Parameter estimation apparatus, air-conditioning system evaluation apparatus, parameter estimation method, and non-transitory computer readable medium
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating air-conditioning systems in buildings face challenges in accurately simulating energy consumption and performance due to the need for iterative calculations with a large number of parameters, leading to increased time and computational load, and often result in variable parameter values that should be constant, such as heat transmissibility coefficients, affecting simulation accuracy.
Innovation Solution
A parameter estimation apparatus with a model reduction processor and parameter estimator that generates reduced order models based on measurement data and conditions for model order reduction, allowing for the estimation of parameter values in a simplified simulation model, thereby reducing the number of parameters and maintaining constant values where necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If the number of parameters in the simulation model is reduced, then the processing time and computational load are reduced, but the simulation accuracy may decrease
Solution Approach 1:
The patent changes the parameter identification approach by using iterative calculation with an objective function to optimize parameter values. This allows for accurate parameter estimation even in reduced-order models by systematically adjusting parameters to minimize the difference between simulated and actual measurement data, thus maintaining simulation accuracy while working with fewer parameters.
Solution Approach 2:
The patent implements a feedback mechanism where simulation results are compared with actual measurement data, and parameter values are iteratively adjusted based on this comparison. The objective function quantifies the error between simulation and measurement, guiding the iterative optimization process to achieve accurate parameter identification without requiring a full-complexity model.
2Device complexity
If multiple regression analysis is used to reduce parameters, then the number of parameters is reduced, but individual values for constant parameters vary across classification conditions
Solution Approach 1:
The patent applies a universal parameter identification approach that works across different classification conditions without requiring separate parameter sets. By using iterative optimization with an objective function, the same methodology can identify consistent parameter values regardless of the specific classification condition, making the approach universally applicable while maintaining parameter consistency.
Solution Approach 2:
The patent performs preliminary classification of measurement data into different conditions before parameter identification. This preliminary action allows the system to recognize when data belongs to different classification categories, enabling the subsequent iterative parameter identification to maintain consistency for parameters that should be constant across conditions while still accommodating condition-specific variations where appropriate.
Data Source
AI summary
A parameter estimation apparatus of an embodiment of the present invention is provided with a model reduction processor and a parameter estimator. The model reduction processor generates reduced order models by reducing order of a simulation model on the basis of measurement data sets and conditions for model order reduction possibility. The parameter estimator estimates values of parameters of the reduced order models on the basis of the reduced order models and the measurement data sets corresponding to the reduced order models. Further, after estimating a first value for a first parameter of a first reduced order model based on a first measurement data set, the parameter estimator applies the first value to the first parameter of a second reduced order model based on a second measurement data set.