Rockfill dam concrete panel strain soft sensing method and system

Abstract

The invention discloses a rock-fill dam concrete panel strain soft sensing method and system in the technical field of structural safety monitoring, and the method comprises the steps: obtaining the monitoring data of a rock-fill deformation sensor, inputting the monitoring data into a pre-constructed mapping model, and calculating the monitoring data of a concrete panel strain virtual sensor; construction of a mapping model: establishing a finite element simulation model according to an actual rock-fill dam project; simulating conditions of rock-fill deformation and concrete panel strain under an environmental load by using a finite element simulation model to obtain simulation result data; the method comprises the following steps of: establishing a mapping relation between rockfill deformation and concrete panel strain through actually acquired rockfill deformation monitoring data, concrete panel strain monitoring data and simulation result data; and constructing a final mapping model based on the mapping relation. The rock-fill dam concrete panel strain soft sensing method can provide more comprehensive panel strain virtual monitoring data under the condition of low monitoring cost.

Description

Technical Field

[0001] The invention relates to a soft-sensing method and system for strain of a rockfill dam concrete panel, belonging to the technical field of structural safety monitoring. Background Art

[0002] Strain gauges are usually placed at key locations on the concrete face rockfill dam panels to monitor the stress state of the panel structure. The locations of the strain gauges are generally selected during the design phase based on engineering experience and finite element simulation results.

[0003] For ultra-high dams or concrete-faced rockfill dams with poor dam material (foundation) conditions, the panels are usually subjected to large and complex stresses, making the accurate arrangement and stable operation of strain gauges even more important. However, due to inaccurate estimation of engineering parameter conditions, the strain gauges sparsely arranged during the design phase based on engineering experience and finite element simulation results are difficult to accurately capture the maximum panel strain values. In addition, the harsh working conditions also lead to a low survival rate. Summary of the Invention

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a soft sensing method and system for rockfill dam concrete panel strain, which can provide more comprehensive panel strain virtual monitoring data at a low monitoring cost.

[0005] To achieve the above object, the present invention is implemented by adopting the following technical solutions:

[0006] In a first aspect, the present invention provides a soft sensing method for strain of a rockfill dam concrete panel, comprising:

[0007] Obtain monitoring data from the rockfill deformation sensor, input it into a pre-built mapping model, and calculate monitoring data from the concrete panel strain virtual sensor;

[0008] The construction of the mapping model includes:

[0009] Establish a finite element simulation model based on the actual rockfill dam project;

[0010] Using the actual collected rockfill deformation monitoring data and concrete panel strain monitoring data, the material parameters in the finite element simulation model are inversely calibrated to obtain the finite element simulation model after parameter calibration;

[0011] Using the finite element simulation model after the parameter calibration, the deformation of the rockfill body and the strain of the concrete panel under the real environmental load are simulated to obtain simulation result data;

[0012] The mapping relationship between rockfill deformation and concrete panel strain is established by using the actual collected rockfill deformation monitoring data, concrete panel strain monitoring data and simulation result data.

[0013] Based on the mapping relationship, a final mapping model is constructed.

[0014] Furthermore, the actually collected rockfill deformation monitoring data includes monitoring data of rockfill deformation sensors. The actual collected concrete panel strain monitoring data includes the monitoring data of the concrete panel strain sensor ;

[0015] Y ̃ = [ y 1 ̃ y 2 ̃ ⋮ y p ̃ ] ; S ̃ = [ s 1 ̃ s 2 ̃ ⋮ s q ̃ ] ;

[0016] Where: Indicates the number The monitoring data of the rockfill deformation sensor, , Indicates the total number of rockfill deformation sensors; Indicates the number Monitoring data of concrete panel strain sensors, , represents the total number of concrete panel strain sensors; and are all time series vectors.

[0017] Furthermore, the finite element simulation model after parameter calibration is used to simulate the deformation of the rockfill body and the strain of the concrete panel under real environmental loads, and the simulation result data are obtained, including:

[0018] The finite element simulation model after parameter calibration includes several rockfill calculation nodes and concrete panel calculation nodes, and calculates and extracts the deformation simulation result data of the rockfill calculation nodes. and concrete panel calculation node strain simulation result data ;

[0019] Y ¯ = [ y 1 ¯ y 2 ¯ ⋮ y i ¯ ] ; S ¯ = [ s 1 ¯ s 2 ¯ ⋮ s J ¯ ] ;

[0020] Where: Indicates the number The deformation simulation results at the calculation nodes of the rockfill body are as follows: , Indicates the total number of rockfill calculation nodes; Indicates the number The strain simulation results at the calculation nodes of the concrete panel are as follows: , Indicates the total number of calculation nodes of the concrete panel; and are all time series vectors.

