Aspects of the present invention relate to a method for controlling a renewable power plant connected to a power network to reduce deviation of a measured frequency of the power network from a target frequency. The method comprises determining a forecasted power gradient over a forecast interval defined between a first time point and a second time point, and, at a third time point during the forecast interval, controlling the power plant to output active power according to a minimum active power level if the measured frequency at the third time point is below the target frequency, controlling the power plant to output active power according to a maximum active power level if the measured frequency at the third time point is above the target frequency. The maximum and minimum active power levels are based on the forecasted power gradient.

Implementations of the disclosure describe an Open Automated Demand Response (OADR) endpoint device. A method of the disclosure includes receiving, by a processing device of an endpoint device, a first message comprising a first market energy price, determining whether a first quantity of energy has been consumed within a first time period by at least one energy device associated with the endpoint device, responsive to a determination that the first quantity of energy has been consumed by the at least one energy device within the first time period, determining whether the first market energy price is greater than a price threshold and responsive to a determination that the first market energy price is greater than the price threshold, causing the at least one energy device to stop consuming additional amounts of energy.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a neural network used to select actions performed by a reinforcement learning agent interacting with an environment. In one aspect, a method includes maintaining a replay memory, where the replay memory stores pieces of experience data generated as a result of the reinforcement learning agent interacting with the environment. Each piece of experience data is associated with a respective expected learning progress measure that is a measure of an expected amount of progress made in the training of the neural network if the neural network is trained on the piece of experience data. The method further includes selecting a piece of experience data from the replay memory by prioritizing for selection pieces of experience data having relatively higher expected learning progress measures and training the neural network on the selected piece of experience data.

Techniques for customizing knowledge representation systems including identifying, based on a plurality of concepts in a knowledge representation (KR), a group of one or more concepts relevant to user context information, and providing the identified group of one more concepts to a user. The KR may include a combination of modules. The modules may include a kernel and a customized module customized for the user. The kernel may accessible via a second KR.

A method includes collecting energy resource data for the specific geographic location over a predetermined time period, calculating power curves and matrices for at least two energy technologies based on the collected energy resource data, estimating the potential of generated electric power over time of the at least two energy technologies based on the calculated power curves and matrices, the time period, and the characteristic parameters of each of the at least two energy technologies, simulating different base load and power variations based on the estimation of the potential generated electric power and different distribution of the electric power generation of the at least two energy technologies, identifying an optimal distribution of the at least two energy technologies by analyzing the base load and power variations for each simulation, and choosing the distribution of the electric power generation with the highest base load and lowest power variation.

An operation control system for a gas-steam combined cycle generator unit-includes a real-time operation data acquisition module, to collect the operation parameter data and power generation data of a power plant unit, an operation status evaluation index mining module, to mine and analyze the operation parameter data of a power plant unit to get key parameters, an operation status evaluation index extraction module, to obtain characteristic variables, an operation characteristic parameter prediction module, to predict the said characteristic variables and obtain the predicted value and the corresponding change trend, and an operation intelligent control module, to realize the intelligent control of parameters. The operation control system can guide the optimal operation of power plant units, and improve their operation reliability and economy.

A method includes generating a time-varying charge/discharge control signal for an electrical storage device, wherein generating the time-varying charge/discharge control signal comprises identifying a prioritization order of a stack of simultaneously operating control modes, the stack of simultaneously operating control modes including a staging mode and at least two additional control modes, each control mode of the stack comprising a plurality of control signal candidate values; identifying an intersection of one or more control signal candidate values from the plurality of control signal candidate values of each control mode of the stack according to the prioritization order; and determining, based on the prioritization order, at least one time-varying charge/discharge control signal for the electrical energy storage device from the intersection of control signal candidate values.

A computer-implemented method for providing an estimated utility expenditure to a user may include: obtaining, via one or more processors, historical transactional data of one or more customers other than the user from one or more transactional entities, wherein the historical transactional data includes: at least one address of a given customer of the one or more customers; and a historical utility expenditure associated with the at least one address; generating, via the one or more processors, a heatmap based on the historical transactional data of the one or more customers via one or more algorithms, wherein the heatmap is indicative of at least the estimated utility expenditure associated with the at least one address during a predetermined period; and causing a display of a user device associated with the user to demonstrate the heatmap.

An example system includes an aggregator configured to receive a service collaboration request and iteratively determine, based on minimum and maximum power values for DERs under its management, an optimized operation schedule. The aggregator may also be configured to iteratively determine, based on the optimized operation schedule, an estimated flexibility range for devices under its management and output an indication thereof. The system may also include a power management unit (PMU) configured to iteratively receive the indication and determine, based on a network model that includes the estimated flexibility range, a reconfiguration plan and an overall optimized operation schedule for the network. The PMU may also be configured to iteratively cause reconfiguration of the network based on the plan. The PMU and aggregator may also be configured to iteratively, at a fast timescale, cause energy resources under their management to modify operation based on the overall optimized operation schedule.

A method for controlling an electrical installation at least one of an electrical energy source or an energy sink. The electrical installation is coupled to a power grid. The method includes a period of time having a start time and a duration being specified, an upward flexibility Fo, which includes a forecast maximum feed-in power increase or feed-out power decrease, and a downward flexibility Fu, which includes a forecast maximum feed-out power increase or feed-in power decrease, being set for the period of time, a selling threshold price Pv and a purchasing threshold price Pe being set for the period of time, and an electricity trading transaction being concluded for the period of time. The electricity trading transaction includes a base value, a base quantity, a base price, a date on which the electricity trading transaction is to be carried out.