A method for controlling a wind turbine includes detecting a plurality of analytic outputs relating to power performance of the wind turbine from a plurality of different analytics. The method also includes analyzing the plurality of analytic outputs relating to power performance of the wind turbine. Further, the method includes generating at least one computer-based model of the power performance of the wind turbine using at least a portion of the analyzed plurality of analytic outputs. Moreover, the method includes training the computer-based model(s) of the power performance of the wind turbine using annotated analytic outputs relating to the power performance of the wind turbine. In addition, the method includes estimating a power magnitude of the wind turbine using the machine-learned computer-based model(s). As such, the method includes implementing a control action when the power magnitude of the wind turbine is outside of a selected range.

A method may include obtaining, at a server or analysis device, sensor data comprising at least one of vibration data and impulse data from one or more sensor devices coupled to a first utility infrastructure; obtaining training sensor data associated with at least one of the first utility infrastructure from a previous time period and one or more second utility infrastructures; comparing the sensor data with the training sensor data associated with the at least one of the first utility infrastructure from the previous time period and the one or more second utility infrastructures; and identifying or predicting a fault occurrence associated with the first utility infrastructure based on the comparing the sensor data associated with the first utility infrastructure to the training sensor data associated with the at least one of the first utility infrastructure from the previous time period and the one or more second utility infrastructures.

A diagnostic system is provided and includes a plurality of processors configured to receive data acquired by a sensor. A first processor can execute a safety function that determines a plurality of sensor measurement from the received data, compares selected sensor measurements to predetermined alarm set points, and determines a safety function output for the selected sensor measurements based upon the sensor measurement comparison. A second processor can determine a shadow function output corresponding to the safety function outputs during different respective portions a diagnostic interval. The shadow function output can be configured to replicate the safety function output under conditions where the safety function output is free from error. A third processor can be configured to validate the safety function output by comparing the safety function output to its corresponding shadow function output and outputting a condition based upon the validation comparison.

A method for controlling a wind turbine includes detecting a plurality of analytic outputs relating to power performance of the wind turbine from a plurality of different analytics. The method also includes analyzing the plurality of analytic outputs relating to power performance of the wind turbine. Further, the method includes generating at least one computer-based model of the power performance of the wind turbine using at least a portion of the analyzed plurality of analytic outputs. Moreover, the method includes training the computer-based model(s) of the power performance of the wind turbine using annotated analytic outputs relating to the power performance of the wind turbine. In addition, the method includes estimating a power magnitude of the wind turbine using the machine-learned computer-based model(s). As such, the method includes implementing a control action when the power magnitude of the wind turbine is outside of a selected range.

Systems and methods for automatically generating a set of meta-parameters used to train invariant-based anomaly detectors are provided. Data is transformed into a first set of time series data and a second set of time series data. A fitness threshold search is performed on the first set of time series data to automatically generate a fitness threshold, and a time resolution search is performed on the set of second time series data to automatically generate a time resolution. A set of meta-parameters including the fitness threshold and the time resolution are sent to one or more user devices across a network to govern the training of an invariant-based anomaly detector.

An arc detection and intervention system for a solar energy system. One or more arc detectors are strategically located among strings of solar panels. In conjunction with local management units (LMUs), arcs can be isolated and affected panels disconnected from the solar energy system.

To provide a manufacturing process analysis method for specifying a hindering factor that causes a variation in product performance and for stabilizing product performance. A manufacturing process analysis method comprises: a step for collecting product data indicating the quality of a product and process data indicating manufacturing conditions of a product; a step for standardizing the process data so that the data are converted into an intermediate function; a step for performing principal component analysis on the intermediate function to derive a principal component load amount and a principal component score of the process data; a step for applying cluster analysis to the principal component score to classify manufacturing process lots into a plurality of groups; a step for determining relative merit of each group on the basis of product data soundness corresponding to the principal component score belonging to the group; and a step for specifying a hindering factor that contributes to the relative merit of the group.

An abnormality determination device acquires observation data observed during an operation of an industrial machine, extracts partial time-series data, including a portion representative of a feature of an operating state at a specified timing, from the observation data, calculates a statistical amount from the extracted partial time-series data, and performs processing for machine learning related to determination of operation abnormality of the industrial machine, based on the calculated statistical amount.

A method for the automatic process monitoring and/or process diagnosis of a piece-based process, in particular a production process, in particular an injection-molding process, including the steps: a) performing an automated reference finding in order to obtain reference values (r.sub.1 . . . r.sub.n) from values (x.sub.0 . . . x.sub.j) of at least one process variable; b) performing an anomaly detection on the basis of the reference values (r.sub.1 . . . r.sub.n) found in step (a); c) performing an automated cause analysis and/or an automated fault diagnosis on the basis of a qualitative model of process relationships and/or on the basis of dependencies of various process variables on each other.

Methods and systems for monitoring a brush holder assembly and/or detecting wear of a brush in a brush holder assembly are disclosed. One method includes sending data from a plurality of remote monitoring locations to a central control unit, where the data may be evaluated in order to monitor states of brushes at a plurality of remote electrical facilities. For example, multiple images of a marker tracking longitudinal movement of the brush may be acquired. A comparison of the images, for example, a comparative imaging technique, such as pixel-by-pixel comparison, may then be performed in order to evaluate a condition of the brush, such as the wear rate, wear state, or life expectancy of the brush.