[0021] Furthermore, the computing nodes are divided into monitoring nodes and virtual nodes. The monitoring nodes include rockfill monitoring nodes and concrete panel monitoring nodes, and their positions correspond one-to-one to the actual positions of the rockfill deformation sensor and the concrete panel strain sensor, respectively. The virtual nodes include rockfill virtual nodes and concrete panel virtual nodes.

[0022] The deformation simulation result data and strain simulation result data of the rockfill monitoring node and the concrete panel monitoring node are recorded as and The deformation simulation result data and strain simulation result data of the virtual node of the rockfill body and the virtual node of the concrete panel are recorded as and , and is subject to the following relations:

[0023] ; ; .

[0024] Furthermore, a mapping relationship between rockfill deformation and concrete panel strain is established using the actual collected rockfill deformation monitoring data, concrete panel strain monitoring data, and simulation result data, including:

[0025] According to the deformation simulation result data of the rockfill monitoring node and strain simulation results of concrete panel monitoring nodes , establish the mapping relationship between the rockfill monitoring node and the concrete panel monitoring node , as shown below, the mapping relationship By transferring the function tensor express:

[0026] ;

[0027] According to the deformation simulation result data of the rockfill monitoring node Strain simulation results of virtual nodes of concrete panels , establish the mapping relationship between the rockfill monitoring node and the concrete panel virtual node , as shown below, the mapping relationship By transferring the function tensor express:

[0028] ;

[0029] According to the monitoring data of the rockfill deformation sensor and monitoring data from concrete panel strain sensors , establish the actual monitoring data mapping relationship between the rockfill monitoring node and the concrete panel monitoring node , as shown below, the mapping relationship By transferring the function tensor express:

[0030] ; ; ;

[0031] Where: represents monitoring data provided by the concrete panel strain sensor at the concrete panel monitoring node; Indicates monitoring data provided by rockfill deformation sensors at rockfill monitoring nodes; and are all time series vectors.

[0032] Furthermore, based on the mapping relationship, a final mapping model is constructed, including:

[0033] Based on the mapping relationship and Building a mapping model , the mapping model It is used to combine the simulation result data with the actual monitoring data, as shown in the following formula: By transferring the function tensor express:

[0034] ; ; ;

[0035] Where: for The generalized inverse matrix of ;

[0036] Based on the mapping relationship and mapping models and build the final mapping model , as shown below, the final mapping model By transferring the function tensor express:

[0037] ; ;

[0038] The monitoring data of the rockfill deformation sensor Input to the final mapping model , calculate the monitoring data of the concrete panel strain virtual sensor , as shown below:

[0039] ; ;

[0040] Where: Represents the monitoring data provided by the concrete panel strain virtual sensor at the virtual node of the concrete panel.

[0041] In a second aspect, the present invention provides a soft sensing system for strain of a rockfill dam concrete panel, comprising:

[0042] The simulation and parameter calibration module is used to establish a finite element simulation model based on the actual rockfill dam project and to perform inverse calibration of the material parameters in the finite element simulation model using the actual collected rockfill deformation monitoring data and concrete panel strain monitoring data;

[0043] The environmental load simulation module is used to simulate the deformation of the rockfill body and the strain of the concrete panel under real environmental loads through the finite element simulation model after parameter calibration, and obtain simulation result data;

[0044] a mapping model establishment module for establishing a mapping relationship between rockfill deformation and concrete panel strain using actually collected rockfill deformation monitoring data, concrete panel strain monitoring data, and simulation result data, and constructing a final mapping model based on the mapping relationship;

[0045] The virtual monitoring data calculation module is used to calculate the monitoring data of the concrete panel strain virtual sensor using the final mapping model.

[0046] Furthermore, the actually collected rockfill deformation monitoring data includes monitoring data of rockfill deformation sensors. The actual collected concrete panel strain monitoring data includes the monitoring data of the concrete panel strain sensor ;

[0047] Y ̃ = [ y 1 ̃ y 2 ̃ ⋮ y p ̃ ] ; S ̃ = [ s 1 ̃ s 2 ̃ ⋮ s q ̃ ] ;

