基于FMU标准与优化算法的管网模型参数反演方法和装置
By constructing a pipeline network physical model in the Modelica environment and using a neural network surrogate model and staggered mesh method, combined with a three-stage optimization algorithm, the problems of long time consumption and low accuracy in pipeline network model parameter calibration in the existing technology are solved, and efficient and accurate parameter inversion is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HANGZHOU STEAM TURBINE ENG
- Filing Date
- 2026-02-13
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies suffer from time-consuming, reliance on human experience, and difficulty in finding the global optimal solution in pipeline network model parameter calibration. Furthermore, existing toolchains struggle to achieve efficient and accurate parameter inversion.
A hybrid strategy based on the FMU standard and optimization algorithms is adopted. By constructing a pipeline physical model in the Modelica modeling environment, a neural network proxy model is used to replace the time-consuming physical simulation. Combined with the staggered mesh method and the three-stage optimization algorithm, cross-platform model coupling and parameter inversion are achieved.
It achieves efficient and accurate automated inversion of pipeline system parameters, significantly improving calibration efficiency and accuracy, avoiding local optima traps, and is suitable for complex pipeline systems.
Smart Images
Figure CN121706616B_ABST