Coal mining subsidence area plant foundation monitoring method and system based on multi-source data fusion

Abstract

The invention relates to a coal mining subsidence area plant foundation monitoring method and system based on multi-source data fusion, and the method comprises the steps: obtaining multi-source monitoring data of a target coal mining subsidence area plant area, the multi-source monitoring data comprising satellite remote sensing data, ground measurement data and interference radar data; performing time synchronization and space unification on the multi-source monitoring data to obtain monitoring data with unified space; performing fusion processing on the monitoring data with uniform space to obtain comprehensive deformation data; constructing a three-dimensional settlement deformation field according to the comprehensive deformation data, and extracting deformation rate and acceleration information of each monitoring point through a gradient calculation method; and according to the deformation rate and the acceleration information, obtaining risk assessment of the plant foundation in the target coal mining subsidence area. The technical problems of multi-source heterogeneous data fusion and real-time settlement monitoring and early warning are effectively solved, and the precision and timeliness of foundation settlement monitoring are improved.

Description

Technical Field

[0001] This invention relates to the field of foundation monitoring technology, and in particular to a method and system for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion. Background Technology

[0002] Monitoring the settlement of plant foundations in coal mining subsidence areas is a core component of infrastructure safety assurance, directly impacting production safety and economic loss control. With accelerating industrialization and increasingly complex geological environments, monitoring plant foundation deformation has become an indispensable key technology in engineering safety management.

[0003] Current methods for monitoring foundation settlement generally suffer from the limitation of relying on a single data source. Traditional monitoring methods either depend on satellite remote sensing to obtain large-scale deformation information or use ground-based measuring equipment for point-based monitoring. This fragmented monitoring approach makes it difficult to form a comprehensive and accurate deformation assessment. At the same time, existing technologies often struggle to balance data accuracy, monitoring range, and real-time performance, failing to meet the refined monitoring needs of complex industrial environments.

[0004] The core challenge in this field stems from the fundamental technical bottleneck of effectively integrating heterogeneous data from multiple sources. Satellite remote sensing data, ground-based measurement data, and interferometric radar data differ significantly in time references, spatial coordinate systems, and data formats, and the accuracy and confidence levels of these data also vary. This data heterogeneity directly leads to technical difficulties in multidimensional information fusion processing; data from different sources cannot form a unified deformation analysis model, thus affecting the accurate construction of the comprehensive deformation field. More critically, due to the lack of an effective data fusion mechanism, existing monitoring systems struggle to achieve dynamic weight adjustment and real-time early warning functions. When abnormal deformation occurs in the foundation, the system cannot promptly identify the risk level and trigger corresponding safety response measures.

[0005] Therefore, how to construct a collaborative fusion mechanism for multi-source heterogeneous data, realize unified processing of satellite remote sensing, ground monitoring and interferometric radar data, and establish an accurate three-dimensional deformation model and intelligent early warning system on this basis has become the key issue in solving the problem of accurate monitoring of foundation settlement of coal mining subsidence areas. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for monitoring the foundation of powerhouses in coal mining subsidence areas based on multi-source data fusion, which effectively solves the technical problems of multi-source heterogeneous data fusion and real-time settlement monitoring and early warning, and improves the accuracy and timeliness of foundation settlement monitoring.

[0007] To achieve the above objectives, the present invention provides the following solution:

[0008] A method for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion, comprising:

[0009] Acquire multi-source monitoring data of the target coal mining subsidence area's powerhouse region, wherein the multi-source data includes satellite remote sensing data, ground measurement data, and interferometric radar data;

[0010] The multi-source monitoring data is synchronized in time and unified in space to obtain spatially unified monitoring data;

[0011] The unified monitoring data of the space is fused to obtain comprehensive deformation data;

[0012] A three-dimensional settlement deformation field is constructed based on the comprehensive deformation data, and the deformation rate and acceleration information of each monitoring point are extracted by gradient calculation method.

[0013] Based on the deformation rate and acceleration information, a risk assessment of the powerhouse foundation in the target coal mining subsidence area is obtained.

