Heuristic method of automated and learning control, and building automation systems thereof
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current building automation systems lack a unified operational model to effectively manage energy interrelationships and adapt to complex, ad hoc building topologies, relying on model-free control loops that are difficult to optimize and require human intervention, limiting their ability to scale and perform continuous commissioning.
Innovation Solution
A closed-loop, heuristically tuned, model-based control algorithm that simulates external factors like weather and occupancy, allowing for adaptive control and real-time monitoring, fault detection, and optimization, using a physical system model to generate predictive control sequences and adjust control actions based on past data and system performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If model-free control loops are used, then implementation is simple, but the system becomes difficult to manage and optimize as complexity increases
Solution Approach 1:
The patent transforms the control approach from model-free to model-based, changing the fundamental parameter of control methodology. This enables the system to handle complex building topologies and energy interrelationships by using physical models that capture system behavior, making management and optimization feasible even as system complexity increases
2Measurement precision
If automated commissioning methods are used with occupancy-free training periods, then system calibration is achieved, but retro-commissioning and continuous commissioning are limited
Solution Approach 1:
The patent implements dynamic commissioning capability where the system can adapt and learn continuously during occupancy rather than requiring static occupancy-free periods. The model-based approach allows the system to update its parameters and improve calibration accuracy while the building is in normal operation, enabling both retro-commissioning and continuous commissioning
Solution Approach 2:
The system uses feedback from actual building operation data to continuously refine and update the physical models. This closed-loop learning process allows the system to maintain calibration accuracy while adapting to changing conditions, removing the limitation of requiring occupancy-free training periods
3Reliability
If physical models are used as reference with data mining, then control strategies are created, but human intervention is required limiting applicability
Solution Approach 1:
The patent implements self-service capability where the model-based control system automatically creates and optimizes control strategies without requiring human intervention. The system uses its physical models to autonomously analyze building data, identify optimization opportunities, and implement control actions, making the technology applicable to buildings without highly skilled engineering teams
4Adaptability or versatility
If model-based control is implemented, then adaptive control and optimization are achieved, but system complexity increases
Solution Approach 1:
The patent segments the building system into distinct physical models representing different components and energy flows. This modular approach to modeling allows the system to capture complex interrelationships while maintaining manageable complexity through structured, component-level representations that can be independently developed and updated
Data Source
AI summary
Apparatuses, systems, and methods of physical-model based building automation using in-situ regression to optimize control systems are presented. A simulation engine is configured to simulate a behavior or a controlled system using a physical model for the controlled system. A data stream comprises data from a controlled system. A training loop is configured to compare an output of a simulation engine to a data stream using a heuristic so that a physical model is regressed in a manner that the output of the simulation engine approaches the data stream.