.png){kind=icon}
.png){kind=icon}
Return
A novel method for identifying thermal parameters based on Virtual Element Method and Genetic Algorithm-optimized Backpropagation
.png){kind=icon}
PRE
AI
DOI：10.1016/j.icheatmasstransfer.2026.111310.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In this work, a new parameter identification method combining the Virtual Element Method (VEM) with the Genetic Algorithm-optimized Backpropagation (GA-BP) intelligent algorithm is proposed. This approach fully exploits the complementary strengths of both components: the virtual element method can effectively solve partial differential equations on complex geometric domain with arbitrary polyhedral meshes, while the GA-BP intelligent algorithm combines the strong nonlinear mapping capability of the BP neural networks with the powerful global search capability of the genetic algorithm. The proposed combination is designed to overcome the performance limitations of traditional parameter identification methods under large measurement errors. Random noise is added to temperature data of measurement points to simulate measurement errors and prior uncertainty, thereby constructing identification scenarios that better approximate real conditions. Multiple types of unknown parameters, such as nonlinear thermal conductivity and transient boundary heat flux, are then identified. In addition, the study also analyzes the impact of sampling point locations on identification results and conducts comparative research between the GA-BP algorithm and other heuristic algorithms. The final results demonstrate that the method proposed in this study can rapidly and efficiently identify multiple thermal parameters, exhibiting exceptional robustness towards erroneous data while maintaining high accuracy.
Keywords:
Virtual Element Method
Genetic Algorithm-optimized Backpropagation
Thermal Parameter Identification
Nonlinear Thermal Conductivity
Transient Boundary Heat Flux
Journal
IF:
6.4
Papers:
1.0W
Citations:
2.5W