Techniques for performing an emissions demand response (EDR) event are described. In an example, a cloud-based HVAC control system may obtain a first emissions rate forecast and generate an EDR event with a start time and end time based on the first emissions rate forecast. The EDR event may then be transmitted to a thermostat and stored in a memory of the thermostat. At the start time, the thermostat may commence controlling an HVAC system according to the EDR event. After the start time and prior to the end time, the cloud-based HVAC control system may obtain a second emissions rate forecast and generate a modified EDR event with a modified end time. The modified EDR event may be transmitted to the thermostat before the end time and/or the modified end time whereupon the thermostat may control the HVAC system accordingly until the modified end time is reached.

A system for providing air conditioning to a living space and heating potable water. The system comprising a heat pump circuit comprising a compressor for circulating a refrigerant around the heat pump circuit, a first condenser, a second condenser and an evaporator. The evaporator being adapted to receive a first flow of air from an air inlet to transfer heat from the first flow of air to the refrigerant. The first condenser being adapted to receive a flow of water to transfer heat from the refrigerant to the water. The second condenser being adapted to receive a second flow of air to transfer heat from the refrigerant to the second flow of air. The first flow being provided from the evaporator to a living space by an air outlet.

An environmental control system for a building space including heating, ventilation, or air conditioning (HVAC) equipment that operates to control a temperature of the building space. The system includes lighting equipment that operates to control a luminosity and affect a heat load disturbance for the building space. The system includes an environmental controller including a processing circuit configured to predict the heat load disturbance based on potential operating states of the lighting equipment over a time period. The heat load disturbance affects the temperature of the building space. The processing circuit is configured to generate control decisions for the HVAC and lighting equipment based on the predicted heat load disturbance and subject to constraints on the temperature and luminosity of the building space. The processing circuit is configured to operate the HVAC and lighting equipment based on the control decisions.

An environmental control system for a building space including heating, ventilation, or air conditioning (HVAC) equipment that operates to control a temperature of the building space. The system includes lighting equipment that operates to control a luminosity and affect a heat load disturbance for the building space. The system includes an environmental controller including a processing circuit configured to predict the heat load disturbance based on potential operating states of the lighting equipment over a time period. The heat load disturbance affects the temperature of the building space. The processing circuit is configured to generate control decisions for the HVAC and lighting equipment based on the predicted heat load disturbance and subject to constraints on the temperature and luminosity of the building space. The processing circuit is configured to operate the HVAC and lighting equipment based on the control decisions.

Data including a power consumption amount and an ambient temperature of a heat exchanger that discharges heat to an outside of a cooling facility in an operation mode in which the facility is operated in an operation state that requires less power consumption than usual is extracted as degradation determination reference data before degradation determination, and data including the power consumption amount and the ambient temperature in the operation mode in which the cooling facility is operated in the operation state that requires less power consumption than usual is extracted as determination data, whereby the increase in the power consumption amount due to the influence of the disturbance that changes a temperature of an inside of a box-shaped housing can be excluded, and the degradation of the cooling facility can be correctly determined by grasping the increase in the power consumption amount due to aging degradation.

Architectures or techniques are presented that can prioritize operating a consumption device in a manner that is efficient in terms of consumption of a resource over satisfying a specified demand assigned to the consumption device. This re-prioritizing can be performed in response to a price of the resource exceeding a threshold.

Architectures or techniques are presented that can prioritize operating a consumption device in a manner that is efficient in terms of consumption of a resource over satisfying a specified demand assigned to the consumption device. This re-prioritizing can be performed in response to a price of the resource exceeding a threshold.

Systems and methods for training a reinforcement learning (RL) model for HVAC control are disclosed herein. Simulated experience data for the HVAC system is generated or received. The simulated experience data is used to initially train the RL model for HVAC control. The HVAC system operates within a building using the RL model and generates real experience data. A determination may be made to retrain the RL model. The real experience data is used to retrain the RL model. In some embodiments, both the simulated and real experience data are used to retrain the RL model. Experience data may be sampled according to various sampling functions. The RL model may be retrained multiple times over time. The RL model may be retrained less frequently over time as more real experience data is used to train the RL model.

Systems and methods for training a reinforcement learning (RL) model for HVAC control are disclosed herein. Simulated experience data for the HVAC system is generated or received. The simulated experience data is used to initially train the RL model for HVAC control. The HVAC system operates within a building using the RL model and generates real experience data. A determination may be made to retrain the RL model. The real experience data is used to retrain the RL model. In some embodiments, both the simulated and real experience data are used to retrain the RL model. Experience data may be sampled according to various sampling functions. The RL model may be retrained multiple times over time. The RL model may be retrained less frequently over time as more real experience data is used to train the RL model.

An electricity automation application may automate control over an HVAC system in a consumer's home or other building to reduce or eliminate electricity consumption during a high-price or high-demand time interval. The consumer may grant authorization for the application to control the HVAC system at appropriate timepoints while maintaining an inside air temperature that is between minimum and maximum temperature setpoints set by the consumer. The application models temperature gains and losses for the home or building as a function of inside and outside temperatures and HVAC operating status. If a high-price or high-demand time interval is predicted, the application may determine a timepoint to precool or preheat the home or building to reduce or eliminate electricity use during the duration of the high-price or high-demand time interval.