Intelligent scheduling method and system for tunnel drainage system based on multi-source data fusion

Abstract

The invention relates to the field of intelligent scheduling of drainage systems, in particular to an intelligent scheduling method and system for a tunnel drainage system based on multi-source data fusion, and the method comprises the steps: obtaining a final rainfall model after space-time calibration through obtaining data and carrying out space-time alignment; acquiring a seepage field and a surface runoff field of the tunnel to obtain a nonlinear mapping relation between rainfall intensity and tunnel catchment flow; obtaining tunnel predicted catchment flow through a nonlinear mapping relation, collecting water-collecting well liquid level data of a tunnel drainage system and operation output data of a drainage pump group, and generating drainage pump group control reference parameters; and when it is detected that the difference between the predicted catchment flow and the actually measured drainage flow of the adjacent tunnel is smaller than a preset difference threshold value, a frequency adjustment instruction and a start-stop control sequence are output in combination with the drainage pump set control reference parameters. The space-time precision of rainfall prediction is improved, the interference influence of catchment flow among tunnels is fully considered, and the accurate scheduling of a tunnel drainage system is improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent scheduling of drainage systems, and specifically to an intelligent scheduling method and system for tunnel drainage systems based on multi-source data fusion. Background Art

[0002] With the acceleration of the urbanization process and the continuous advancement of traffic infrastructure construction, tunnel engineering, as an important part of urban underground space development, the safety and reliability of its drainage system have received increasing attention. The tunnel drainage system undertakes the important functions of draining accumulated water in the tunnel, preventing groundwater leakage, and ensuring the safety of the tunnel structure. Especially in seasons or regions with heavy rainfall, the intelligent scheduling of the drainage system is crucial for ensuring the safe operation of the tunnel.

[0003] The operating conditions of the tunnel drainage system change dynamically with the rainfall intensity, tunnel seepage characteristics, and the influence of adjacent tunnel drainage systems. However, the existing technologies lack an effective dynamic adjustment mechanism. Traditional methods often rely on relatively single data sources, such as rain gauges or water level sensors in the tunnel, and cannot comprehensively and accurately reflect the rainfall conditions around the tunnel and the seepage state inside the tunnel. Especially at the intersections of tunnel groups, there is a problem of water convergence interference between different drainage systems, and the existing hydrological models cannot accurately reflect the water flow coupling characteristics under dynamic boundary conditions, resulting in inaccurate water volume prediction results and unreasonable drainage scheduling.

[0004] Therefore, there is an urgent need for an intelligent scheduling method for tunnel drainage systems based on multi-source data fusion to achieve the intelligent scheduling of drainage pump groups and improve the safety and reliability of tunnel drainage systems. Summary of the Invention

[0005] (1) Technical Problems to be Solved The purpose of the present invention is to provide an intelligent scheduling method and system for tunnel drainage systems based on multi-source data fusion, so as to solve the problems that the traditional tunnel drainage system scheduling method has a single data source, insufficient spatio-temporal calibration ability, does not consider the non-linear mapping relationship between rainfall and water convergence, and the drainage interference effect of adjacent tunnels, resulting in low drainage prediction accuracy and unreasonable drainage scheduling.

[0006] (2) Technical Solutions To achieve the above purpose, on the one hand, the present invention provides an intelligent scheduling method for tunnel drainage systems based on multi-source data fusion, and the method includes: S1. Obtain meteorological satellite data, ground-based radar data, and seepage pressure data in the tunnel, and perform spatio-temporal alignment through a spatio-temporal nested data fusion method and dynamically assign weight coefficients to obtain a finally spatio-temporally calibrated rainfall model.

[0007] S2. Obtain the seepage field and surface runoff field of the tunnel, and obtain the non-linear mapping relationship between rainfall intensity and tunnel catchment flow through iterative calculation.

[0008] S3. Obtain the predicted catchment flow of the tunnel through the non-linear mapping relationship from the rainfall intensity spatio-temporal distribution data of the final rainfall model, obtain the sump liquid level data of the tunnel drainage system and the operating output data of the drainage pump group, and generate the control reference parameters of the drainage pump group; output the frequency adjustment command and start-stop control sequence of the drainage pump group according to the predicted catchment flow of the tunnel and the control reference parameters of the drainage pump group.

[0009] Further, the method for obtaining the finally rainfall model after spatio-temporal calibration by spatio-temporal nested data fusion method for spatio-temporal alignment and dynamic weight coefficient allocation includes: Improve the spatial resolution of meteorological satellite data to the same grid scale as the ground-based radar data by bilinear interpolation method, and synchronize the timestamps of the ground-based radar data to match the sampling frequency of the seepage pressure data; construct a multi-dimensional feature matrix based on the meteorological satellite data, ground-based radar data and seepage pressure data after spatio-temporal alignment, and each grid node of the multi-dimensional feature matrix includes the rainfall intensity of the meteorological satellite data, the rainfall intensity of the ground-based radar data, seepage pressure, spatial coordinates and timestamps.

[0010] Dynamically calculate the weight coefficients of the meteorological satellite data, ground-based radar data and seepage pressure data. Among them, the weight coefficient of the meteorological satellite data is determined according to the exponential decay characteristic of its historical error, the weight coefficient of the ground-based radar data is calculated according to the spatial variance of the rainfall intensity within its grid, and the weight coefficient of the seepage pressure data is dynamically adjusted according to its real-time Pearson correlation coefficient with the surface runoff data, and the surface runoff data is obtained by inverting the geometric parameters of the surface runoff section.

[0011] Perform weighted calculation on the rainfall intensity of the meteorological satellite data, the rainfall intensity of the ground-based radar data, and the seepage pressure of each grid node in the multi-dimensional feature matrix with the corresponding weight coefficients to generate an initial calibrated rainfall model.

[0012] Real-time detect the correlation coefficient between the time series data of the seepage rate and the time series data of the rainfall intensity in the preset fan-shaped area along the tunnel axis direction. When the detected correlation coefficient is lower than the preset correlation coefficient threshold, recalculate the weight coefficient of the seepage pressure data and locally correct the spatial variance of the ground-based radar data; perform iterative optimization on the rainfall intensity of the grid nodes in the tunnel group intersection area in the initial calibrated rainfall model according to the corrected weight coefficients until the difference between the predicted rainfall intensity and the measured rainfall intensity is less than the preset rainfall intensity difference threshold, and then output the finally rainfall model after spatio-temporal calibration.

[0013] Furthermore, the method for recalculating the weight coefficient of seepage pressure data and locally correcting the spatial variance of ground radar data when the correlation coefficient between the time series data of seepage rate and the time series data of rainfall intensity within a preset fan-shaped area in the direction of the tunnel axis is detected to be lower than the preset correlation coefficient threshold includes: Arrange a seepage rate sensor array within the preset fan-shaped area at a preset sampling interval and collect the time series data of seepage rate. Calculate the correlation coefficient within the current time window for the time series data of seepage rate and the time series data of rainfall intensity in the corresponding fan-shaped area in the initial calibrated rainfall model through a sliding time window.

[0014] When the correlation coefficient is lower than the preset correlation coefficient threshold, obtain the real-time Pearson correlation coefficient between the seepage pressure data and the surface runoff data within the current time window and perform dynamic correction through a correction factor to obtain the weight coefficient of the seepage pressure data. The correction factor is the moving average deviation of the seepage pressure data.

