Thermal Energy Distribution Modeling With ML for Unmonitored Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing thermal energy distribution models in systems like district heating fail to accurately account for consumer behavior and system characteristics beyond outdoor temperature, leading to inefficiencies and high computational complexity, especially in complex networks with unmonitored parameters.
Innovation Solution
A data-driven machine learning model characterizes thermal energy distribution by lumping unmonitored connections into a single supply and return line, using historical data to predict properties like supply temperature and flow rate without requiring detailed system component information, enabling efficient control and optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If white-box models based on physical laws are used to characterize thermal energy distribution, then measurement precision and physical accuracy are improved, but device complexity and computational complexity increase significantly
Solution Approach 1:
The patent replaces complex physical law-based models (white-box models requiring detailed system parameters, network layout, insulation characteristics) with a data-driven machine learning model (black-box approach). The neural network learns thermal energy distribution patterns directly from operational data without needing explicit physical equations, thereby reducing computational complexity while maintaining characterization accuracy.
Solution Approach 2:
The invention changes the modeling approach from parameter-intensive physical models to a data-driven approach where the system learns effective parameters from operational data. The machine learning model processes input features (temperatures, flow rates, outdoor conditions) and outputs thermal energy distribution characteristics without requiring explicit knowledge of physical parameters like pipe insulation, network topology, or component specifications.
2Measurement precision
If detailed system component information and sensor data are collected for accurate modeling, then measurement precision is improved, but loss of time and ease of operation worsen due to labor-intensive data collection
Solution Approach 1:
The machine learning model performs self-learning from available operational data without requiring manual data collection efforts. The system automatically processes sensor inputs (temperatures, flow rates, outdoor conditions) and learns thermal energy distribution patterns on its own, eliminating the need for manual surveys, expert assessments, and time-consuming model setup procedures.
Solution Approach 2:
The patent creates a virtual copy of the thermal energy system's behavior through the machine learning model. Instead of physically measuring and documenting every system parameter, the model learns to replicate system characteristics from operational data, providing accurate predictions without requiring complete physical system knowledge or extensive data collection campaigns.
3Ease of operation
If heating curve method based on outdoor temperature is used, then ease of operation is improved, but adaptability worsens as it cannot account for consumer behavior and system characteristics
Solution Approach 1:
The machine learning model incorporates feedback from multiple sources including consumer-side temperatures, flow rates, and outdoor conditions to dynamically adjust supply temperature predictions. Unlike simple heating curves that only respond to outdoor temperature, the model learns from actual system responses and consumer behavior patterns, adapting to varying demand characteristics while maintaining operational simplicity through automated predictions.
Data Source
AI summary
A method of characterizing thermal energy distribution in a thermal energy exchange system using a data-driven model, the system having a thermal energy supply unit configured to supply heating/cooling, and a plurality of receiving units configured to consume heating/cooling supplied by said thermal energy supply unit. Connections of at least a subset of multiple receiving units to a primary supply line and a primary return line of the supply unit are modelled as respectively a single supply line and a single return line, wherein values indicative of temperature and flow rate of thermal energy exchange medium at the single supply line and the single return line are unknown to the data-driven model. An input parameter set is provided to a trained machine learning model system which is configured to output at least one prediction value, the at least one prediction value including a value indicative of a property of the thermal energy exchange system, and wherein the input parameter set includes at least a value indicative of a temperature of thermal energy exchange medium supplied by the thermal energy supply unit via the primary supply line.