Intelligent inversion analysis of thermal parameters for distributed monitoring data

Thermal parameters are essential for temperature field calculation and construction site temperature control of massive concrete. The thermal parameters are generally obtained by laboratory measurements, which differ from the actual values. To obtain accurate thermal parameters, this paper proposes an intelligent inversion model using in-site distributed monitoring data and numerical simulation. Firstly, the distributed monitoring data with the spatial-temporal feature is clustered and denoised with a point cloud segmentation procession according to the smoothness and the distance constraint criterion. Subsequently, the form of parameters is optimized analytically according to the heat conduction equation. Afterward, the partition weighted objective function based on the classification is established. The risky regions where the temperature changes drastically are assigned higher weights. Finally, an improved whale swarm algorithm is implemented to find the optimum solution. An inversion analysis of in-situ pouring of a flow channel is conducted to validate the proposed methods. The validity of the intelligent inversion model is validated in three aspects: noise reduction effect, computational convergence speed, and inversion accuracy. The model has the potential to be applied to massive concrete structures of various fields.