[0015] Synchronously locate the grid cells with abnormal spatial variance of rainfall intensity in the ground radar data within the fan-shaped area, calculate the residual distribution based on the measured seepage rate and the rainfall intensity of the corresponding grid cells, perform spatial interpolation on the residual distribution using a Gaussian kernel function to obtain a spatial variance correction coefficient matrix, and perform local weighted correction on the spatial variance of the ground radar data according to the spatial variance correction coefficient matrix.

[0016] Furthermore, the method for obtaining the seepage field and the surface runoff field of the tunnel and obtaining the non-linear mapping relationship between rainfall intensity and tunnel water inflow through alternating iterative calculation includes: Obtain the dynamic change data of the tunnel lining permeability coefficient and the geometric parameters of the surface runoff section; calculate the water pressure distribution behind the tunnel lining based on the dynamic change data of the tunnel lining permeability coefficient and the seepage pressure data, and dynamically correct the permeability coefficient according to the water pressure distribution behind the tunnel lining; obtain the surface runoff flow rate through data analysis using the corrected permeability coefficient and the geometric parameters of the surface runoff section; use the surface runoff flow rate as the boundary condition of the seepage field and recalculate the water pressure distribution of the seepage field; perform alternating iterative calculation on the seepage field and the surface runoff field until the difference in water pressure distribution between two iterations is less than the preset water pressure distribution difference threshold, and then generate a steady-state coupling relationship between the seepage field and the surface runoff field.

[0017] Obtain the real-time water level gradient data of the adjacent tunnel drainage system, and calculate the water convergence interference intensity coefficient according to the distance between adjacent tunnels and the change rate of the water level gradient; use the water convergence interference intensity coefficient as a dynamic weight factor to correct the surface runoff boundary conditions in the intersection area of the tunnel group in the surface runoff model; update the coupling relationship between the seepage field and the surface runoff field according to the corrected surface runoff boundary conditions to obtain the corrected coupling relationship.

[0018] Establish a non-linear mapping relationship between the rainfall intensity and the tunnel water convergence flow rate according to the steady-state coupling relationship and the corrected coupling relationship.

[0019] Further, the method for obtaining the surface runoff flow rate through data analysis of the corrected permeability coefficient and the geometric parameters of the surface runoff cross-section includes: Establish a surface runoff model of the surface runoff field according to the corrected permeability coefficient and the geometric parameters of the surface runoff cross-section; input the spatio-temporal distribution data of the rainfall intensity at the current time step into the surface runoff model, and perform spatial discretization to solve for the water depth and flow velocity distribution data of each grid cell, and calculate the surface runoff flow rate according to the water depth and flow velocity distribution data and the geometric parameters of the surface runoff cross-section.

[0020] Further, the method for obtaining the real-time water level gradient data of the adjacent tunnel drainage system and calculating the water convergence interference intensity coefficient according to the distance between adjacent tunnels and the change rate of the water level gradient includes: Through a pressure-type water level sensor array arranged in the connection channel of the adjacent tunnel drainage system, real-time collect the water level gradient data of each drainage well along the tunnel axis direction.

[0021] Calculate the geometric center distance between adjacent tunnel drainage systems according to the spatial coordinate data in the tunnel group design drawings, and convert the geometric center distance into the equivalent hydraulic conductivity distance in the hydrological model; synchronously extract the temporal variation characteristics of the water level gradient data within the current time window, and use the sliding difference method to calculate the change rate of the water level gradient between adjacent tunnel sumps per unit time; calculate the water convergence interference intensity coefficient after non-dimensional processing of the reciprocal of the equivalent hydraulic conductivity distance and the change rate of the water level gradient.

[0022] Further, the method for establishing a non-linear mapping relationship between the rainfall intensity and the tunnel water convergence flow rate according to the steady-state coupling relationship and the corrected coupling relationship includes: Extract the water pressure distribution characteristics of the seepage field under different permeability coefficients and the temporal sequence data of the surface runoff flow rate of the surface runoff field from the steady-state coupling relationship, and obtain the surface runoff boundary condition parameters corrected by the water convergence interference intensity coefficient in the intersection area of the tunnel group from the corrected coupling relationship.

[0023] Perform multi-dimensional feature fusion on the water pressure distribution characteristics, surface runoff flow time series data, and corrected surface runoff boundary condition parameters with the rainfall intensity spatio-temporal distribution data of the final rainfall model to generate an input feature vector including rainfall intensity spatio-temporal distribution, dynamic change amount of tunnel lining permeability coefficient, water level gradient between adjacent tunnels, and water collection interference intensity coefficient.

[0024] According to the correspondence between the measured data of tunnel water collection flow in historical rainfall events and the input feature vector, use a neural network to perform non-linear regression modeling, where the number of nodes in the input layer is the dimension of the input feature vector, and the number of nodes in the output layer is the predicted tunnel water collection flow; when the root mean square error between the predicted water collection flow and the measured data of water collection flow is less than the preset error threshold, a non-linear mapping relationship between rainfall intensity and tunnel water collection flow is generated.

[0025] Further, the method for outputting the frequency adjustment instruction and start-stop control sequence of the drainage pump group according to the predicted tunnel water collection flow and the control reference parameters of the drainage pump group includes: Calculate the flow difference value between the predicted tunnel water collection flow and the current actual drainage flow of the drainage pump group in real time, and generate the frequency adjustment direction and frequency adjustment amplitude according to the positive and negative signs and change rate of the flow difference value.

[0026] Generate a start-stop priority sequence according to the liquid level gradient of the sump and the current output ratio of the drainage pump group. When the rising rate of the liquid level gradient of the sump exceeds the preset safety threshold, the standby drainage pump group is preferentially started and the maximum frequency adjustment amplitude is allocated; the frequency adjustment instruction and the start-stop priority sequence are logically coupled and then output to the drainage pump group controller for execution.

[0027] On the other hand, based on the same inventive concept, the present invention also provides an intelligent scheduling system for a tunnel drainage system based on multi-source data fusion. The system includes: a final rainfall model acquisition module, a non-linear mapping relationship generation module, and a drainage scheduling management module, and the modules are communicatively connected in sequence; The final rainfall model acquisition module is used to acquire meteorological satellite data, ground-based radar data, and seepage pressure data in the tunnel, perform spatio-temporal alignment through a spatio-temporal nested data fusion method, and dynamically allocate weight coefficients to obtain a spatio-temporally calibrated final rainfall model.

[0028] The non-linear mapping relationship generation module is used to acquire the seepage field and surface runoff field of the tunnel, and obtain the non-linear mapping relationship between rainfall intensity and tunnel water collection flow through iterative calculation.

[0029] The drainage scheduling management module is used to obtain the predicted water inflow of the tunnel through the non - linear mapping relationship with the rainfall intensity spatio - temporal distribution data of the final rainfall model, acquire the sump liquid level data of the tunnel drainage system and the operation output data of the drainage pump set, and generate the control reference parameters for the drainage pump set; and output the frequency adjustment instruction and start - stop control sequence of the drainage pump set according to the predicted water inflow of the tunnel and the control reference parameters of the drainage pump set.

