Single zone HVAC prediction system and method for open plan large-scale building
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional building energy management systems (BEMS) face challenges in communicating between distant sensors and distribution boards, particularly in open plan large-scale buildings, leading to high computational costs and poor real-time performance when simulating HVAC environments.
Innovation Solution
A single zone HVAC prediction system utilizing IoT terminals and a server that applies deep learning-based prediction models to process HVAC, control, and environmental information, incorporating time embedding to provide accurate and efficient HVAC predictions without hydrodynamic calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If computational fluid dynamics (CFD) is used to physically analyze air flow and simulate HVAC environment, then simulation accuracy is improved, but computational cost increases and calculation time increases exponentially
Solution Approach 1:
The patent replaces the mechanical CFD computation system with a data-driven deep learning model. Instead of performing complex hydrodynamic calculations to simulate air flow, the system uses a pre-trained prediction model that processes sensor data (temperature, humidity, occupancy) to directly predict HVAC performance. This substitution eliminates the need for resource-intensive computational fluid dynamics while maintaining prediction accuracy through learned patterns from historical data.
Solution Approach 2:
The patent creates a virtual copy of the HVAC system behavior through a deep learning model trained on historical operational data. Rather than physically simulating air flow dynamics, the model learns to replicate HVAC performance patterns from past data, enabling accurate predictions without repeating expensive CFD computations. The model captures the essential relationships between environmental conditions and HVAC responses.
2Measurement precision
If computational fluid dynamics (CFD) is used to physically analyze air flow and simulate HVAC environment, then simulation accuracy is improved, but calculation time increases exponentially, resulting in poor real-time performance
Solution Approach 1:
The patent performs the computationally intensive work in advance by training a deep learning model on historical HVAC data and CFD results. Once trained, the model can make rapid predictions in real-time without requiring expensive computations during operation. The heavy lifting of learning complex air flow patterns is done beforehand, enabling fast inference when real-time predictions are needed.
Solution Approach 2:
The patent replaces the time-consuming CFD computation system with a lightweight deep learning inference system. The trained model processes sensor inputs and generates predictions in seconds, whereas equivalent CFD simulations would take hours or days. This substitution fundamentally changes the time complexity from exponential to linear or constant time operations.
3Ease of operation
If conventional BEMS is used to manage building energy, then basic energy monitoring is achieved, but communication between distant sensors and distribution boards in open plan large-scale buildings is difficult
Solution Approach 1:
The patent implements a centralized server that performs multiple functions: collecting data from all sensors, preprocessing information, training and hosting the prediction model, and generating control commands. This multi-functional approach eliminates the need for complex point-to-point communication infrastructure between distant sensors and distribution boards, as all communication converges at the centralized server equipped with wireless communication capabilities.
Solution Approach 2:
The patent introduces a centralized server as an intermediary between sensors and distribution boards. Instead of requiring direct communication links between distant components, the server acts as a hub that receives data from sensors via wireless communication, processes it through the prediction model, and sends control commands back to the HVAC system. This intermediary approach simplifies the communication architecture significantly.
Data Source
AI summary
Disclosed are a single zone HVAC prediction system and method for an open plan large-scale building. The single zone HVAC prediction system for an open plan large-scale building includes: a plurality of IoT terminals each installed inside and outside the building to acquire at least one of HVAC information, control information, facility-related information, and environmental information; and a server collecting the HVAC information, the control information, the facility-related information, and the environmental information from the plurality of IoT terminals, preprocessing the collected information to form a data set, applying the data set to a trained deep learning-based prediction model, and reflecting a correlation between the HVAC information and the control information and a daily pattern of the environmental information through the time embedding to output an HVAC prediction value.