Building energy system with energy data stimulation for pre-training predictive building models

Abstract

A building energy system for a building, the building energy system including one or more memory devices configured to store instructions thereon, that, when executed by one or more processors, cause the one or more processors to receive a building model, the building model defining a physical construction of the building, generate predicted energy usage data of the building based on the building model and historical weather data, train a predictive building model based on the predicted energy usage data, wherein the predictive building model indicates a prediction of an optimal equipment operating setting, generate the optimal equipment operating setting based on the predictive building model, and operate one or more pieces of building equipment based on the optimal equipment setting to control an environmental condition of the building.

What is claimed is: 
     
         1 . A building energy system for a building, the building energy system comprising one or more memory devices configured to store instructions thereon, that, when executed by one or more processors, cause the one or more processors to:
 receive a building model, the building model defining a physical construction of the building;   generate predicted energy usage data of the building based on the building model and historical weather data;   train a predictive building model based on the predicted energy usage data, wherein the predictive building model indicates a prediction of an optimal equipment operating setting;   generate the optimal equipment operating setting based on the predictive building model; and   operate one or more pieces of building equipment based on the optimal equipment setting to control an environmental condition of the building.   
     
     
         2 . The system  claim 1 , wherein the instructions cause the one or more processors to generate a user interface comprising:
 an input interface element for receiving or defining the building model of the building;   a first output interface element indicating the optimal equipment setting; and   a second output interface element indicating energy savings resulting from operating at the optimal equipment setting.   
     
     
         3 . The system of  claim 1 , wherein the instructions cause the one or more processors to:
 generate a user input interface for graphically defining the building model;   receive user input defining the building model via the user input interface; and   generate the building model based on the user input.   
     
     
         4 . The system of  claim 1 , wherein the instructions cause the one or more processors to:
 record an electric load amount associated with the optimal equipment setting; generate a plot of the electric load amount associated with the optimal equipment setting and a plurality of other electric load amounts associated with a plurality of other equipment settings; and   cause a user device to display the plot.   
     
     
         5 . The system of  claim 1 , wherein the environmental condition is at least one of a temperature, a humidity, an air quality, or a light level;
 wherein the optimal equipment operating setting is at least one of a temperature setpoint, a humidity setpoint, an air quality setting, or a light level setting.   
     
     
         6 . The system of  claim 1 , wherein the instructions cause the one or more processors to train the predictive building model by minimizing energy usage associated with the one or more pieces of building equipment and maintaining a comfortable level of the environmental condition of the building. 
     
     
         7 . The system of  claim 1 , wherein the instructions cause the one or more processors to:
 pseudo-randomly generate one or more equipment learning settings;   generate the predicted energy usage data based on the pseudo-randomly generated one or more equipment learning settings; and   train the predictive building model based on the predicted energy usage data, the pseudo-randomly generated one or more equipment learning settings, and the historical weather data.   
     
     
         8 . The system of  claim 1 , wherein the instructions cause the one or more processors to:
 collect actual building data of the building from one or more sensors; and   re-train the predictive building model based on the actual building data.   
     
     
         9 . The system of  claim 1 , wherein the instructions cause the one or more processors to:
 receive user feedback from a user device, wherein the user feedback indicates a comfort level of a user of the user device; and   re-train the predictive building model based on the user feedback.   
     
     
         10 . A method of building control of a building, the method comprising:
 receiving, by one or more processing circuits, a building model, the building model defining a physical construction of the building;   generating, by the one or more processing circuits, predicted energy usage data of the building based on the building model and historical weather data;   training, by the one or more processing circuits, a predictive building model based on the predicted energy usage data, wherein the predictive building model indicates a prediction of an optimal equipment operating setting;   generating, by the one or more processing circuits, the optimal equipment operating setting based on the predictive building model; and   operating, by the one or more processing circuits, one or more pieces of building equipment based on the optimal equipment setting to control an environmental condition of the building.   
     
     
         11 . The method  claim 10 , further comprising generating, by the one or more processing circuits, a user interface comprising:
 an input interface element for receiving or defining the building model of the building;   a first output interface element indicating the optimal equipment setting; and   a second output interface element indicating energy savings resulting from operating at the optimal equipment setting.   
     
     
         12 . The method of  claim 10 , further comprising:
 generating, by the one or more processing circuits, a user input interface for graphically defining the building model; and   receiving, by the one or more processing circuits, user input defining the building model via the user input interface; and   generating, by the one or more processing circuits, the building model based on the user input.   
     
     
         13 . The method of  claim 10 , further comprising:
 recording, by the one or more processing circuits, an electric load amount associated with the optimal equipment setting;   generating, by the one or more processing circuits, a plot of the electric load amount associated with the optimal equipment setting and a plurality of other electric load amounts associated with a plurality of other equipment settings; and   causing, by the one or more processing circuits, a user device to display the plot.   
     
     
         14 . The method of  claim 10 , wherein the environmental condition is at least one of a temperature, a humidity, an air quality, or a light level;
 wherein the optimal equipment operating setting is at least one of a temperature setpoint, a humidity setpoint, an air quality setting, or a light level setting.   
     
     
         15 . The method of  claim 10 , further comprising training, by the one or more processing circuits, the predictive building model by minimizing energy usage and maintaining a comfortable level of the environmental condition of the building. 
     
     
         16 . The method of  claim 10 , further comprising:
 pseudo-randomly generating, by the one or more processing circuits, one or more equipment learning settings;   generating, by the one or more processing circuits, the predicted energy usage data based on the pseudo-randomly generated one or more equipment learning settings; and   training, by the one or more processing circuits, the predictive building model based on the predicted energy usage data, the pseudo-randomly generated one or more equipment learning settings, and the historical weather data.   
     
     
         17 . The method of  claim 10 , further comprising:
 collecting, by the one or more processing circuits, actual building data of the building from one or more sensors; and   re-training, by the one or more processing circuits, the predictive building model based on the actual building data.   
     
     
         18 . The method of  claim 10 , further comprising:
 receiving, by the one or more processing circuits, user feedback from a user device, wherein the user feedback indicates a comfort level of a user of the user device; and   re-training, by the one or more processing circuits, the predictive building model based on the user feedback.   
     
     
         19 . A building management system of a building, the building management system comprising:
 building equipment configured to operate based on an equipment operating setting to control an environmental condition of the building; and   a processing circuit configured to:
 receive a building model, the building model defining a physical construction of the building; 
 generate predicted energy usage data of the building based on the building model and historical weather data; 
 train a predictive building model based on the predicted energy usage data, wherein the predictive building model indicates a prediction of an optimal equipment operating setting; 
 generate the optimal equipment operating setting based on the predictive building model; and 
 cause the building equipment to operate based on the optimal equipment setting to control the environmental condition of the building. 
   
     
     
         20 . The building management system of  claim 19 , wherein the processing circuit is configured to:
 collect actual building data of the building from one or more sensors;   receive user feedback from a user device, wherein the user feedback indicates a comfort level of a user of the user device; and   re-train the predictive building model based on the actual building data and the user feedback.