[0030] (3)Beneficial effects Compared with the prior art, the beneficial effects of the present invention are as follows: 1. By fusing multi - source data such as meteorological satellite data, ground - based radar data, and seepage pressure data in the tunnel, and using the spatio - temporal nested data fusion method for spatio - temporal alignment and dynamic weight allocation, a final rainfall model after spatio - temporal calibration is generated, significantly improving the spatio - temporal accuracy of rainfall prediction and effectively overcoming the problems of single data source and insufficient spatio - temporal calibration ability in traditional methods.

[0031] 2. According to the final rainfall model after spatio - temporal calibration, the non - linear mapping relationship between rainfall intensity and tunnel water inflow is obtained through iterative calculation, making the prediction of tunnel water inflow more accurate and reliable. It provides a key basis for the intelligent scheduling of the drainage pump set, and can dynamically adjust the operation parameters of the drainage pump set according to real - time prediction results, greatly enhancing the ability of the tunnel drainage system to cope with complex rainfall conditions.

[0032] 3. On the basis of considering the interference effect between adjacent tunnel drainage systems, combined with the control reference parameters of the drainage pump set, the frequency adjustment instruction and start - stop control sequence of the drainage pump set are output, realizing the coordinated and optimized scheduling between the tunnel drainage system and adjacent tunnel drainage systems. It not only improves the operation efficiency of a single tunnel drainage system, but also optimizes the drainage layout of the tunnel group as a whole, effectively avoiding the problem of poor drainage caused by the interference of adjacent tunnel drainage systems, and further improving the safety and reliability of the tunnel drainage system. Brief description of the drawings

[0033] Figure 1 It is a flow chart of the intelligent scheduling method for a tunnel drainage system based on multi - source data fusion in Embodiment 1 of the present invention.

[0034] Figure 2 It is a schematic diagram of the module composition of the intelligent scheduling system for a tunnel drainage system based on multi - source data fusion in Embodiment 2 of the present invention. Detailed implementation manners

[0035] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0036] Before giving examples, it is necessary to elaborate on the application scenario of the inventive concept. The present invention is an intelligent scheduling method and system for tunnel drainage systems based on multi-source data fusion, which is applied to solve the problems of traditional tunnel drainage system scheduling methods, such as single data source, insufficient spatio-temporal calibration ability, failure to consider the non-linear mapping relationship between rainfall and catchment water, and interference effects of adjacent tunnel drainage, resulting in low drainage prediction accuracy and unreasonable drainage scheduling.

[0037] Example 1: As Figure 1 shown, this embodiment provides an intelligent scheduling method for a tunnel drainage system based on multi-source data fusion, and the method includes: S1. Obtain meteorological satellite data, ground-based radar data, and seepage pressure data in the tunnel, and perform spatio-temporal alignment through a spatio-temporal nested data fusion method and dynamically allocate weight coefficients to obtain a finally spatio-temporally calibrated rainfall model; meteorological satellite data can provide large-scale and macroscopic meteorological information, ground-based radar data can obtain more accurate local meteorological characteristics, and seepage pressure data in the tunnel reflects the hydrological conditions inside the tunnel. Since there are differences in the accuracy and coverage of different data sources in terms of time and space, by comprehensively utilizing the advantages of each data source through spatio-temporal nested data fusion and eliminating spatio-temporal inconsistencies, the spatio-temporal distribution characteristics of rainfall can be more accurately reflected.

[0038] S2. Obtain the seepage field and surface runoff field of the tunnel, and obtain the non-linear mapping relationship between rainfall intensity and tunnel catchment water flow through iterative calculation; the seepage field describes the flow of groundwater around the tunnel, and the surface runoff field reflects the distribution and flow of surface water. Rainfall intensity is an important factor affecting the tunnel catchment water flow, and the relationship between the two is not a simple linear relationship. Through iterative calculation, the complex non-linear relationship between the two can be more accurately captured.

[0039] S3. Obtain the predicted catchment flow of the tunnel from the rainfall intensity spatio-temporal distribution data of the final rainfall model, acquire the sump liquid level data of the tunnel drainage system and the operating output data of the drainage pump set, and generate the control reference parameters for the drainage pump set; output the frequency adjustment command and start-stop control sequence for the drainage pump set according to the predicted catchment flow of the tunnel and the control reference parameters for the drainage pump set. The sump liquid level data reflects the water accumulation situation in the tunnel, and the operating output data of the drainage pump set reflects the working state of the drainage pump set. The control reference parameters for the drainage pump set comprehensively consider the actual drainage demand in the tunnel and the working ability of the drainage pump set.

[0040] The method for obtaining the finally spatio-temporally calibrated rainfall model through spatio-temporal alignment by the spatio-temporal nested data fusion method and dynamically allocating weight coefficients includes: Improve the spatial resolution of the meteorological satellite data to the same grid scale as the ground-based radar data through bilinear interpolation, and synchronize the timestamps of the ground-based radar data to match the sampling frequency of the seepage pressure data; construct a multi-dimensional feature matrix based on the spatio-temporally aligned meteorological satellite data, ground-based radar data, and seepage pressure data. Each grid node of the multi-dimensional feature matrix includes the rainfall intensity of the meteorological satellite data, the rainfall intensity of the ground-based radar data, seepage pressure, spatial coordinates, and timestamps. Meteorological satellite data usually has a relatively low spatial resolution, while ground-based radar data has a relatively high spatial resolution. By improving the spatial resolution of the meteorological satellite data to the same grid scale as the ground-based radar data through bilinear interpolation, the two are comparable in the spatial dimension. At the same time, synchronize the timestamps of the ground-based radar data to ensure the consistency of different data sources in the time dimension.

[0041] Dynamically calculate the weight coefficients of the meteorological satellite data, ground-based radar data, and seepage pressure data. The weight coefficient of the meteorological satellite data is determined according to the exponential decay characteristic of its historical error. The weight coefficient of the ground-based radar data is calculated based on the spatial variance of the rainfall intensity within its grid. The weight coefficient of the seepage pressure data is dynamically adjusted according to its real-time Pearson correlation coefficient with the surface runoff data, and the surface runoff data is obtained by inverting the geometric parameters of the surface runoff cross-section; the greater the historical error, the smaller the weight coefficient of the meteorological satellite data, and as time goes by, the influence of the error gradually decreases, and the weight coefficient will change accordingly, realizing the dynamic evaluation of the reliability of historical data. The reliability of the ground-based radar data in this area may be affected, so the weight coefficient will be adjusted accordingly. The surface runoff data is obtained by inverting the geometric parameters of the surface runoff cross-section. The higher the correlation between the seepage pressure data and the surface runoff data, the more accurately the seepage pressure data reflects the rainfall situation, and the greater its weight coefficient.

[0042] Perform weighted calculations on the rainfall intensity of meteorological satellite data, the rainfall intensity of ground-based radar data, and the seepage pressure of each grid node in the multi-dimensional feature matrix with the corresponding weight coefficients to generate an initial calibrated rainfall model; the weighted calculation comprehensively considers the contributions and reliabilities of different data sources, making the initial calibrated rainfall model more accurate.