[0014] Optionally, time synchronization and spatial unification of the multi-source monitoring data include:

[0015] A timestamp standardization algorithm is used to unify the time base of the multi-source monitoring data, obtain the initial time-synchronized multi-source monitoring data, and preprocess the initial time-synchronized multi-source monitoring data to obtain the time-synchronized multi-source monitoring data.

[0016] A coordinate transformation matrix algorithm is used to spatially unify the time-synchronized multi-source monitoring data. If the coordinate deviation exceeds the preset accuracy, the least squares method is used for coordinate correction to obtain the spatially unified monitoring data.

[0017] Optionally, preprocessing the initial time-synchronized multi-source monitoring data includes:

[0018] The acquisition time interval of each data point is calculated for the initial time synchronization of the multi-source monitoring data. If the time interval exceeds a preset threshold, it is determined that there are missing time points.

[0019] The missing time points are filled with data values ​​using a linear interpolation algorithm to generate time-synchronized multi-source monitoring data.

[0020] A data consistency processing method is used to unify the format and detect outliers in the time-synchronized multi-source monitoring data after completion, thereby obtaining the time-synchronized multi-source monitoring data.

[0021] Optionally, the fusion processing of the spatially unified monitoring data to obtain comprehensive deformation data includes: performing weighted fusion processing on the spatially unified monitoring data, using a Kalman filter algorithm to eliminate data noise and outliers, and if the variance of the fused data exceeds a preset range, readjusting the weight coefficients and repeating the fusion calculation to obtain the comprehensive deformation data.

[0022] Optionally, constructing a three-dimensional settlement deformation field based on the comprehensive deformation data includes:

[0023] S1. Perform three-dimensional spatial interpolation on the comprehensive deformation data at multiple time points to generate an initial three-dimensional settlement field;

[0024] S2. For the initial three-dimensional settlement field, calculate the interpolation error. If the error exceeds a preset threshold, record the error-exceeding area and obtain the error distribution area.

[0025] S3. Update the initial three-dimensional settlement field according to the error distribution area, return to S2 until the error meets the preset threshold, and obtain the three-dimensional settlement deformation field.

[0026] Optionally, updating the initial three-dimensional settlement field based on the error distribution region includes:

[0027] Based on the error distribution area, the control point density is increased, deformation data is regenerated, and spline interpolation is used to re-interpolate the updated deformation data to generate a new three-dimensional settlement field.

[0028] Optionally, the risk assessment for obtaining the foundation of the powerhouse in the target coal mining subsidence area includes:

[0029] If the deformation rate exceeds a preset warning threshold, the monitoring point is marked as an abnormal state;

[0030] Input the abnormal state and deformation acceleration into the random forest model to obtain the risk level.

[0031] This invention also provides a monitoring system for the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion, comprising:

[0032] The data acquisition module is used to acquire multi-source monitoring data of the target coal mining subsidence area, wherein the multi-source data includes satellite remote sensing data, ground measurement data and interferometric radar data;

[0033] The data alignment module is used to synchronize the multi-source monitoring data in time and spatially unify it to obtain spatially unified monitoring data.

[0034] The fusion module is used to fuse the unified monitoring data of the space to obtain comprehensive deformation data;

[0035] The data extraction module is used to construct a three-dimensional settlement deformation field based on the comprehensive deformation data, and to extract the deformation rate and acceleration information of each monitoring point through gradient calculation method;

[0036] The risk assessment module is used to obtain a risk assessment of the foundation of the powerhouse in the target coal mining subsidence area based on the deformation rate and acceleration information.

[0037] The beneficial effects of this invention are as follows: This invention discloses a method and system for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion. It achieves spatiotemporal synchronization of multi-source data through timestamp standardization and coordinate transformation, followed by data fusion. Kalman filtering is used to eliminate noise, constructing a three-dimensional settlement deformation field and extracting dynamic deformation parameters. When the settlement rate or acceleration exceeds the warning threshold, a risk warning and emergency response are automatically triggered. This method effectively solves the technical challenges of multi-source heterogeneous data fusion and real-time settlement monitoring and early warning, improving the accuracy and timeliness of foundation settlement monitoring and providing reliable technical support for engineering safety. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a flowchart of a method for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion, according to an embodiment of the present invention. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0042] Example 1:

