Systems and methods for detecting changes in energy usage in a building

A system for generating a baseline energy usage model for a building management system comprising:

• a processing circuit configured to receive an initial set of variables for potential use in a baseline energy usage model for predicting energy use in a building,
• wherein the processing circuit is further configured to perform statistical hypothesis testing on each of the initial set of variables to determine which of the initial set of variables are statistically significant to the energy use in the building, the statistical hypothesis testing comprising:
• generating a first set of coefficients for the baseline energy usage model based on the initial set of variables;
• removing one of the variables from the initial set of variables to create a subset of the initial set of variables;
• generating a second set of coefficients for the baseline energy usage model based on the subset of the initial set of variables;
• and
• calculating a test statistic for the removed variable using a difference between the first set of coefficients and the second set of coefficients and comparing the calculated test statistic to a critical value;
• wherein the processing circuit is further configured to automatically select one or more of the initial set of variables for use in the baseline energy usage model based on a result of the statistical hypothesis testing, wherein a variable is selected for use in the baseline energy usage model when the test statistic calculated for the variable exceeds the critical value,
• wherein the processing circuit is further configured to cause the one or more automatically selected variables to be displayed on a user interface device,
• wherein the processing circuit is further configured to receive, from the user interface, a selection of one or more variables which differ from the one or more automatically selected variables, and
• wherein the processing circuit is further configured to use the selection of the one or more variables which differ from the one or more automatically selected variables to generate a new baseline energy usage model.

Non-transitory computer-readable storage media with computer-executable instructions embodied thereon that when executed by a computer system perform a method for use with a building management system in a building, wherein the instructions comprise:

• instructions for receiving an initial set of variables for potential use in a baseline energy usage model for predicting energy use in a building;
• instructions for performing statistical hypothesis testing on each of the initial set of variables to determine which of the initial set of variables are statistically significant to the energy use in the building, the statistical hypothesis testing comprising:
• generating a first set of coefficients for the baseline energy usage model based on the initial set of variables;
• removing one of the variables from the initial set of variables to create a subset of the initial set of variables;
• generating a second set of coefficients for the baseline energy usage model based on the subset of the initial set of variables;
• and
• calculating a test statistic for the removed variable using a difference between the first set of coefficients and the second set of coefficients and comparing the calculated test statistic to a critical value;
• instructions for automatically selecting one or more of the initial set of variables for use in the baseline energy usage model based on a result of the statistical hypothesis testing, wherein a variable is selected for use in the baseline energy usage model when the test statistic calculated for the variable exceeds the critical value;
• instructions for causing the one or more automatically selected variables to be displayed on a user interface device;
• instructions for receiving a selection of one or more variables which differ from the one or more automatically selected variables;
• and
• instructions for using the selection of the one or more variables which differ from the one or more automatically selected variables to generate a new baseline energy usage model.