[0043] Real-time detect the correlation coefficient between the time series data of seepage rate and the time series data of rainfall intensity in a preset fan-shaped area along the tunnel axis direction. When the detected correlation coefficient is lower than the preset correlation coefficient threshold, recalculate the weight coefficient of the seepage pressure data and perform local correction on the spatial variance of the ground-based radar data; perform iterative optimization on the rainfall intensity of the grid nodes in the tunnel group intersection area in the initial calibrated rainfall model according to the corrected weight coefficient until the difference between the predicted rainfall intensity and the measured rainfall intensity is less than the preset rainfall intensity difference threshold, and then output the finally rainfall model calibrated in space and time. When the detected correlation coefficient is lower than the preset correlation coefficient threshold, it indicates that the relationship between the current seepage pressure data and the rainfall intensity may have changed, and it is necessary to recalculate the weight coefficient of the seepage pressure data and perform local correction on the spatial variance of the ground-based radar data. By continuously adjusting the weight coefficient and performing iterative calculations, the difference between the predicted rainfall intensity and the measured rainfall intensity is made less than the preset rainfall intensity difference threshold.

[0044] The method for real-time detecting the correlation coefficient between the time series data of seepage rate and the time series data of rainfall intensity in a preset fan-shaped area along the tunnel axis direction and, when the detected correlation coefficient is lower than the preset correlation coefficient threshold, recalculating the weight coefficient of the seepage pressure data and performing local correction on the spatial variance of the ground-based radar data includes: An array of seepage rate sensors is arranged in the preset fan-shaped area at a preset sampling interval, and time series data of the seepage rate is collected. The time series data of the seepage rate and the time series data of the rainfall intensity in the corresponding fan-shaped area in the initial calibration rainfall model are used to calculate the correlation coefficient within the current time window through a sliding time window. The seepage rate sensors can measure the seepage rates at different positions in the area in real time and accurately. The size of the sampling interval depends on the required monitoring accuracy and the area range. A smaller sampling interval can obtain more detailed seepage rate distribution information, but it will also increase the cost and the complexity of data processing. The sensor array continuously collects seepage rate data to form time series data of the seepage rate, reflecting the change of the seepage rate in the area over time. The sliding time window is a commonly used time series analysis method, which divides the entire time series data into multiple consecutive time windows with a certain length. In each time window, statistical analysis is performed on the time series data of the seepage rate and the time series data of the rainfall intensity to calculate the correlation coefficient. The correlation coefficient is used to measure the linear correlation degree between the two time series data. By calculating the correlation coefficient within the current time window through the sliding time window, the relationship change between the seepage rate and the rainfall intensity can be monitored in real time. If the correlation coefficient is high, it indicates that there is a strong linear correlation between the two, that is, the change of the rainfall intensity will cause a corresponding change in the seepage rate. If the correlation coefficient is low, it indicates that the linear correlation between the two is weak.

[0045] When the correlation coefficient is lower than the preset correlation coefficient threshold, the real-time Pearson correlation coefficient of the seepage pressure data and the surface runoff data within the current time window is obtained and dynamically corrected through a correction factor to obtain the weight coefficient of the seepage pressure data. The correction factor is the sliding average deviation of the seepage pressure data. The Pearson correlation coefficient is a statistic used to measure the linear correlation degree between two variables. Within the current time window, the seepage pressure data and the surface runoff data are respectively collected and the correlation coefficient is calculated using the calculation formula of the Pearson correlation coefficient. The correction factor is the sliding average deviation of the seepage pressure data. The sliding average deviation refers to the average of the deviations between each data point and the average value of the data within a series of consecutive data points. By calculating the sliding average deviation of the seepage pressure data, the fluctuation of the seepage pressure data is reflected, thus providing a reference for dynamically correcting the weight coefficient of the seepage pressure data. The specific method of correction can be designed according to the actual situation. For example, the weight of the correction factor can be adjusted according to the size of the correlation coefficient, and then the correction factor is multiplied by a coefficient and added or subtracted from the original weight coefficient to obtain a new weight coefficient.

[0046] The grid cells with abnormal spatial variance of rainfall intensity in the ground-based radar data in the sector area are synchronously located, and the residual distribution is calculated according to the measured seepage rate and the rainfall intensity of the corresponding grid cell, and the spatial variance correction coefficient matrix is obtained by spatial interpolation of the residual distribution using the Gaussian kernel function, and the spatial variance of the ground-based radar data is locally weighted corrected according to the spatial variance correction coefficient matrix. The ground-based radar data in the sector area is divided into a number of grid cells according to a certain spatial resolution, and each grid cell corresponds to a specific spatial position. This gridding process helps to perform spatial analysis and processing on the data, and calculate the spatial variance of rainfall intensity in each grid cell. The spatial variance reflects the discrete degree of rainfall intensity in the grid cell, and the larger the variance, the more drastic the change of rainfall intensity in the grid cell. According to the position information of the sensor, the grid cell corresponding to each measured seepage rate data point is determined, and the rainfall intensity data of the grid cell is obtained. For each measured seepage rate data point, the residual between it and the rainfall intensity of the corresponding grid cell is calculated. In this embodiment, the linear or nonlinear relationship between rainfall intensity and seepage rate is obtained by statistical analysis of historical data, and the residual reflects the deviation between the two. All residual data are statistically analyzed according to spatial positions to obtain the spatial distribution of residuals. The Gaussian kernel function is a commonly used spatial interpolation function with good smoothness and locality. The Gaussian kernel function is used to perform spatial interpolation on the residual distribution to obtain the spatial variance correction coefficient corresponding to each grid cell in the fan-shaped area. The spatial variance correction coefficient reflects the degree of correction of the spatial variance of rainfall intensity in the grid cell. The larger the spatial variance correction coefficient, the greater the degree to which the spatial variance of rainfall intensity in the grid cell needs to be corrected. The spatial variance correction coefficients of all grid cells are arranged in the order of the grid to construct a spatial variance correction coefficient matrix. For each grid cell, its original spatial variance is multiplied by the corresponding spatial variance correction coefficient to obtain the corrected spatial variance. The spatial variance of the ground-based radar data is locally weighted corrected, so that the grid cells with abnormal spatial variance of rainfall intensity are properly adjusted, thereby improving the accuracy and reliability of the ground-based radar data.