[0048] Where: Indicates the number The monitoring data of the rockfill deformation sensor, , Indicates the total number of rockfill deformation sensors; Indicates the number Monitoring data of concrete panel strain sensors, , represents the total number of concrete panel strain sensors; and are all time series vectors.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] The proposed soft-sensing method for rockfill dam concrete panel strain, developed by the present invention, achieves soft sensing of concrete panel strain by establishing a finite element simulation model and integrating it with actual monitoring data, significantly improving monitoring efficiency and coverage. This method utilizes actual monitoring data to inversely calibrate model parameters, ensuring the accuracy of simulation results. By establishing a mapping relationship between rockfill deformation and concrete panel strain, high-precision virtual monitoring data is derived. This approach not only reduces reliance on physical sensors and monitoring costs, but also provides real-time and continuous information on concrete panel strain, providing reliable data support for safety assessment and early warning of rockfill dams, with significant engineering application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a flowchart of the mapping model construction provided by the first embodiment of the present invention;

[0052] Figure 2 This is a schematic diagram of a concrete face rockfill dam provided in Example 1 of the present invention;

[0053] Figure 3 This is a cross-sectional schematic diagram of a finite element simulation model of a concrete face rockfill dam provided in the first embodiment of the present invention;

[0054] Figure 4 This is a schematic diagram of the mapping relationship between a concrete face rockfill dam and a finite element simulation model provided in the first embodiment of the present invention;

[0055] The meanings of the reference numerals in the figures are as follows:

[0056] 101. Concrete-faced rockfill dam; 102. Rockfill body; 103. Concrete panel; 104. Rockfill body deformation sensor; 105. Concrete panel strain sensor; 106. Concrete panel strain virtual sensor; 201. Finite element simulation model; 202. Finite element simulation model of rockfill body; 203. Finite element simulation model of concrete panel; 204. Rockfill body monitoring node; 205. Rockfill body virtual node; 206. Concrete panel monitoring node; 207. Concrete panel virtual node. DETAILED DESCRIPTION

[0057] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solution of the present application, rather than limitations on the technical solution of the present application. Unless there is a conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.

[0058] Example 1:

[0059] The soft sensing method for strain of rockfill dam concrete panels provided in this embodiment specifically includes the following steps:

[0060] Obtain monitoring data from the rockfill deformation sensor, input it into a pre-built mapping model, and calculate monitoring data from the concrete panel strain virtual sensor;

[0061] The construction of the mapping model is as follows: Figure 1 Shown, including:

[0062] Based on the actual rockfill dam project, a finite element simulation model 201 was established;

[0063] Using the actually collected rockfill deformation monitoring data and concrete panel strain monitoring data, the material parameters in the finite element simulation model 201 are inversely calibrated to obtain a finite element simulation model after parameter calibration;

[0064] Using the finite element simulation model after the parameter calibration, the deformation of the rockfill body and the strain of the concrete panel under the real environmental load are simulated to obtain simulation result data;

[0065] The mapping relationship between rockfill deformation and concrete panel strain is established by using the actual collected rockfill deformation monitoring data, concrete panel strain monitoring data and simulation result data.

[0066] Based on the mapping relationship, a final mapping model is constructed.

[0067] Traditionally, engineers install strain gauges (sensors) on concrete panels to monitor their deformation. However, these strain gauges are limited in number and their placement may not be optimal. Especially in the case of extremely high dams or complex geological conditions, the strain gauges may not accurately capture the panel's maximum deformation point. Furthermore, strain gauges are easily damaged in harsh environments, resulting in incomplete data.

[0068] Therefore, this embodiment uses the deformation data of the rockfill to infer the strain of the concrete panel. Using the deformation data of the rockfill, combined with a computer-implemented finite element simulation model 201, the strain of the concrete panel is "virtually" calculated, thus overcoming the shortcomings of traditional strain gauges and providing more comprehensive and accurate strain data.

[0069] The present invention's soft-sensing method for rockfill dam concrete panel strain aims to establish a mapping relationship between rockfill deformation and concrete panel strain using a finite element simulation model 201 and actual monitoring data, thereby enabling soft sensing of concrete panel strain. The core of this method lies in utilizing finite element simulation technology, combined with actual monitoring data, to construct a mathematical model that accurately reflects the deformation and strain behavior of a rockfill dam under real-world loads. This model is then used to infer virtual strain data for the concrete panel, thereby supplementing the concrete panel strain.

[0070] First, based on the actual rockfill dam project, a finite element simulation model containing information such as the geometric structure, material properties, and boundary conditions of the rockfill body and concrete panel was established201. The main structure of the rockfill dam is mainly composed of the rockfill body and the concrete panel. As the core supporting part of the dam, the rockfill body is made of stone filling and bears the weight of the dam body itself, the upstream and downstream water pressure, and other loads. The concrete panel is located on the upstream side of the dam body and is an important anti-seepage structure, mostly made of reinforced concrete. Figure 2 As shown, the finite element simulation model 201 established in this embodiment includes a rockfill finite element simulation model 202 and a concrete face finite element simulation model 203. The establishment of finite element simulation model 201 is fundamental to the entire process, and its accuracy directly impacts subsequent simulation results and the establishment of mapping relationships. Using finite element analysis software (such as ANSYS and ABAQUS), the structural characteristics of the rockfill dam are accurately simulated, providing a foundation for subsequent material parameter inversion calibration.

