussian mixture distribution is fitted by software. For certain power system parameters, a two-side chance-constrained stochastic dispatching model is established. The hyperbolic tangent function is used to analyze and approximate cumulative distribution functions of random variables in the reserve demand constraint and the power flow constraint, to convert the two-side chance constraint into a deterministic constraint. The disclosure can have the advantage of using the hyperbolic tangent function to convert the two-side chance constraint containing risk levels and random variables into the solvable deterministic convex constraint, effectively improving the solution efficiency of the model, and providing decision makers with a more reasonable dispatching basis.

Disclosed herein is a method and system for sharing power or energy across various power supply and control modules. More specifically, disclosed herein are systems and methods for distributing energy. As explained herein, the method discloses receiving, at a microgrid, data from a plurality of data sources. The data is then analyzed to forecast power needs associated with the microgrid. Using the data, the microgrid may determine whether and when to share power with the requesting module.

A simulated protected loads panel system for managing energy consumption and obviating the need to install a physical protected loads panel in conjunction with an energy storage system, comprising a controller, in operable communication with electrical current and/or voltage sensors and relays, which is configured to control the amount of and/or distribution of electrical power from a source of electrical power to an electrical load based on user preference, energy storage system charge, and/or available or anticipated power generation and/or usage.

One embodiment of the present invention provides an energy-asset control system for utilizing an energy asset to provide one of more modes of operation services. The system includes an economic optimizer configured to identify at least one mode of operation opportunity based on current and/or future market conditions; a prognostics module configured to perform a prognostic analysis associated with the mode of operation opportunity for the energy asset using an existing model, and determine a confidence level associated with the prognostic analysis; and an operation controller. The economic optimizer is further to configured to, in response to the prognostics module determining the confidence level exceeding a predetermined threshold, determine an expected profit of the mode of operation opportunity based on outcomes of the prognostic analysis; and optimize, over a predetermined time period, a usage of the energy asset based on the expected profit of the mode of operation opportunity.

Disclosed in the present invention is a method for predicting a delay at multiple corners for a digital integrated circuit, which is applicable to the problem of timing signoff at multiple corners. In the aspect of feature engineering, a path delay relationship at adjacent corners is extracted by using a dilated convolutional neural network (Dilated CNN), and learning is performed by using a bi-directional long short-term memory model (Bi-directional Long Short-Term Memory, BLSTM) to obtain topology information of a path. Finally, prediction results of a path delay at a plurality of corners are obtained by using an output of a multi-gate mixture-of-experts network model (Multi-gate Mixture-of-Experts, MMoE). Compared with a conventional machine learning method, the present invention can achieve prediction with higher precision through more effective feature engineering processing in a case of low simulation overheads, and is of great significance for timing signoff at multiple corners of a digital integrated circuit.

A method, apparatus, system and computer program is provided for controlling an electric power system, including implementation of voltage measurement using paired comparison analysis applied to calculating a shift in average usage per customer from one time period to another time period for a given electrical use population where the pairing process is optimized using a novel technique to improve the accuracy of the measurement.