[0047] The method of obtaining the seepage field and the surface runoff field of the tunnel and obtaining the nonlinear mapping relationship between rainfall intensity and tunnel water flow through alternating iterative calculation includes: Obtain the dynamic change data of the permeability coefficient of the tunnel lining and the geometric parameters of the surface runoff cross-section; calculate the water pressure distribution behind the tunnel lining based on the dynamic change data of the permeability coefficient of the tunnel lining and the seepage pressure data, and dynamically correct the permeability coefficient according to the water pressure distribution behind the tunnel lining; obtain the surface runoff flow rate through data analysis using the corrected permeability coefficient and the geometric parameters of the surface runoff cross-section; use the surface runoff flow rate as the boundary condition of the seepage field and recalculate the water pressure distribution of the seepage field; perform iterative calculations on the seepage field and the surface runoff field alternately until the difference in water pressure distribution between two iterations is less than the preset water pressure distribution difference threshold, then generate the steady-state coupling relationship between the seepage field and the surface runoff field; the permeability coefficient of the tunnel lining is affected by various factors, such as the deterioration of the lining material and the erosion of groundwater, which will cause the permeability coefficient to change over time. Obtaining these dynamic change data can more accurately reflect the actual permeability performance of the tunnel lining. The geometric parameters of the surface runoff cross-section include the shape, size, etc. of the cross-section, which are crucial for calculating the surface runoff flow rate. For example, the cross-section shape is rectangular, trapezoidal or circular, etc. Different cross-section shapes and sizes will cause changes in the flow velocity and flow rate of the water passing through the cross-section. According to the dynamic change data of the permeability coefficient of the tunnel lining and the seepage pressure data, use the basic principles of seepage mechanics, such as Darcy's law, etc., to calculate the water pressure distribution behind the tunnel lining. The seepage pressure data is obtained by setting pressure sensors behind the tunnel lining. Since the water pressure distribution will affect the permeability performance of the tunnel lining, the permeability coefficient of the tunnel lining is dynamically corrected according to the calculated water pressure distribution. For example, when the water pressure increases, it will cause changes in the porosity of the lining material, thus affecting the permeability coefficient. Input the corrected permeability coefficient and the geometric parameters of the surface runoff cross-section into a suitable hydraulic model, such as Manning's formula, etc., and calculate the surface runoff flow rate through data analysis. The surface runoff flow rate is an important parameter of the surface runoff field, reflecting the strength of the surface water flow. Use the calculated surface runoff flow rate as the boundary condition of the seepage field and recalculate the water pressure distribution of the seepage field. The surface runoff flow rate will affect the seepage field. For example, the surface runoff may enter the rock mass around the tunnel through infiltration, etc., changing the water pressure distribution of the seepage field. Perform iterative calculations alternately on the recalculated seepage field water pressure distribution and the surface runoff field. In each iteration, update the boundary conditions of the surface runoff field, such as the infiltration amount, etc., according to the water pressure distribution of the seepage field, and at the same time update the boundary conditions of the seepage field, such as the surface runoff recharge amount, etc., according to the changes in the surface runoff field. Calculate the difference in water pressure distribution between two iterations. When the difference is less than the preset water pressure distribution difference threshold, the iterative process has converged, and at this time, generate the steady-state coupling relationship between the seepage field and the surface runoff field, reflecting the interaction and balance state between the seepage field and the surface runoff field.

[0048] Obtain the real-time water level gradient data of the adjacent tunnel drainage system, and calculate the water convergence interference intensity coefficient according to the adjacent tunnel spacing and the change rate of the water level gradient; use the water convergence interference intensity coefficient as a dynamic weight factor to correct the surface runoff boundary conditions in the intersection area of the tunnel group in the surface runoff model; update the coupling relationship between the seepage field and the surface runoff field according to the corrected surface runoff boundary conditions to obtain the corrected coupling relationship; install water level sensors in the drainage systems of adjacent tunnels, and these sensors can measure the water level heights at different positions in the tunnels in real time and accurately. By regularly collecting these water level data, the water level distribution of the adjacent tunnel drainage systems at each moment can be obtained. The water level gradient refers to the change rate of the water level in space, that is, the change amount of the water level per unit distance. For adjacent tunnels, the water level gradients at different positions between the adjacent tunnels can be calculated according to the water level data at different positions. For example, at key positions such as the connection and intersection of adjacent tunnels, calculating the water level gradient can reflect the flow trend and intensity of the water flow at these positions. Measuring the actual distance between adjacent tunnels is one of the important parameters for calculating the water convergence interference intensity coefficient. The size of the adjacent tunnel spacing will affect the flow and interaction of the water flow between the tunnels. In addition to the water level gradient itself, it is also necessary to analyze the change rate of the water level gradient over time. The change rate of the water level gradient can reflect the dynamic change situation of the water flow between adjacent tunnels, such as whether the water flow is accelerating or decelerating. The water convergence interference intensity coefficient comprehensively reflects the degree of water flow interference between adjacent tunnels. The larger the water convergence interference intensity coefficient, the greater the water convergence interference intensity. The surface runoff model is usually used to describe the movement and distribution of surface water flow. The surface runoff boundary conditions in the intersection area of the tunnel group are crucial for the accuracy of the model. Correct the surface runoff boundary conditions in the intersection area of the tunnel group according to the size of the water convergence interference intensity coefficient. For example, when the water convergence interference intensity coefficient is large, it indicates that the water flow interference between adjacent tunnels is strong. At this time, it is necessary to adjust the surface runoff boundary conditions to more accurately reflect the water flow situation in this area. There is a close interaction relationship between the seepage field and the surface runoff field. Surface runoff will enter the ground through infiltration and other means, affecting the water pressure distribution and flow situation of the seepage field; and the change of the seepage field will in turn affect the generation and flow of surface runoff. Input the corrected surface runoff boundary conditions into the coupling model of the seepage field and the surface runoff field, and recalculate the water pressure distribution, flow rate and other parameters of the seepage field and the surface runoff field. The corrected coupling relationship between the seepage field and the surface runoff field more accurately reflects the actual situation of the water flow interference between adjacent tunnels.

[0049] According to the steady-state coupling relationship and the corrected coupling relationship, establish a non-linear mapping relationship between the rainfall intensity and the tunnel water convergence flow rate.

[0050] The obtaining the surface runoff flow rate by data analysis of the corrected permeability coefficient and the geometric parameters of the surface runoff cross-section includes: A surface runoff model of the surface runoff field is established based on the corrected permeability coefficient and the geometric parameters of the surface runoff cross-section. After inputting the rainfall intensity spatio-temporal distribution data of the current time step into the surface runoff model, spatial discretization is performed for solution to obtain the water depth and flow velocity distribution data of each grid cell. Based on the water depth and flow velocity distribution data and the geometric parameters of the surface runoff cross-section, the surface runoff flow rate is calculated. The surface runoff model is based on hydraulic principles such as the continuity equation and the momentum equation, and is used to describe the generation, flow, and distribution of surface runoff. Surface runoff is a dynamic process, and the spatio-temporal distribution of rainfall intensity changes over time. Therefore, it is necessary to obtain the rainfall intensity data of the current time step to accurately simulate the situation of surface runoff. The rainfall intensity not only changes over time but also varies spatially. For example, the rainfall intensity is relatively large in some areas and relatively small in other areas. Inputting the spatio-temporal distribution data of rainfall intensity into the surface runoff model can more realistically reflect the generation of surface runoff. To solve the surface runoff model, the study area needs to be divided into multiple grid cells through spatial discretization. Each grid cell represents a small spatial area, and by calculating each grid cell, the surface runoff situation of the entire study area can be obtained. After inputting the rainfall intensity spatio-temporal distribution data into the surface runoff model, numerical methods are used for spatial discretization and solution. During the solution process, various factors such as terrain, soil properties, and vegetation cover are considered to obtain the water depth and flow velocity distribution data of each grid cell. The water depth reflects the degree of surface water accumulation, and the flow velocity reflects the flow speed of the water. Based on the water depth and flow velocity distribution data of each grid cell and combined with the geometric parameters of the surface runoff cross-section, the surface runoff flow rate of each grid cell can be calculated. For example, for a rectangular cross-section, the flow rate can be calculated based on the water depth and the cross-section width; for a trapezoidal cross-section, factors such as the slope coefficient of the cross-section also need to be considered. Summing up the surface runoff flow rates of all grid cells can obtain the surface runoff flow rate of the entire study area, which is of great significance for evaluating flood risks, designing drainage systems, etc.

