Abstract: A solar tracking system includes a database to store data associated with a solar field comprising physical location parameters of each of a set of solar panels. The physical location parameters can include a relative row-to-row height and spacing of the solar panels. The system also includes a parameter aggregation tool to aggregate the physical location parameters to generate aggregate time-based output power data associated with the solar field based on geolocation data associated with the solar field. The system also includes a solar tracking tool to generate a solar tracking control scheme for the solar field based on the aggregate time-based output power data, the geolocation data associated with the solar field, and a row-to-row backtracking constraint. The set of solar panels of the solar field can implement solar tracking based on the solar tracking scheme.

Abstract: Systems and methods for monitoring an operational system. A data set with output power values and associated environmental data values for an electrical generation system are accumulated. Statistical relationships are determined for output power values and environmental data values. Outlying data is determined based on the statistical relationships and are removed from the data set to create selected data. A regression model is developed from the selected data to map predicted output power values to values of environmental data. Data with present output power values and present associated environmental data for the electrical generation system are later received. Predicted output power values are predicted by the regression model for the present associated environmental data. An output power discrepancy is identified by comparing the predicted output power to the present output power. A notification of an anomaly is provided based on identification of the output power discrepancy.

Abstract: One example includes a forecast engine that generates forecast data that characterizes predicted operating conditions of an energy storage system for a given time period in the future, wherein the predicted operating conditions are based on a load history for a power consuming premises coupled to the energy storage system and on a value history for power provided to and consumed from a power grid. The load history of the power consuming premises characterizes unmetered power transferred to the power consuming premises, metered powered transferred from the power grid to the power consuming premises and metered powered exchanged from the energy storage system to the power grid. In the example, a schedule manager generates an operation schedule for operating the energy storage system. The operation schedule includes charge and discharge patterns for an energy storage source that are tuned to curtail power costs and/or elevate power revenue value.

Abstract: Systems and methods for monitoring an operational system. A data set with output power values and associated environmental data values for an electrical generation system are accumulated. Statistical relationships are determined for output power values and environmental data values. Outlying data is determined based on the statistical relationships and are removed from the data set to create selected data. A regression model is developed from the selected data to map predicted output power values to values of environmental data. Data with present output power values and present associated environmental data for the electrical generation system are later received. Predicted output power values are predicted by the regression model for the present associated environmental data. An output power discrepancy is identified by comparing the predicted output power to the present output power. A notification of an anomaly is provided based on identification of the output power discrepancy.

Abstract: One example includes a forecast engine that generates forecast data that characterizes predicted operating conditions of an energy storage system for a given time period in the future, wherein the predicted operating conditions are based on a load history for a power consuming premises coupled to the energy storage system and on a value history for power provided to and consumed from a power grid. The load history of the power consuming premises characterizes unmetered power transferred to the power consuming premises, metered powered transferred from the power grid to the power consuming premises and metered powered exchanged from the energy storage system to the power grid. In the example, a schedule manager generates an operation schedule for operating the energy storage system. The operation schedule includes charge and discharge patterns for an energy storage source that are tuned to curtail power costs and/or elevate power revenue value.

Abstract: One example includes a forecast engine that generates forecast data that characterizes predicted operating conditions of an energy storage system for a given time period in the future, wherein the predicted operating conditions are based on a load history for a power consuming premises coupled to the energy storage system and on a value history for power provided to and consumed from a power grid. The load history of the power consuming premises characterizes unmetered power transferred to the power consuming premises, metered powered transferred from the power grid to the power consuming premises and metered powered exchanged from the energy storage system to the power grid. In the example, a schedule manager generates an operation schedule for operating the energy storage system. The operation schedule includes charge and discharge patterns for an energy storage source that are tuned to curtail power costs and/or elevate power revenue value.

Abstract: One example includes a forecast engine that generates forecast data that characterizes predicted operating conditions of an energy storage system for a given time period in the future, wherein the predicted operating conditions are based on a load history for a power consuming premises coupled to the energy storage system and on a value history for power provided to and consumed from a power grid. The load history of the power consuming premises characterizes unmetered power transferred to the power consuming premises, metered powered transferred from the power grid to the power consuming premises and metered powered exchanged from the energy storage system to the power grid. In the example, a schedule manager generates an operation schedule for operating the energy storage system. The operation schedule includes charge and discharge patterns for an energy storage source that are tuned to curtail power costs and/or elevate power revenue value.