[0043] like Figure 1 The embodiment shown provides a method for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion, including:

[0044] Acquire multi-source monitoring data of the target coal mining subsidence area, including satellite remote sensing data, ground measurement data, and interferometric radar data;

[0045] Synchronize and unify multi-source monitoring data in time and space to obtain spatially unified monitoring data;

[0046] Spatial unified monitoring data are fused and processed to obtain comprehensive deformation data;

[0047] A three-dimensional settlement deformation field is constructed based on comprehensive deformation data, and the deformation rate and acceleration information of each monitoring point are extracted by gradient calculation method.

[0048] Based on deformation rate and acceleration information, a risk assessment of the powerhouse foundation in the target coal mining subsidence area is obtained.

[0049] Furthermore, time synchronization and spatial unification of multi-source monitoring data include:

[0050] A timestamp standardization algorithm is used to unify the time base of multi-source monitoring data, obtain the initial time-synchronized multi-source monitoring data, and preprocess the initial time-synchronized multi-source monitoring data to obtain the time-synchronized multi-source monitoring data.

[0051] A coordinate transformation matrix algorithm is used to spatially unify the time-synchronized multi-source monitoring data. If the coordinate deviation exceeds the preset accuracy, the least squares method is used for coordinate correction to obtain spatially unified monitoring data.

[0052] Furthermore, preprocessing of the initial time-synchronized multi-source monitoring data includes:

[0053] The acquisition time interval of each data point is calculated for the multi-source monitoring data of the initial time synchronization. If the time interval exceeds the preset threshold, it is determined that there are missing time points.

[0054] The missing time points are filled with data values ​​using a linear interpolation algorithm to generate time-synchronized multi-source monitoring data.

[0055] A data consistency processing method is adopted to unify the format and detect outliers in the time-synchronized multi-source monitoring data after completion, so as to obtain the time-synchronized multi-source monitoring data.

[0056] Specifically, satellite remote sensing data, ground measurement data, and interferometric radar data are acquired. A timestamp standardization algorithm is used to align the time base of each data source, generating a preliminary aligned dataset. For this preliminary aligned dataset, the acquisition time interval for each data source is calculated. If the time interval exceeds a preset threshold, missing time points are identified. A linear interpolation algorithm is used to complete the missing time points, generating a completed time-aligned dataset. Data consistency processing methods are employed to unify the format and detect outliers in the completed time-aligned dataset, resulting in a consistent dataset. Based on the consistent dataset, multi-source data is fused to generate a time-synchronized multi-source monitoring dataset. Time series analysis is used to check the data integrity and continuity in the time-synchronized dataset and determine if any abnormal fluctuations exist. If abnormal fluctuations are detected, a sliding window algorithm is used to perform local interpolation repair on the abnormal time periods, resulting in the final time-synchronized monitoring dataset.

[0057] Obtain the spatial coordinates of each data source in the time synchronization dataset, and use a coordinate transformation matrix algorithm to calculate the transformation matrix T, where T is a 3×3 matrix with elements t ij Let the rotation and translation parameters from coordinate axis i to j represent the initial coordinate point set P1 under the unified coordinate system. Extract the deviation value D of each point from the standard coordinate system from the coordinate point set P1, and determine whether the deviation value D exceeds the millimeter-level accuracy threshold T. d The set of deviation exceeding the limit points P is obtained. d If the deviation exceeds the limit, the set of points P... d If not empty, then for P d For points in the matrix, the coordinates are optimized using the least squares method, and the objective function F = ∑(D) is constructed. i ^2), where D i For the deviation of point i, the minimum value of F is obtained by iteratively solving for the corrected coordinate point set P2. Based on the corrected coordinate point set P2, the unified spatial reference is updated.

[0058] Furthermore, the spatially unified monitoring data is fused to obtain comprehensive deformation data, including: weighted fusion processing of the spatially unified monitoring data, using the Kalman filter algorithm to eliminate data noise and outliers, and if the variance of the fused data exceeds the preset range, the weight coefficients are readjusted and the fusion calculation is repeated to obtain comprehensive deformation data.

