Predictive building control system with neural network based constraint generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current building control systems lack an efficient method to optimize energy consumption by HVAC equipment, as they fail to effectively generate and enforce constraints on temperature setpoints, leading to suboptimal energy usage and increased costs.
Innovation Solution
A predictive building control system utilizing a neural network model to generate constraints for HVAC equipment, which classifies and trains on setpoints to optimize energy usage, incorporating inputs such as outdoor temperature, occupancy status, and valve positions to adjust setpoints dynamically.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If traditional building control systems are used without neural network-based constraint generation, then the system complexity is lower, but energy consumption optimization is insufficient
Solution Approach 1:
A neural network model is introduced as an intermediary component between the optimization controller and the equipment controller. The neural network generates temperature setpoint constraints that the optimization controller uses to minimize energy consumption while maintaining comfort. This intermediary enables sophisticated energy optimization without requiring complete system redesign.
Solution Approach 2:
The neural network model pre-generates constraint values for temperature setpoints based on historical and real-time data before the optimization process occurs. By preparing constraints in advance, the system reduces computational complexity during real-time optimization while achieving energy efficiency.
2Measurement precision
If neural network model training is implemented to improve constraint accuracy, then constraint performance improves, but computational time and processing requirements increase
Solution Approach 1:
The neural network model is trained offline using historical operational data to learn optimal constraint patterns. Once trained, the model rapidly generates constraints during real-time operation without requiring intensive computation. This preliminary training phase separates the computationally intensive learning process from real-time control, achieving both accuracy and speed.
Solution Approach 2:
The neural network creates a simplified computational representation (copy) of the complex relationship between environmental conditions, equipment operation, and energy consumption. This copied knowledge in the form of trained network weights enables fast constraint generation without repeatedly performing complex simulations or calculations.
3Loss of energy
If dynamic setpoint adjustment based on real-time data is implemented, then energy cost reduction improves, but system adaptability requirements increase
Solution Approach 1:
The system continuously receives feedback from sensors measuring outdoor temperature, humidity, occupancy, and equipment status. This real-time feedback is fed into the neural network and optimization controller to dynamically adjust temperature setpoints and equipment operation, reducing energy costs while adapting to changing conditions. The feedback loop enables the system to respond automatically to environmental and operational changes.
Data Source
AI summary
A predictive building control system includes equipment operable to provide heating or cooling to a building and a predictive controller. The predictive controller includes one or more optimization controllers configured to perform an optimization to generate setpoints for the equipment at each time step of an optimization period subject to one or more constraints, a constraint generator configured to use a neural network model to generate the one or more constraints, and an equipment controller configured to operate the equipment to achieve the setpoints generated by the one or more optimization controllers at each time step of the optimization period.