[0051] The method for obtaining the real-time water level gradient data of the adjacent tunnel drainage system and calculating the water collection interference intensity coefficient according to the adjacent tunnel spacing and the water level gradient change rate includes: Through a pressure-type water level sensor array arranged in the connection channel of the adjacent tunnel drainage system, the water level gradient data of each drainage well along the tunnel axis direction is collected in real time. Reasonably arranging a pressure-type water level sensor array in the connection channel of the adjacent tunnel drainage system can accurately measure the water level pressure at the drainage well, and then based on the corresponding relationship between pressure and water level, the water level data of each drainage well along the tunnel axis direction can be obtained. By calculating the difference between the water level data of adjacent drainage wells, the water level gradient between each drainage well can be obtained.

[0052] Calculate the geometric center distance between adjacent tunnel drainage systems based on the spatial coordinate data in the tunnel group design drawings, and convert the geometric center distance into the equivalent hydraulic conductivity distance in the hydrological model; synchronously extract the temporal variation characteristics of the water level gradient data within the current time window, and use the sliding difference method to calculate the water level gradient change rate between adjacent tunnel sumps per unit time; perform dimensionless processing on the reciprocal of the equivalent hydraulic conductivity distance and the water level gradient change rate, and then calculate the water convergence interference intensity coefficient. Obtain the water convergence interference intensity coefficient through the water convergence interference intensity calculation formula , and the interference intensity calculation formula is: ; where is the water level gradient change rate, is the equivalent hydraulic conductivity distance; is an empirical coefficient, which is calibrated based on the measured data of the mutual interference of adjacent tunnel drainage systems in historical rainfall events. Determine the spatial positions of adjacent tunnel drainage systems based on the spatial coordinate data in the tunnel group design drawings. By calculating the geometric center coordinates between adjacent tunnel drainage systems, the geometric center distance between them can be obtained. Convert the geometric center distance into the equivalent hydraulic conductivity distance in the hydrological model. In the actual hydrological process, the conduction of water flow between tunnel drainage systems is not a simple linear conduction, and is affected by various factors, such as the permeability of the soil, the structure of the drainage system, etc. The equivalent hydraulic conductivity distance takes these factors into account and can more accurately reflect the conduction characteristics of water flow between adjacent tunnel drainage systems. Synchronously extract the temporal variation characteristics of the water level gradient data within the current time window. The time window can be set according to actual needs. By analyzing the water level gradient data within the time window, understand the change trend of the water level gradient over time. Use the sliding difference method to calculate the water level gradient change rate between adjacent tunnel sumps per unit time. The sliding difference method is a commonly used method for temporal data analysis. By calculating the ratio of the difference in water level gradients at adjacent time points to the time interval, the water level gradient change rate is obtained. For example, if the water level gradients between adjacent tunnel sumps at times t1 and t2 (t2 > t1) are h1 and h2 respectively, and the time interval is =t2 - t1, then the water level gradient change rate . Dimensionless processing is to eliminate the dimensional influence between different physical quantities, so that comparison and calculation can be carried out on the same scale. Usually, methods such as standardization and normalization can be used for dimensionless processing

[0053] The method for establishing the non-linear mapping relationship between rainfall intensity and tunnel water convergence flow according to the steady-state coupling relationship and the modified coupling relationship includes: Extract the water pressure distribution characteristics of the seepage field under different permeability coefficients and the time-series data of the surface runoff flow rate of the surface runoff field from the steady-state coupling relationship, and obtain the surface runoff boundary condition parameters corrected by the water collection interference intensity coefficient in the intersection area of the tunnel group from the modified coupling relationship; by analyzing the steady-state coupling relationship, the water pressure values and their distribution laws of each point in the seepage field under different permeability coefficients can be extracted. For example, the larger the permeability coefficient, the easier the flow of groundwater, and the relatively smaller the water pressure gradient; on the contrary, the smaller the permeability coefficient, the larger the water pressure gradient may be. In the steady-state coupling relationship, the runoff flow rate data of the surface runoff field at different time points can be obtained to form the time-series data of the surface runoff flow rate, which reflects the change of the surface runoff over time and is affected by various factors such as precipitation intensity, topography, and soil type. There is a mutual recharge relationship between the surface runoff and groundwater. The change of the surface runoff will affect the recharge amount of groundwater, thus affecting the water collection flow rate of the tunnel. The modified coupling relationship is obtained by modifying the steady-state coupling relationship after considering more actual factors (such as the water collection interference intensity coefficient). The surface runoff boundary condition parameters are important parameters describing the boundary characteristics of the surface runoff field, such as boundary water level, flow rate, etc. In the modified coupling relationship, due to considering the influence of the water collection interference intensity coefficient, the surface runoff boundary condition parameters will change accordingly. The corrected surface runoff boundary condition parameters consider the water collection interference intensity coefficient between the drainage systems of adjacent tunnels and more accurately describe the characteristics of the surface runoff field in the intersection area of the tunnel group.

[0054] Perform multi-dimensional feature fusion on the water pressure distribution characteristics, the time-series data of the surface runoff flow rate, and the corrected surface runoff boundary condition parameters with the rainfall intensity spatio-temporal distribution data of the final rainfall model to generate an input feature vector including the rainfall intensity spatio-temporal distribution, the dynamic change amount of the tunnel lining permeability coefficient, the water level gradient between adjacent tunnels, and the water collection interference intensity coefficient; perform multi-dimensional feature fusion on the above various feature data to generate an input feature vector containing rich information, which can comprehensively reflect various factors affecting the water collection flow rate of the tunnel.

[0055] According to the correspondence between the measured data of the tunnel water inflow in historical rainfall events and the input feature vectors, a neural network is used for non - linear regression modeling. The number of nodes in the input layer is the dimension of the input feature vectors, and the number of nodes in the output layer is the predicted tunnel water inflow. When the root - mean - square error between the predicted water inflow and the measured data of the water inflow is less than the preset error threshold, a non - linear mapping relationship between the rainfall intensity and the tunnel water inflow is generated. The number of nodes in the input layer is equal to the dimension of the input feature vectors. Each node corresponds to a feature in the input feature vectors, and multi - dimensional feature data can be input into the neural network to let the neural network learn the relationship between these features and the tunnel water inflow. The number of nodes in the output layer is the predicted tunnel water inflow, and the goal of the neural network is to predict the tunnel water inflow based on the input feature vectors. The root - mean - square error is an index to measure the difference between the predicted water inflow and the measured data of the water inflow, reflecting the accuracy of the prediction model. When the root - mean - square error between the predicted water inflow and the measured data of the water inflow is less than the preset error threshold, it indicates that the neural network model has learned the non - linear relationship between factors such as rainfall intensity, tunnel lining permeability coefficient, adjacent tunnel water level gradient, and water interference intensity coefficient and the tunnel water inflow. At this time, the neural network model can be used as a non - linear mapping relationship to predict the tunnel water inflow under different rainfall intensities and different tunnel conditions, and can more accurately reflect the actual hydro - geological conditions and the complexity of the tunnel drainage system.

