An electrical transformer is a crucial device that transfers electrical energy from one circuit to another through electromagnetic induction, making it an essential component of electrical power systems.
The accelerator demonstrates to end-users the leverage of the IBM solution to generate and deploy the analytic models, including machine learning, without coding requirements. Furthermore, it can integrate with the IBM Maximo Monitor to display the analytic results over the dashboard. To leverage the data sets collected from distribution electrical transformers, we can provide a set of examples that convert this data into business or operation values.
API Documentation is hosted on Developer Hub
Here we include five examples:
-
CO2 Equaivent Emissions - The energy loss to CO2 equaivent estimation provides an approach for taking energy loss calculations and the location of the electrical transformer expressed as a zip code and using that information and CO2 energy conversion data from the EPA to generate an estimate of CO2 emissions. By calculating CO2 emissions, power companies can leverage this information in various ways to create business value and ensure their operations are sustainable and aligned with market demands and regulatory requirements.
-
Rule-based Haromic Anomaly Detection - Voltage Total Harmonic Distortion (THD) in the context of transformers refers to the distortion of the voltage waveform due to multiple harmonic frequencies in addition to the fundamental frequency. The rule-based anomaly detection using the sensors's maximum and minimum threshold set points is a simple and effective method for monitoring sensor. When the observed values exceed these thresholds, it generates a tag (or indicator) for that timestamp. Additional rules are applied to count the indicators in a specific amount of time for different metrics or KPIs to trigger an alert or anomaly. Those additional rules are used to avoid overflooding the anomalies in the system.
-
Energy Loss - Energy loss estimation calculates the amount of energy lost in a system due to various factors such as inefficiencies due to the outdated transformer, too high or too low of capacities, and cooling oil overheating. Energy loss estimation can help identify areas of the system where energy loss is particularly high, allowing for targeted improvements to be made to reduce these losses. By minimizing energy loss in the system, it is possible to increase its efficiency, reduce costs, and minimize environmental impact.
-
Health Score - The health score is a quantitative measure of the electrical transformer's health condition, often ranging from 0 to 100 or some other scale. By continuously monitoring and analyzing the health score, it is possible to identify trends and patterns in the machine's behavior and make adjustments to maximize its operational efficiency and lifespan.
-
Anomaly Detection - Anomaly detection is an essential aspect of transformer monitoring. It enables identifying abnormal behavior that may indicate a potential failure or fault in the transformer. Early detection of such anomalies can facilitate preventative maintenance, reducing the risk of costly downtime or damage to the transformer.
Navigate to the onboarding folder to discover all the resources for utilizing our electrical transformer accelerator. We have provided a detailed outline of the required steps to streamline and automate the entire procedure.
Completing the walkthrough should take around 15 minutes, considering the time needed for training, deployment, and inference on the dataset spanning nine months during the demonstration.
See CONTRIBUTING for more information.
This collection of AI cookbooks is licensed under Apache license. See the LICENSE file.
Please create git issue.