[0059] Specifically, a weighted fusion process is performed on multi-source heterogeneous data using a weight allocation matrix. Initial fusion is achieved using a pre-established weight allocation matrix to obtain initial fused data. A Kalman filter algorithm is then applied to the initial fused data to obtain filtered data, eliminating noise and outliers to obtain smoothed data. The variance of the smoothed data is assessed; if it exceeds a preset threshold, the weight coefficients are adjusted to generate an updated weight allocation matrix. The updated weight allocation matrix is ​​then used to re-perform the weighted fusion process on the multi-source heterogeneous data to obtain secondary fused data. A Kalman filter algorithm is then applied to the secondary fused data to obtain optimized filtered data, eliminating residual noise and outliers. The variance is calculated based on the optimized filtered data, and its accuracy is checked against a preset threshold to obtain satisfactory comprehensive deformed data. Finally, the comprehensive deformed data is used to generate the final output, determining the stability of the data fusion and yielding optimized comprehensive deformed data.

[0060] For example, satellite remote sensing data has a wide coverage but low resolution, so it can be assigned an initial weight of 0.4; ground sensor data has high accuracy but limited coverage, so it has a weight of 0.5; interferometric radar data has high resolution but is affected by weather, so it has a weight of 0.3. The Kalman filter algorithm eliminates noise through prediction and update steps when processing the initial fused data. For example, in ground measurements, the initial fused data may contain outliers introduced by equipment jitter or environmental interference. The Kalman filter predicts the theoretical values ​​of the data by establishing a state transition model and updates the results by combining actual observations. In one embodiment, the data variance is used to assess the stability of the smoothed data. For example, if the preset variance threshold is 0.05, and the variance of a certain smoothed data is 0.08, it indicates that the data fluctuates greatly and there may be unremoved noise. In this case, the weight allocation matrix is ​​regenerated by reducing the weight of the outlier data sources (e.g., reducing the weight of the interferometric radar data from 0.3 to 0.1). This dynamic adjustment ensures the accuracy of the fused data.

[0061] Furthermore, the three-dimensional settlement deformation field constructed based on the comprehensive deformation data includes:

[0062] S1. Perform three-dimensional spatial interpolation on the comprehensive deformation data at multiple time points to generate an initial three-dimensional settlement field;

[0063] S2. For the initial three-dimensional settlement field, calculate the interpolation error. If the error exceeds the preset threshold, record the error exceeding the limit area and obtain the error distribution area.

[0064] S3. Update the initial three-dimensional settlement field according to the error distribution area, return to S2 until the error meets the preset threshold, and obtain the three-dimensional settlement deformation field.

[0065] Furthermore, based on the error distribution region, the initial three-dimensional settlement field is updated, including:

[0066] Based on the error distribution area, the density of control points is increased, deformation data is regenerated, and spline interpolation is used to re-interpolate the updated deformation data to generate a new three-dimensional settlement field.

[0067] Specifically, comprehensive deformation data is acquired, spatial coordinates and settlement are extracted to obtain an initial settlement dataset. Spline interpolation is used to perform three-dimensional spatial interpolation on the initial settlement dataset to generate a continuous three-dimensional settlement field, resulting in a preliminary settlement field distribution. For the preliminary settlement field distribution, the interpolation error is calculated. If the error exceeds a preset threshold, the error-exceeding region is recorded, resulting in an error distribution region. Based on the error distribution region, the control point density is increased, and a high-density point cloud dataset is regenerated to obtain an updated settlement dataset. Spline interpolation is then used to re-interpolate the updated settlement dataset to generate a new three-dimensional settlement field, resulting in an optimized settlement field distribution. For the optimized settlement field distribution, the interpolation error is verified. If the error meets a preset threshold, the final three-dimensional settlement deformation field is determined. Through spatial interpolation accuracy analysis, the quality of the final three-dimensional settlement deformation field is assessed, resulting in an accurate three-dimensional settlement model.