[0056] The method for outputting the frequency adjustment instruction and start - stop control sequence of the drainage pump group according to the predicted tunnel water inflow and the control reference parameters of the drainage pump group includes: Real - time calculate the flow difference value between the predicted tunnel water inflow and the current actual drainage flow of the drainage pump group, and generate the frequency adjustment direction and frequency adjustment amplitude according to the positive - negative sign and change rate of the flow difference value. The frequency adjustment amplitude is mapped to the target frequency increment or decrement of the drainage pump group through a fuzzy control algorithm, and the target frequency adjustment amount is constrained by the liquid level safety threshold in the control reference parameters of the drainage pump group.

[0057] Generate a start - stop priority sequence according to the liquid level gradient of the sump well and the current output ratio of the drainage pump group. When the rising rate of the liquid level gradient of the sump well exceeds the preset safety threshold, the standby drainage pump group is preferentially started and the maximum frequency adjustment amplitude is allocated. Couple the frequency adjustment instruction and the start - stop priority sequence logically, and after eliminating the instantaneous fluctuation interference through a preset delay filtering module, output it to the drainage pump group controller for execution. After execution, re - detect the flow difference value between the predicted and actual drainage flows. If the flow difference value does not return to the threshold range, dynamically optimize the membership function parameters of the fuzzy control algorithm according to the previous adjustment effect and iterate the adjustment process.

[0058] Embodiment 2: Based on the same inventive concept, as Figure 2As shown in the figure, this embodiment also provides an intelligent scheduling system for a tunnel drainage system based on multi-source data fusion. The system includes: a final rainfall model acquisition module, a non-linear mapping relationship generation module, and a drainage scheduling management module. The modules are communicatively connected in sequence; The final rainfall model acquisition module is used to obtain meteorological satellite data, ground-based radar data, and seepage pressure data in the tunnel, perform spatio-temporal alignment through a spatio-temporal nested data fusion method, and dynamically allocate weight coefficients to obtain a spatio-temporally calibrated final rainfall model.

[0059] The non-linear mapping relationship generation module is used to obtain the seepage field and surface runoff field of the tunnel, and obtain the non-linear mapping relationship between rainfall intensity and tunnel water collection flow through iterative calculation.

[0060] The drainage scheduling management module is used to obtain the predicted tunnel water collection flow by using the spatio-temporal distribution data of rainfall intensity of the final rainfall model through the non-linear mapping relationship, obtain the sump liquid level data of the tunnel drainage system and the operating output data of the drainage pump group, and generate control reference parameters for the drainage pump group; output frequency adjustment instructions and start-stop control sequences for the drainage pump group according to the predicted tunnel water collection flow and the control reference parameters of the drainage pump group.

[0061] It should be noted that regarding the system in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment related to the method, and will not be elaborated here.

[0062] Finally, it should be noted that although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An intelligent scheduling method for tunnel drainage systems based on multi-source data fusion, characterized in that The method includes: Obtaining meteorological satellite data, ground-based radar data, and seepage pressure data inside the tunnel, and performing spatio-temporal alignment through a spatio-temporal nested data fusion method and dynamically assigning weight coefficients to obtain a finally rainfall model calibrated in space and time; Obtaining the seepage field and surface runoff field of the tunnel, and obtaining the non-linear mapping relationship between rainfall intensity and tunnel catchment flow through iterative calculation; Obtaining the predicted tunnel catchment flow by using the spatio-temporal distribution data of rainfall intensity of the finally rainfall model through the non-linear mapping relationship, obtaining the sump liquid level data of the tunnel drainage system and the operating output data of the drainage pump group, and generating control reference parameters for the drainage pump group; Outputting frequency adjustment instructions and start-stop control sequences for the drainage pump group according to the predicted tunnel catchment flow and the control reference parameters of the drainage pump group.

2. The intelligent scheduling method for the tunnel drainage system based on multi-source data fusion according to claim 1, wherein The method of obtaining the finally rainfall model calibrated in space and time by performing spatio-temporal alignment through a spatio-temporal nested data fusion method and dynamically assigning weight coefficients includes: Improving the spatial resolution of meteorological satellite data to the same grid scale as ground-based radar data through bilinear interpolation, and synchronizing the timestamps of ground-based radar data to match the sampling frequency of seepage pressure data; Constructing a multi-dimensional feature matrix based on the spatio-temporally aligned meteorological satellite data, ground-based radar data, and seepage pressure data, where each grid node of the multi-dimensional feature matrix includes the rainfall intensity of meteorological satellite data, the rainfall intensity of ground-based radar data, seepage pressure, spatial coordinates, and timestamps; Dynamically calculating the weight coefficients of meteorological satellite data, ground-based radar data, and seepage pressure data, where the weight coefficient of meteorological satellite data is determined according to the exponential decay characteristic of its historical error, the weight coefficient of ground-based radar data is calculated according to the spatial variance of rainfall intensity within its grid, and the weight coefficient of seepage pressure data is dynamically adjusted according to its real-time Pearson correlation coefficient with surface runoff data, and the surface runoff data is obtained by inverting the geometric parameters of the surface runoff section; Performing weighted calculation on the rainfall intensity of meteorological satellite data, the rainfall intensity of ground-based radar data, seepage pressure, and the corresponding weight coefficients of each grid node in the multi-dimensional feature matrix to generate an initial calibrated rainfall model; Real-time detecting the correlation coefficient between the time series data of seepage rate and the time series data of rainfall intensity in a preset fan-shaped area along the tunnel axis direction. When the detected correlation coefficient is lower than the preset correlation coefficient threshold, recalculate the weight coefficient of seepage pressure data and perform local correction on the spatial variance of ground-based radar data; Iteratively optimize the rainfall intensity of grid nodes in the tunnel group intersection area in the initial calibrated rainfall model according to the corrected weight coefficient until the difference between the predicted rainfall intensity and the measured rainfall intensity is less than the preset rainfall intensity difference threshold, and then output the finally rainfall model calibrated in space and time.

3. The intelligent scheduling method for the tunnel drainage system based on multi-source data fusion according to claim 2, wherein The method of real-time detecting the correlation coefficient between the time series data of seepage rate and the time series data of rainfall intensity in a preset fan-shaped area along the tunnel axis direction. When the detected correlation coefficient is lower than the preset correlation coefficient threshold, recalculate the weight coefficient of seepage pressure data and perform local correction on the spatial variance of ground-based radar data includes: Arrange a seepage rate sensor array within the preset fan-shaped area at a preset sampling interval and collect seepage rate time series data. Calculate the correlation coefficient within the current time window for the seepage rate time series data and the rainfall intensity time series data of the corresponding fan-shaped area in the initial calibrated rainfall model through a sliding time window. When the correlation coefficient is lower than the preset correlation coefficient threshold, obtain the real-time Pearson correlation coefficient between the seepage pressure data and the surface runoff data within the current time window and dynamically correct it through a correction factor to obtain the weight coefficient of the seepage pressure data. The correction factor is the sliding average deviation of the seepage pressure data. Synchronously locate the grid cells with abnormal spatial variance of rainfall intensity in the ground radar data within the fan-shaped area. Calculate the residual distribution based on the measured seepage rate and the rainfall intensity of the corresponding grid cells, and perform spatial interpolation on the residual distribution using a Gaussian kernel function to obtain a spatial variance correction coefficient matrix. Locally weighted correct the spatial variance of the ground radar data according to the spatial variance correction coefficient matrix.

