Fast detection of energy consumption anomalies in buildings
AssigneeInternational Business Machines Corporation
InventorsAmadou Ba, Joern Ploennigs
Life of this patent
Estimating the term…
Abstract
Embodiments for detection of energy consumption anomalies in one or more energy consumption systems in a cloud computing environment by a processor. Energy consumption may be predicted for one or more facilities according to one or more energy consumption measurements, weather data, and one or more characteristics of the one or more facilities, or a combination thereof. An onset of an energy consumption anomaly may be detected according to the prediction.
Claims18
A method for detection of energy consumption in a computing environment by a processor, comprising:
- monitoring, by the processor, data collected by one or more Internet of Things (IoT) sensor devices, wherein the data is representative of a combination of one or more energy consumption measurements, weather data, and one or more characteristics of one or more facilities;
- training, using a machine learning operation by the processor, one or more prediction models according to the monitored data, wherein the one or more prediction models compute, in real-time, residuals for the one or more facilities as a difference between a predicted energy consumption and an actual energy consumption of the one or more facilities;
- predicting, by the processor, energy consumption for the one or more facilities according to an analyzation of the monitored data by the trained one or more prediction models, wherein, when at least some of the residuals begin to deviate from a standardized nominal value, one or more tuning parameters of the one or more prediction models are dynamically changed;
- and
- detecting, by the processor, an onset of an energy consumption anomaly according to the prediction.
A system for detection of energy consumption in a computing environment, comprising:
- one or more computers with executable instructions that when executed cause the system to:
- monitor, by a processor associated with the one or more computers and executing the executable instructions, data collected by one or more Internet of Things (IoT) sensor devices, wherein the data is representative of a combination of one or more energy consumption measurements, weather data, and one or more characteristics of one or more facilities;
- train, using a machine learning operation by the processor, one or more prediction models according to the monitored data, wherein the one or more prediction models compute, in real-time, residuals for the one or more facilities as a difference between a predicted energy consumption and an actual energy consumption of the one or more facilities;
- predict, by the processor, energy consumption for the one or more facilities according to an analyzation of the monitored data by the trained one or more prediction models, wherein, when at least some of the residuals begin to deviate from a standardized nominal value, one or more tuning parameters of the one or more prediction models are dynamically changed;
- and
- detect, by the processor, an onset of an energy consumption anomaly according to the prediction.
A computer program product for detection of energy consumption in a building associated with a computing environment by a processor, the computer program product comprising a non-transitory computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions comprising:
- an executable portion that monitors, by the processor, data collected by one or more Internet of Things (IoT) sensor devices, wherein the data is representative of a combination of one or more energy consumption measurements, weather data, and one or more characteristics of one or more facilities;
- an executable portion that trains, using a machine learning operation by the processor, one or more prediction models according to the monitored data, wherein the one or more prediction models compute, in real-time, residuals for the one or more facilities as a difference between a predicted energy consumption and an actual energy consumption of the one or more facilities;
- an executable portion that predicts, by the processor, energy consumption for one or more facilities according to an analyzation of the monitored data by the trained one or more prediction models, wherein, when at least some of the residuals begin to deviate from a standardized nominal value, one or more tuning parameters of the one or more prediction models are dynamically changed;
- and
- an executable portion that detects, by the processor, an onset of an energy consumption anomaly according to the prediction.
Dependencies are inferred from the claim text.