Optimized precooling of structures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Model predictive control for building precooling faces challenges due to simplifications and errors in weather forecasts, leading to potential comfort limit excesses and consequential losses, especially in multi-zone buildings where optimization becomes computationally complex.
Innovation Solution
An aggregated single-zone model predictive control (MPC) method is formulated to simplify the optimization problem by aggregating multiple zones, determining optimal actions, simulating indoor quality trajectories, and setting zone temperature setpoints to comply with comfort limits, using zone weighting and disaggregation techniques to minimize temperature differences and changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple zones are controlled individually with model predictive control, then comfort limits can be maintained, but computational complexity increases significantly
Solution Approach 1:
The patent aggregates multiple thermal zones into a single equivalent zone by combining their thermal parameters (capacities, resistances, temperature coefficients) into aggregate values. This merging approach maintains comfort compliance for the overall building while dramatically reducing computational complexity by solving a single MPC problem instead of multiple independent ones.
Solution Approach 2:
The patent segments the computational problem by separating the aggregation phase (combining zone parameters) from the control phase (solving MPC for aggregate zone), enabling efficient computation while preserving the essential thermal dynamics of individual zones through weighted contributions in the aggregate model.
2Device complexity
If simplified model predictive control is used, then computational complexity is reduced, but accuracy in predicting indoor temperature deteriorates
Solution Approach 1:
The patent transforms the thermal model parameters from zone-specific values to aggregate values that represent the combined thermal behavior of multiple zones. By changing the parameter representation (from individual zone parameters to aggregate parameters), the model maintains predictive accuracy while enabling simplified computation through the aggregated single-zone formulation.
3Use of energy by moving object
If precooling is optimized without considering weather forecast errors, then energy consumption is reduced, but comfort limit excesses occur
Solution Approach 1:
The patent incorporates weather forecast error margins into the MPC optimization by adjusting the precooling strategy to account for potential forecast deviations. This beforehand cushioning approach reduces energy consumption compared to conservative methods while maintaining comfort compliance by building in a safety margin against forecast errors.
Solution Approach 2:
The patent uses a feedback mechanism where the MPC controller continuously monitors actual temperature deviations from forecast predictions and adjusts subsequent precooling actions accordingly. This feedback loop enables energy-efficient operation by learning from forecast accuracy and adapting the control strategy to maintain comfort limits despite weather prediction uncertainties.
Data Source
AI summary
A method includes aggregating multiple zones of an indoor structure, each zone having associated comfort limits, formulating an aggregated single zone model predictive control (MPC) problem representative of the multiple zones for a heating ventilation and air conditioning (HVAC) system, determining optimal aggregated actions as a function of the aggregated single zone model predictive control problem, simulating an optimal trajectory of indoor qualities, and determining zone temperature setpoints to comply with the comfort limits for each zone and pre-cool the indoor structure.