Abstract: To facilitate access to public records, the present inventors devised, among other things, an entity resolution system. The exemplary system includes master records database of 300 million entities, which is partitioned into multiple distinct portions. The exemplary system extracts entity information from input public records and constructs one or more blocking queries against specific portions of the master records database to identify one or more sets of candidate records. Feature vectors are defined for the candidate records and machine learning techniques, such as Support Vector Machine, are used to determine which of the candidate records from the master records database match the input public records. Candidate records that match are logically associated with public records, enabling ready access via direct or indirect queries.

Abstract: The present invention provides a method and system for re-ranking search results in a patent retrieval system where the query text is derived in whole or in part from a patent claim, which may be from an existing patent or a prospective claim. The re-ranking is based on several features of the candidate patent, such as the text similarity to the claim, international patent code or other classification or subject matter relatedness or overlap, and internal citation structure of the candidates. One alternative aspect provides a re-ranker that is trained on automatically generated training data, thus obviating the expensive and time-intensive step of expert annotation.

Abstract: A method of using unlabeled data to train a classifier is disclosed. In one embodiment related to record linkage, the method entails retrieving a set of candidate data records from a master database based on a least one update record. Next, a surrogate learning technique is used to identify one of the candidate data records as a match for the one update record. Lastly, the exemplary method links or merges the update record and the identified one of the candidate data records.

Abstract: A method of using unlabeled data to train a classifier is disclosed. In one embodiment related to record linkage, the method entails retrieving a set of candidate data records from a master database based on a least one update record. Next, a surrogate learning technique is used to identify one of the candidate data records as a match for the one update record. Lastly, the exemplary method links or merges the update record and the identified one of the candidate data records.

Abstract: A method of using unlabeled data to train a classifier is disclosed. In one embodiment related to record linkage, the method entails retrieving a set of candidate data records from a master database based on a least one update record. Next, a surrogate learning technique is used to identify one of the candidate data records as a match for the one update record. Lastly, the exemplary method links or merges the update record and the identified one of the candidate data records.

Abstract: The present invention provides a method and system for re-ranking search results in a patent retrieval system where the query text is derived in whole or in part from a patent claim, which may be from an existing patent or a prospective claim. The re-ranking is based on several features of the candidate patent, such as the text similarity to the claim, international patent code or other classification or subject matter relatedness or overlap, and internal citation structure of the candidates. One alternative aspect provides a re-ranker that is trained on automatically generated training data, thus obviating the expensive and time-intensive step of expert annotation.

Abstract: A method of using unlabeled data to train a classifier is disclosed. In one embodiment related to record linkage, the method entails retrieving a set of candidate data records from a master database based on a least one update record. Next, a surrogate learning technique is used to identify one of the candidate data records as a match for the one update record. Lastly, the exemplary method links or merges the update record and the identified one of the candidate data records.

Abstract: For automated text processing, the inventors devised, among other things, an exemplary system that includes an entity tagger, an entity resolver, a text segment classifier, and a relationship extractor. The entity tagger receives an input text segment, and tags named entities with the segment as being a person, company, or place. The entity resolver accesses authority files, and associates the persons and companies named in the text segment with specific entries in the files. The text segment classifier determines whether the text segment includes a relationship event, such as job-change event or merger and acquisition event, and if an event is detected, the relationship extractor determines the event role of entities named in the segment. For example, the extractor determines for a merger and acquisition event, which named company was the acquirer and which was acquired.

Abstract: To facilitate access to public records, the present inventors devised, among other things, an entity resolution system. The exemplary system includes master records database of 300 million entities, which is partitioned into multiple distinct portions. The exemplary system extracts entity information from input public records and constructs one or more blocking queries against specific portions of the master records database to identify one or more sets of candidate records. Feature vectors are defined for the candidate records and machine learning techniques, such as Support Vector Machine, are used to determine which of the candidate records from the master records database match the input public records. Candidate records that match are logically associated with public records, enabling ready access via direct or indirect queries.