Building Control Model Configuration for Accurate System Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional system identification methods for building equipment require hard-coding model forms, leading to inaccurate representations of physical systems, necessitating repeated code rewriting and recompilation, which is time-consuming and requires high technical qualifications.
Innovation Solution
A building management system that allows users to input and edit model forms through a graphical user interface, generating machine-executable steps to determine parameter values, calculate cost functions, and update system models without requiring code editing, facilitating online configurable system identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If model form is hard-coded into software package, then system identification can be performed with conventional approaches, but model form is not sufficiently representative of physical system leading to inaccurate system identification
Solution Approach 1:
The patent implements dynamic configurability of model forms through a graphical user interface that allows users to select and modify model parameters in real-time. The system transitions from static hard-coded models to dynamic user-configurable models, enabling the model structure to adapt to different physical systems and requirements without recompilation.
Solution Approach 2:
The patent enables parameter changes by allowing users to modify model form parameters through a graphical interface. The system parses user-defined model forms, generates executable code dynamically, and recompiles without requiring source code access. This transforms fixed parameters into adjustable parameters that can be changed online to better represent physical systems.
2Reliability
If technician rewrites source code to program new model form operations, then system identification accuracy can be improved, but substantial time is required for rewriting and recompiling code
Solution Approach 1:
The patent implements preliminary action by pre-compiling template code structures and parsing routines that can dynamically generate executable code from user-defined model forms. Instead of rewriting entire source codes, the system prepares reusable code templates that are instantiated with specific model parameters, dramatically reducing the time required to implement new model forms.
Solution Approach 2:
The patent uses copying by generating executable code through parsing and instantiation of model form definitions rather than manual source code rewriting. The system creates copies of template structures filled with specific model parameters, enabling rapid deployment of new model forms without direct source code modification.
3Reliability
If technician repeatedly reprograms software multiple times to identify accurate model form, then sufficiently accurate model can be achieved, but high technical qualifications are required and costs increase
Solution Approach 1:
The patent implements self-service by enabling users to independently configure and modify model forms through a graphical user interface without requiring programming expertise. The system automatically parses user definitions, generates executable code, and performs compilation, allowing non-programmers to perform tasks that previously required skilled technicians.
Solution Approach 2:
The patent introduces an intermediary layer between the user and the compiled code through a graphical user interface and automatic code generation system. This intermediary translates user-friendly model form definitions into executable code automatically, eliminating the need for users to directly manipulate source code and reducing qualification requirements.
4Device complexity
If conventional system identification approaches are used, then code structure remains simple and fixed, but the system lacks adaptability to different model forms and requires frequent modifications
Solution Approach 1:
The patent applies segmentation by separating the model form definition from the executable code through a parsing architecture. The system divides the code structure into template segments that can be dynamically instantiated with different model parameters, maintaining code simplicity while enabling adaptability to various model forms through modular composition.
Data Source
AI summary
A building management system includes building equipment operable to affect a variable state or condition of a building and a control system configured to receive a user input indicating a model form. The model form includes a plurality of matrices having a plurality of elements defined in terms of a plurality of parameters. The control system is configured to parse the model form to generate a sequence of machine-executable steps for determining a value of each of the plurality of elements based on a set of potential parameter values, identify a system model by executing the sequence of machine-executable steps to generate a set of parameter values for the plurality of parameters, generate a graphical user interface that illustrates a fit between predictions of the identified system model and behavior of the variable state or condition of the building, and control the building equipment using the identified system model.