A heat exchanger heat transfer coefficient identification method based on grey box model and pattern clustering
By using gray box model and pattern clustering method, combined with Gaussian mixture model and genetic algorithm to optimize heat transfer coefficient, the deployment problem of heat exchanger heat transfer coefficient identification is solved, realizing dynamic reflection of performance changes and accurate identification, which is suitable for real-time control of district heating systems.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- TIANJIN UNIV
- Filing Date
- 2026-04-27
- Publication Date
- 2026-07-21
AI Technical Summary
Existing methods for identifying the heat transfer coefficient of heat exchangers have problems such as being unable to be directly deployed, ignoring thermal inertia and heat transfer delay, and failing to dynamically reflect performance changes.
A gray box model and pattern clustering method is adopted to calculate the drag coefficient by collecting data, perform clustering using Gaussian mixture model and Bayesian information criterion, and optimize the heat transfer coefficient and water volume by combining genetic algorithm to establish a dynamic heat transfer model, construct a heat transfer coefficient spectrum and embed it into the control system.
It enables deployment without cumbersome calibration and large amounts of tag data, can dynamically identify changes in heat exchanger performance, improves the accuracy and stability of heat transfer coefficient identification, and is suitable for real-time prediction and control.
Smart Images

Figure CN122432741A_ABST