Error correction for predictive schedules for a thermostat
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing HVAC systems lack the capability to determine whether a user is present or away without user input, leading to inefficiencies in energy consumption and wear on components, as they cannot automatically adjust set point temperatures without this information.
Innovation Solution
Employing machine learning to predict user behavior and preferences by collecting data from user devices, interactions, and network connections to generate a predicted occupancy schedule, allowing the HVAC system to adjust temperatures for energy savings and reduce wear.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If the HVAC system automatically adjusts set point temperatures without user input, then energy consumption is reduced and component wear is minimized, but the system cannot accurately determine user presence or absence leading to incorrect temperature adjustments
Solution Approach 1:
The patent introduces an intermediary machine learning model that indirectly infers user presence by analyzing patterns in thermostat interactions and environmental data. This mediator bridges the gap between available data and accurate presence determination, allowing the system to predict occupancy without direct user input while maintaining reasonable accuracy for energy-saving adjustments
Solution Approach 2:
The system employs self-service by using its own historical interaction data and environmental sensors to train a machine learning model that autonomously determines user presence patterns. The thermostat analyzes its own usage history, temperature adjustment patterns, and local data to independently make presence predictions without requiring external user input or additional specialized sensors
2Measurement precision
If the HVAC system collects and analyzes extensive user interaction data to improve presence prediction accuracy, then user presence detection improves, but system complexity increases
Solution Approach 1:
The patent uses copying by creating a virtual model (machine learning model) that replicates user presence patterns based on historical data. Instead of requiring complex real-time analysis infrastructure, the system creates a simplified computational copy of user behavior patterns that can be executed on standard thermostat hardware, reducing the complexity burden while maintaining prediction accuracy
Solution Approach 2:
The system performs preliminary action by pre-training the machine learning model during off-peak periods using accumulated historical data. This advance preparation allows the model to be ready for real-time predictions without requiring complex real-time processing, as the heavy computational lifting of pattern recognition is completed beforehand during model training phases
3Reliability
If the HVAC system makes conservative temperature adjustments to maintain user comfort, then user comfort is preserved, but energy saving benefits are reduced
Solution Approach 1:
The patent applies dynamics by making temperature adjustments adaptive rather than static. The machine learning model continuously refines its presence predictions based on new data, allowing the system to dynamically adjust its confidence level and corresponding temperature adjustment aggressiveness. This dynamic approach enables the system to be more aggressive when confidence is high and more conservative when confidence is lower, optimizing both energy savings and comfort reliability
Solution Approach 2:
The system implements feedback by monitoring user responses to temperature adjustments and using this information to refine the machine learning model. When users manually override predicted temperature settings, this feedback is fed back into the model to improve future predictions, allowing the system to learn from actual user behavior patterns and progressively optimize the balance between energy savings and comfort reliability
Data Source
AI summary
A heating, ventilation, and air conditioning (HVAC) control device is configured to record a plurality of actual occupancy statuses, to determine a plurality of corresponding predicted occupancy statuses, and to compare the plurality of predicted occupancy statuses to the plurality of actual occupancy statuses. The device is further configured to identify conflicting occupancy statuses based on the comparison. A conflicting occupancy status indicates a difference between an actual occupancy status and a corresponding predicted occupancy status. The device is further configured to identify timestamps corresponding with the conflicting occupancy statuses, to identify historical occupancy statuses corresponding with the identified timestamps, and to update the conflicting occupancy statuses in the predicted occupancy schedule with the historical occupancy statuses.