Embodiments of the present invention provide hydroponics farming apparatus, and systems including the same. The farming apparatus comprises a frame, a plurality of functional systems, a first plurality of sensors configured to monitor conditions associated with farming of the one or more plants, and one or more modular storage cabinets removably attached to the frame. The one or more modular storages include electronics that are pre-assembled and configured to communicate with one or more of the first plurality of sensors and the plurality of functional systems. The electronics includes a main controller configured to collect data from the first plurality of sensors.

Provided is a method for operating an electrical storage station on an electrical supply network. The network has electrical consumers, the storage station, and at least one wind power installation to generate electrical power from wind. The method includes generating electrical power by way of the installation as generated wind power, and feeding a feed-in power into the network. The electrical feed-in power at least results from the generated wind power and a storage power taken up or output by the storage station. The feeding of the feed-in power into the network is controlled depending on a station state of charge and a wind and/or power forecast. Changes in the feed-in power over time are controlled depending on the wind and/or power forecast and a limit gradient is specified to limit the changes in the feed-in power thereto. The limit gradient is specified depending on the wind and/or power forecast.

A computer-implemented system and method for training a machine learning model are disclosed, the method includes: maintaining a data set representing a neural network having a plurality of weights; receiving input data comprising a plurality of time series data sets ending with timestamp t−1; generating, using the neural network and based on the input data, a probabilistic forecast distribution prediction at timestamp t and a selection value associated with the probabilistic forecast distribution prediction at timestamp t; computing a loss function based on the selection value; and updating at least one of the plurality of weights of the neural network based on the loss function.

Disclosed is a method for recognizing distribution network equipment based on Raspberry Pi multi-scale feature fusion. The method includes obtaining an initial sample data set; constructing an object detection network composed of EfficientNe-B0 backbone network, multi-scale feature fusion module and a regression classification prediction head; training the object detection network by taking the initial sample data set as a training sample; finally, detecting inspection pictures by using a the trained object detection network. A light-weight EfficientNet-B0 backbone network feature extraction method obtains more features of objects. Meanwhile, an introduction of multi-scale feature fusion better adapts to small object detection, and a light-weight y_pred regression classification detection head is effectively deployed and realized in Raspberry Pi embedded equipment with tight resources and limited computing power.

Systems, methods and apparatus for electric power grid management and communications are disclosed. At least one active grid element is constructed and configured in network-based communication with a server via at least one coordinator. The at least one active grid element communicates Internet Protocol (IP)-based messages with the server via the at least one coordinator in real time or less than 15 minutes interval. The at least one active grid element participates actively in an electric power grid. The at least one active grid element has an energy consumption pattern or an energy supply pattern. The IP-based messages comprise at least one IP packet including a content, a priority, a security, and a transport route. The content comprises an amount of power available for the electric power grid or an amount of curtailment power available at an attachment point of the at least one grid element.

A transformer health state evaluation method based on a leaky-integrator echo state network includes the following steps: collecting monitoring information in each substation; performing data filtering, data cleaning and data normalization on the collected monitoring information to obtain an input matrix; inputting the input matrix into a leaky-integrator echo state network to generate trainable artificial data, and dividing the artificial data into a training set and a test set in proportion; constructing a deep residual neural network based on a squeeze-and-excitation network, and inputting the training set and the test set for network training; and performing health state evaluation and network weight update based on actual test data. Considering that a deep learning-based neural network needs a large amount of data, the present disclosure uses the leaky-integrator echo state network to generate the artificial training data.

A system incorporating a fluid dispenser for dispensing chemicals includes at least one sensor for sensing fluid flow and a dispenser management unit. The dispenser management unit includes a processor and corresponding storage memory for storing firmware, at least one input for receiving sensor data from the or each sensor, a data transfer device comprising storage memory for storing the received sensor data, and a power source for powering at least the processor. The system further incorporates a portable terminal configured to receive data from the dispenser, and a remote server configured to receive the data from the portable terminal.

In some embodiments, the system is directed to an autonomous inspection system for electrical grid components. In some embodiments, the system collects electrical grid component data using an autonomous drone and then transmits the inspection data to one or more computers. In some embodiments, the system includes artificial intelligence that analysis the data and identifies electrical grid components defects and provides a model highlighting the defects to a user. In some embodiments, the system enables a user to train the artificial intelligence by providing feedback for models where defects or components are not properly identified.

A system for load identification in collaboration with a cloud end includes: a smart Internet of Things electricity meter module, used for matching extracted feature quantity data with a first load feature library of the smart Internet of Things electricity meter module, and determining load feature data corresponding to unmatched feature quantity data in the feature quantity data as target load feature data; a use information front-end/acquisition module, used for calling the target load feature data and transmitting the target load feature data to a main station load identification module; and a main station load identification module, used for receiving the target load feature data, performing data direction processing on the target load feature data, determining an optimal matching strategy matching the feature data to be identified, and identifying an optimal matching solution corresponding to the feature quantity data to be identified in a second load feature library.

A method and apparatus for estimating capacity of a system including an energy generation system, an energy storage system or both. The method and apparatus initially estimate the system capacity based on a facility location and size. The initial estimate may be adjusted through adjustment of at least one parameter. An updated capacity estimate is generated and displayed.