4. The intelligent scheduling method for the tunnel drainage system based on multi-source data fusion according to claim 1, wherein, The method for obtaining the seepage field and surface runoff field of the tunnel and obtaining the non-linear mapping relationship between rainfall intensity and tunnel catchment flow through alternating iterative calculations includes: Obtain the dynamic change data of the tunnel lining permeability coefficient and the geometric parameters of the surface runoff section. Calculate the water pressure distribution behind the tunnel lining based on the dynamic change data of the tunnel lining permeability coefficient and the seepage pressure data, and dynamically correct the permeability coefficient according to the water pressure distribution behind the tunnel lining. Obtain the surface runoff flow through data analysis using the corrected permeability coefficient and the geometric parameters of the surface runoff section. Use the surface runoff flow as the boundary condition of the seepage field and recalculate the water pressure distribution of the seepage field. Alternately iterate the seepage field and the surface runoff field until the difference in water pressure distribution between two iterations is less than the preset water pressure distribution difference threshold, and then generate a steady-state coupling relationship between the seepage field and the surface runoff field. Obtain the real-time water level gradient data of the drainage system of adjacent tunnels, and calculate the catchment interference intensity coefficient according to the distance between adjacent tunnels and the change rate of the water level gradient. Use the catchment interference intensity coefficient as a dynamic weight factor to correct the surface runoff boundary condition in the intersection area of the tunnel group in the surface runoff model. Update the coupling relationship between the seepage field and the surface runoff field according to the corrected surface runoff boundary condition to obtain a corrected coupling relationship. Establish a non-linear mapping relationship between rainfall intensity and tunnel catchment flow based on the steady-state coupling relationship and the corrected coupling relationship.

5. The intelligent scheduling method for the tunnel drainage system based on multi-source data fusion according to claim 4, wherein The method for obtaining the surface runoff flow through data analysis using the corrected permeability coefficient and the geometric parameters of the surface runoff section includes: Establish a surface runoff model of the surface runoff field based on the corrected permeability coefficient and the geometric parameters of the surface runoff section. Input the rainfall intensity spatio-temporal distribution data of the current time step into the surface runoff model and perform spatial discretization to solve for the water depth and flow velocity distribution data of each grid cell. Calculate the surface runoff flow according to the water depth and flow velocity distribution data and the geometric parameters of the surface runoff section.

6. The intelligent scheduling method for tunnel drainage systems based on multi-source data fusion according to claim 4, characterized in that The method for obtaining real-time water level gradient data of adjacent tunnel drainage systems and calculating the water collection interference intensity coefficient according to the distance between adjacent tunnels and the change rate of water level gradient includes: Real-time collecting water level gradient data of each drainage well along the tunnel axis direction through a piezometric water level sensor array arranged in the connection channel of the adjacent tunnel drainage systems; Calculating the geometric center distance between adjacent tunnel drainage systems according to the spatial coordinate data in the tunnel group design drawings, and converting the geometric center distance into the equivalent hydraulic conductivity distance in the hydrological model; synchronously extracting the temporal variation characteristics of the water level gradient data within the current time window, and calculating the change rate of the water level gradient between adjacent tunnel sump wells per unit time by using the sliding difference method; calculating the water collection interference intensity coefficient after non-dimensionalizing the reciprocal of the equivalent hydraulic conductivity distance and the change rate of the water level gradient.

7. The intelligent scheduling method for tunnel drainage systems based on multi-source data fusion according to claim 4, characterized in that, The method for establishing a non-linear mapping relationship between rainfall intensity and tunnel water collection flow according to the steady-state coupling relationship and the modified coupling relationship includes: Extracting the water pressure distribution characteristics of the seepage field under different permeability coefficients and the time series data of the surface runoff flow of the surface runoff field from the steady-state coupling relationship, and obtaining the modified surface runoff boundary condition parameters of the intersection area of the tunnel group affected by the water collection interference intensity coefficient from the modified coupling relationship; Performing multi-dimensional feature fusion on the water pressure distribution characteristics, the time series data of the surface runoff flow, and the modified surface runoff boundary condition parameters with the rainfall intensity spatio-temporal distribution data of the final rainfall model to generate an input feature vector including the rainfall intensity spatio-temporal distribution, the dynamic change amount of the tunnel lining permeability coefficient, the water level gradient between adjacent tunnels, and the water collection interference intensity coefficient; According to the corresponding relationship between the measured data of the tunnel water collection flow in historical rainfall events and the input feature vector, performing non-linear regression modeling by using a neural network, where the number of nodes in the input layer is the dimension of the input feature vector, and the number of nodes in the output layer is the predicted tunnel water collection flow; when the root mean square error between the predicted water collection flow and the measured data of the water collection flow is less than the preset error threshold, a non-linear mapping relationship between rainfall intensity and tunnel water collection flow is generated.

8. The intelligent scheduling method for the tunnel drainage system based on multi-source data fusion according to claim 1, characterized in that, The method for outputting the frequency adjustment instruction and the start-stop control sequence of the drainage pump group according to the predicted tunnel water collection flow and the control reference parameters of the drainage pump group includes: Real-time calculating the flow difference value between the predicted tunnel water collection flow and the current actual drainage flow of the drainage pump group, and generating the frequency adjustment direction and the frequency adjustment amplitude according to the positive and negative signs and the change rate of the flow difference value; Generating a start-stop priority sequence according to the liquid level gradient of the sump well and the current output ratio of the drainage pump group. When the rising rate of the liquid level gradient of the sump well exceeds the preset safety threshold, preferentially starting the standby drainage pump group and allocating the maximum frequency adjustment amplitude; logically coupling the frequency adjustment instruction and the start-stop priority sequence and then outputting them to the drainage pump group controller for execution.

9. An intelligent scheduling system for a tunnel drainage system based on multi-source data fusion, characterized in that, The system includes: a final rainfall model acquisition module, a non-linear mapping relationship generation module, and a drainage scheduling management module, and the modules are communicatively connected in sequence; The final rainfall model acquisition module is used to obtain meteorological satellite data, ground-based radar data, and seepage pressure data in the tunnel, perform spatio-temporal alignment through a spatio-temporal nested data fusion method, and dynamically allocate weight coefficients to obtain the final rainfall model after spatio-temporal calibration; The non-linear mapping relationship generation module is used to obtain the seepage field and surface runoff field of the tunnel, and obtain the non-linear mapping relationship between rainfall intensity and tunnel water inflow through iterative calculation; The drainage scheduling management module is used to obtain the predicted tunnel water inflow through the non-linear mapping relationship with the spatio-temporal distribution data of rainfall intensity of the final rainfall model, obtain the sump liquid level data of the tunnel drainage system and the operating output data of the drainage pump group, and generate the control reference parameters for the drainage pump group; output the frequency adjustment instruction and start-stop control sequence of the drainage pump group according to the predicted tunnel water inflow and the control reference parameters of the drainage pump group.

Images