Decision Tree Control Engine for Simplified Building Automation MPC
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Building automation systems (BAS) often operate sub-optimally due to the complexity and knowledge barrier of implementing Model-based Predictive Control (MPC), which requires significant historical data and computational resources, making it difficult for practitioners to configure and debug, especially in resource-constrained environments like building control edge devices.
Innovation Solution
A control engine system and method that utilizes a graphical software interface to generate empirical rule extraction heuristics through decision trees, including a data preprocessing engine, random forest generator, and BAS code generator, to simplify MPC configuration and deployment, enabling energy savings while being explainable and implementable by field practitioners.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If Model-based Predictive Control (MPC) is implemented in building automation systems, then control performance and energy savings are improved, but system complexity and knowledge barrier increase
Solution Approach 1:
The patent introduces an intermediary processing layer that transforms complex MPC algorithms into simplified control decisions. This intermediary layer handles the computational complexity and mathematical operations, presenting only simple control outputs to the building automation system, thus resolving the contradiction between achieving energy savings through MPC and maintaining system simplicity.
Solution Approach 2:
The patent creates simplified copies or approximations of the full MPC algorithm that can be executed in resource-constrained environments. By using reduced-order models or pre-computed control strategies, the system achieves comparable energy savings without requiring the full computational complexity of detailed MPC implementations.
2Manufacturing precision
If detailed MPC algorithms are used, then control precision is improved, but ease of operation and debugging deteriorate
Solution Approach 1:
The patent segments the control system into modular components with clearly defined interfaces. By dividing the complex MPC algorithm into separate functional blocks (prediction module, optimization module, control module), each with specific responsibilities, the system maintains high control precision while enabling easier operation and debugging through modular architecture.
3Measurement precision
If historical time series data is collected for MPC training, then model accuracy is improved, but loss of time and computational resources increase
Solution Approach 1:
The patent performs preliminary actions by pre-processing and pre-training MPC models offline using historical data. By completing the computationally intensive model training and validation before deployment, the system achieves high model accuracy without requiring extensive real-time data collection and processing, thus reducing time loss during operational phases.
Data Source
AI summary
The present disclosure provides a control engine system connected to a BAS, the control engine system includes a data preprocessing engine, a random forest generator, a tree selector, and a BAS code generator. The data preprocessing engine receives a data from the BAS and is configured for performing preprocessing on the data and generating a training data accordingly. The random forest generator is connected to the data preprocessing engine, the random forest generator receives the training data and generates a random forest accordingly. The tree selector is connected to the random forest generator, and the tree selector receives the random forest for selecting a final decision tree. The BAS code generator is connected to the tree selector, the BAS code generator receives the final decision tree and encodes a supervisory decision tree according to the final decision tree, the BAS code generator outputs the supervisory decision tree to the BAS.