Configure Energy Anomaly Detection

Overview

In the configuration of Energy Anomaly Detection, you define:

• how sensitive anomaly detection should be

• how often anomalies are checked

• which financial thresholds are used

• which attributes are included in the analysis

The configuration page is divided into two areas:

• Module settings

• Analyzed attributes

Open the configuration

1. In the sidebar, click on Platform core [1].

2. In the top navigation, click on Modules [2]. ▷ The module configuration page opens.

3. In the module list, open Energy Anomaly Detection [3]. ▷ The configuration page opens.

Configure module settings

In Anomaly Module settings, define how anomalies are detected and evaluated.

1. Set the Anomaly sensitivity under [A]. ◈ This value defines the anomaly score threshold at which the system classifies an event as significant and triggers an alert. ◈ Lower sensitivity detects more anomalies, including smaller deviations. ◈ Higher sensitivity requires stronger evidence before an event is classified as an anomaly.

2. Set the Anomaly check interval under [B]. ◈ This value defines how often anomaly detection is executed.

3. Set the Anomaly financial impact alert threshold under [C]. ◈ An alert is only triggered if the estimated financial impact of an anomaly exceeds this value.

4. Set the Anomaly financial impact meter threshold under [D]. ◈ Meters with a possible daily impact below this value are excluded from the anomaly analysis.

5. Enter the Average energy price under [E]. ◈ This value is used to calculate the estimated financial impact of detected anomalies.

Open the filter for analyzed attributes

On the right side in Analyzed attributes, you can filter the asset list before selecting the attributes to analyze.

1. Click on [F]. ▷ The window Filter asset list opens.

Configure the filter

1. In [F1], select one or more Sites.

2. In the Tags area, choose how tags should be applied under [F2]:

   • All the selected required

   • One of the selected

3. In [F3], select one or more tags.

4. In [F4], select the Asset type.

5. In [F5], select the attribute Class.

6. Click on Apply [F6]. ▷ The filtered asset list is shown in Analyzed attributes.

Select analyzed attributes

After applying the filter, define which attributes are included in anomaly detection.

1. If you want to enable all filtered attributes, click on [G]. ▷ All filtered attributes are enabled for anomaly detection.

2. If you want to disable all filtered attributes, click on [H]. ▷ All filtered attributes are disabled for anomaly detection.

3. If required, use [I] to search for a specific asset or attribute.

4. In the asset tree, enable or disable individual attributes under [J]. ◈ This defines on which high-level attributes anomaly detection is performed.

5. Click on Save [K]. ▷ The configuration is saved.

Notes

• Enabled attributes define where Energy Anomaly Detection actively searches for anomalies.

• Attributes that are not enabled can still show anomalies if an anomaly is detected higher in the asset hierarchy and cascades down.

• This means the enable or disable setting controls the active anomaly search, not whether an attribute can ever be affected by an anomaly result.

• All relevant meters are still considered when evaluating anomalies from the cascade, provided they pass the Anomaly financial impact meter threshold.

A practical configuration sequence is:

1. Define the anomaly sensitivity.

2. Define the check interval.

3. Define the financial thresholds.

4. Enter the average energy price.

5. Filter the asset list.

6. Select the attributes for active anomaly detection.

7. Save the configuration.