[0068] A continuous 3D settlement field is generated using spline interpolation. Spline interpolation ensures the continuity of the settlement field by constructing smooth curves between point cloud data. If, after interpolation, a region is found to have drastic settlement variations, with a maximum settlement value reaching 8 mm, it indicates potential local foundation instability. By visualizing the preliminary settlement field, engineers can intuitively identify high-risk areas. Calculating the interpolation error is crucial for assessing the quality of the preliminary settlement field. If the error exceeds a preset threshold, such as 0.5 mm, the area exceeding the error limit must be recorded. The density of control points is increased for these areas. For example, 10 GNSS control points are added in the southwest corner, with the point spacing reduced from 5 meters to 2 meters. A high-density point cloud dataset is generated by combining the extracted spatial coordinates and settlement data. The updated dataset contains more local details, and the settlement distribution is more refined. After re-interpolating with splines, the error of the new 3D settlement field is reduced to within 0.3 mm, meeting the preset threshold. This method improves the accuracy and reliability of the settlement field. In one possible implementation, spatial interpolation accuracy analysis can be quality-assessed through statistical error distribution and the smoothness of the settlement field. For example, the smoothness of the model is determined by curvature analysis; a curvature change rate below 0.1 indicates a smooth and reliable settlement field. This assessment ensures that the model can provide reliable deformation monitoring data for subway construction, supporting engineering decisions and optimizing construction plans.

[0069] Furthermore, the risk assessment for obtaining the foundation of the powerhouse in the target coal mining subsidence area includes:

[0070] If the deformation rate exceeds the preset warning threshold, the monitoring point is marked as abnormal.

[0071] Input the abnormal state and deformation acceleration into the random forest model to obtain the risk level.

[0072] Specifically, in the scenario of foundation settlement monitoring, the acquisition of monitoring point data from the three-dimensional settlement deformation field can be achieved by using high-precision sensors to collect spatial coordinates and settlement amounts, generating a point cloud dataset containing X, Y, and Z coordinates and settlement values. Assuming a monitoring area contains 100 monitoring points, each point records coordinates such as (100.5, 200.3, 50.2) and a settlement amount such as -5.2 mm. Using time-series data, the deformation rate of each monitoring point is calculated. For example, if the settlement of a point changes from -5.2 mm to -6.8 mm within 30 days, the deformation rate is approximately -0.053 mm / day. Acceleration is calculated based on the rate change; if the rate changes to -0.060 mm / day over the next 30 days, the acceleration is -0.00023 mm / day.

[0073] In one possible implementation, the warning threshold for deformation rate is set to -0.050 mm / day. If the rate at a monitoring point reaches -0.053 mm / day, the system automatically marks that point as an abnormal state and generates an "abnormal" label. The abnormal state, combined with acceleration data, is used to determine the risk level through a random forest algorithm.

[0074] If the settlement rate exceeds the first-level warning threshold for three consecutive monitoring cycles, a high-risk warning signal will be automatically triggered. If the settlement acceleration suddenly exceeds the second-level warning threshold, the emergency response mode will be activated simultaneously to obtain real-time foundation settlement safety status assessment results.

[0075] Example 2:

[0076] This embodiment provides a monitoring system for the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion, implemented according to the method of Embodiment 1, including:

[0077] The data acquisition module is used to acquire multi-source monitoring data of the target coal mining subsidence area, including satellite remote sensing data, ground measurement data and interferometric radar data.

[0078] The data alignment module is used to synchronize and unify the time and space of multi-source monitoring data to obtain spatially unified monitoring data.

[0079] The fusion module is used to fuse spatially unified monitoring data to obtain comprehensive deformation data;

[0080] The data extraction module is used to construct a three-dimensional settlement deformation field based on comprehensive deformation data, and to extract the deformation rate and acceleration information of each monitoring point through gradient calculation method;

[0081] The risk assessment module is used to obtain risk assessments of the foundation of the powerhouse in the target coal mining subsidence area based on deformation rate and acceleration information.

