PERCEPTUAL BOUNDARY EXTERNAL FLOW FIELD RECONSTRUCTION METHOD BASED ON CROSS-TIME STEP SOLUTION FORMAT

In long-distance water conveyance projects involving deep-buried long tunnel sections, due to the limitations of terrain and geological conditions, it is difficult to deploy sensors along the entire line, resulting in numerous "blind sections" in the water conveyance tunnel that cannot be ignored, which limits the scope and increases the difficulty of hydraulic safety monitoring. Based on the characteristic line method, this study proposes a cross-time-step solution format that uses sensor data to reconstruct the global flow field. This solution format uses the known data of the previous and subsequent time steps to reconstruct the hydraulic parameters of the time step to be solved. Through numerical simulation and experimental verification, this paper confirms the practicality and accuracy of this method, and proves that the cross-time-step solution format can extract key physical information from the sensor data, such as the frictional resistance and hydraulic characteristics of the pipeline, and thereby reconstruct the flow field distribution of the actual pipeline. This paper considers the possible influence of noise in practical engineering and gives a recommended method based on the comparison of three filtering methods. In addition, this study also explores the influence of grid size on the flow field reconstruction results and verifies the applicability of this method in the pipe diameter change system through a case of variable pipe diameter. Finally, this study analyzes the interference of the initial flow and the constant friction coefficient, and evaluates the anti-interference performance of this method. The sensitivity analysis results reveal that this solution format has strong robustness to the changes of the initial flow and the constant friction coefficient.