System identification and model development

A controller for a building system, the controller comprising:

• a communications interface configured to receive training data comprising input data and output data, the output data indicative of a state of the building system affected by both the input data and an extraneous disturbance comprising an uncontrolled input to the building system;
• and
• a processing circuit comprising a processor and memory, wherein the processing circuit is configured to perform a two-stage optimization process to identify system parameters and Kalman gain parameters of a dynamic model for the building system, the two-stage optimization process comprising:
• a first stage in which the processing circuit is configured to filter the training data to remove an effect of the extraneous disturbance from the output data and uses the filtered training data to identify the system parameters, and
• a second stage in which the processing circuit is configured to use the non-filtered training data to identify the Kalman gain parameters;
• wherein the processing circuit is configured to use the dynamic model to generate a setpoint for the building system and the building system is configured to use the setpoint to affect the state indicated by the output data.

A controller for a building system, the controller comprising:

• a communications interface configured to receive training data comprising input data and output data, the output data indicative of a state of the building system affected by both the input data and an extraneous disturbance comprising an uncontrolled input to the building system;
• and
• a processing circuit comprising a processor and memory, wherein the processing circuit is configured to perform a two-stage optimization process to identify parameters of a dynamic model for the building system, the two-stage optimization process comprising:
• a first stage in which the processing circuit is configured to identify system parameters of the dynamic model, and
• a second stage in which the processing circuit is configured to identify Kalman gain parameters of the dynamic model while holding the system parameters at constant values identified during the first stage;
• wherein the processing circuit is configured to use the dynamic model to generate a setpoint for the building system and the building system is configured to use the setpoint to affect the state indicated by the output data.

A computer-implemented method for identifying parameters of a dynamic model for a building system, the method comprising:

• receiving, at a controller for the building system, training data comprising input data and output data, the output data indicative of a state of the building system affected by both the input data and an extraneous disturbance comprising an uncontrolled input to the building system;
• performing, by the controller, a two-stage optimization process to identify system parameters and Kalman gain parameters of the dynamic model, the two-stage optimization process comprising:
• a first stage in which the controller filters the training data to remove an effect of the extraneous disturbance from the output data and uses the filtered training data to identify the system parameters, and
• a second stage in which the controller uses the non-filtered training data to identify the Kalman gain parameters;
• and
• using the dynamic model to generate, by the controller, a setpoint for the building system, wherein the building system is configured to use the setpoint to affect the state indicated by the output data.