[0082] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion, characterized in that, include: Acquire multi-source monitoring data of the target coal mining subsidence area's powerhouse region, wherein the multi-source data includes satellite remote sensing data, ground measurement data, and interferometric radar data; The multi-source monitoring data is synchronized in time and unified in space to obtain spatially unified monitoring data; The unified monitoring data of the space is fused to obtain comprehensive deformation data; A three-dimensional settlement deformation field is constructed based on the comprehensive deformation data, and the deformation rate and acceleration information of each monitoring point are extracted by gradient calculation method. Based on the deformation rate and acceleration information, a risk assessment of the powerhouse foundation in the target coal mining subsidence area is obtained.

2. The method for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion according to claim 1, characterized in that, The time synchronization and spatial unification of the multi-source monitoring data include: A timestamp standardization algorithm is used to unify the time base of the multi-source monitoring data, obtain the initial time-synchronized multi-source monitoring data, and preprocess the initial time-synchronized multi-source monitoring data to obtain the time-synchronized multi-source monitoring data. A coordinate transformation matrix algorithm is used to spatially unify the time-synchronized multi-source monitoring data. If the coordinate deviation exceeds the preset accuracy, the least squares method is used for coordinate correction to obtain the spatially unified monitoring data.

3. The method for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion according to claim 2, characterized in that, Preprocessing of the initial time-synchronized multi-source monitoring data includes: The acquisition time interval of each data point is calculated for the initial time synchronization of the multi-source monitoring data. If the time interval exceeds a preset threshold, it is determined that there are missing time points. The missing time points are filled with data values ​​using a linear interpolation algorithm to generate time-synchronized multi-source monitoring data. A data consistency processing method is used to unify the format and detect outliers in the time-synchronized multi-source monitoring data after completion, thereby obtaining the time-synchronized multi-source monitoring data.

4. The method for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion according to claim 1, characterized in that, The process of fusing the spatially unified monitoring data to obtain comprehensive deformation data includes: performing weighted fusing processing on the spatially unified monitoring data, using a Kalman filter algorithm to eliminate data noise and outliers, and if the variance of the fused data exceeds a preset range, readjusting the weight coefficients and repeating the fusing calculation to obtain the comprehensive deformation data.

5. The method for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion according to claim 1, characterized in that, Constructing a three-dimensional settlement deformation field based on the comprehensive deformation data includes: S1. Perform three-dimensional spatial interpolation on the comprehensive deformation data at multiple time points to generate an initial three-dimensional settlement field; S2. For the initial three-dimensional settlement field, calculate the interpolation error. If the error exceeds a preset threshold, record the error-exceeding area and obtain the error distribution area. S3. Update the initial three-dimensional settlement field according to the error distribution area, return to S2 until the error meets the preset threshold, and obtain the three-dimensional settlement deformation field.

6. The method for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion according to claim 5, characterized in that, Updating the initial three-dimensional settlement field based on the error distribution region includes: Based on the error distribution area, the control point density is increased, deformation data is regenerated, and spline interpolation is used to re-interpolate the updated deformation data to generate a new three-dimensional settlement field.

7. The method for monitoring the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion according to claim 1, characterized in that, The risk assessment for obtaining the foundation of the powerhouse in the target coal mining subsidence area includes: If the deformation rate exceeds a preset warning threshold, the monitoring point is marked as an abnormal state; Input the abnormal state and deformation acceleration into the random forest model to obtain the risk level.

8. A monitoring system for the foundation of a powerhouse in a coal mining subsidence area based on multi-source data fusion, implemented according to any one of claims 1-7, characterized in that, include: The data acquisition module is used to acquire multi-source monitoring data of the target coal mining subsidence area, wherein the multi-source data includes satellite remote sensing data, ground measurement data and interferometric radar data; The data alignment module is used to synchronize the multi-source monitoring data in time and spatially unify it to obtain spatially unified monitoring data. The fusion module is used to fuse the unified monitoring data of the space to obtain comprehensive deformation data; The data extraction module is used to construct a three-dimensional settlement deformation field based on the comprehensive deformation data, and to extract the deformation rate and acceleration information of each monitoring point through gradient calculation method; The risk assessment module is used to obtain a risk assessment of the foundation of the powerhouse in the target coal mining subsidence area based on the deformation rate and acceleration information.

Images