[0071] Next, the material parameters in the finite element simulation model 201 are inversely calibrated using the actual monitoring data. Figure 2 As shown, the main structure of the concrete face rockfill dam 101 includes the core supporting part of the dam, the rockfill body 102, and the concrete face 103 located on the upstream side of the dam body. The actual collected rockfill deformation monitoring data and concrete face strain monitoring data include the monitoring data extracted from the concrete face strain sensor 105. and monitoring data extracted from the rockfill deformation sensor 104 , as shown below:

[0072] ; ;

[0073] Where: Indicates the number The monitoring data of the rockfill deformation sensor, , Indicates the total number of rockfill deformation sensors; Indicates the number Monitoring data of concrete panel strain sensors, , represents the total number of concrete panel strain sensors; and are all time series vectors.

[0074] Using optimization algorithms (such as genetic algorithms and particle swarm optimization), the material parameters in finite element simulation model 201 are adjusted to minimize the error between the simulation results and the actual monitoring data. This step ensures that finite element simulation model 201 accurately reflects the mechanical behavior of the rockfill dam under actual operating conditions, thereby improving the reliability of the simulation results.

[0075] After completing the inverse calibration of material parameters, the calibrated finite element simulation model 201 is used to simulate the deformation of the rockfill body and the strain of the concrete panel under real environmental loads. Real environmental loads include external factors such as water pressure and temperature changes, which can have a significant impact on the deformation and strain of the rockfill dam. The finite element simulation model 201 contains several rockfill body calculation nodes and concrete panel calculation nodes. The deformation simulation result data of the rockfill body calculation nodes are extracted through the finite element simulation model 201. and concrete panel calculation node strain simulation result data , as shown below:

[0076] ; ;

[0077] Where: Indicates the number The deformation simulation results at the calculation nodes of the rockfill body are as follows: , Indicates the total number of rockfill calculation nodes; Indicates the number The strain simulation results at the calculation nodes of the concrete panel are as follows: , Indicates the total number of calculation nodes of the concrete panel; and are all time series vectors.

[0078] These simulation results provide important data support for the subsequent establishment of mapping relationships.

[0079] The computing nodes are divided into monitoring nodes and virtual nodes, such as Figure 3 As shown, the monitoring nodes include a rockfill monitoring node 204 and a concrete panel monitoring node 206, whose positions correspond one-to-one to the actual positions of the rockfill deformation sensor 104 and the concrete panel strain sensor 105 respectively; the virtual nodes include a rockfill virtual node 205 and a concrete panel virtual node 207;

[0080] The deformation simulation result data and strain simulation result data of the rockfill monitoring node 204 and the concrete panel monitoring node 206 are respectively recorded as and The deformation simulation result data and strain simulation result data of the rockfill virtual node 205 and the concrete panel virtual node 206 are respectively recorded as and , and is subject to the following relations:

[0081] ; ; .

[0082] Finally, the established mapping relationship is used to calculate virtual monitoring data for the concrete panel, enabling soft sensing of concrete panel strain. Specifically, the monitoring data from the rockfill deformation sensor 104 is input, and the mapping model is used to calculate virtual monitoring data for the strain of the concrete panel virtual node 207. This virtual monitoring data serves as the output of the concrete panel virtual strain sensor 106, enabling real-time monitoring of the concrete panel strain. In this way, comprehensive monitoring of concrete panel strain can be achieved without adding physical sensors, thereby reducing monitoring costs and improving monitoring efficiency.

[0083] The simulation result data calculated by the actual monitoring data and the finite element simulation model 201, such as Figure 4 As shown, a mapping relationship is established between the deformation of the rockfill body and the strain of the concrete panel. The key to this step is to match the actual monitoring data provided by the sensor with the simulation data provided by the finite element simulation model 201 and to establish a mathematical model between the two using a transfer function tensor. The transfer function tensor can be solved using the least squares method or other optimization methods. Its purpose is to map the deformation data of the rockfill body to the strain data of the concrete panel, thereby achieving soft sensing of the concrete panel strain.

[0084] Specifically, the process of establishing each mapping relationship is as follows:

[0085] (1) Establishing a mapping relationship between the rockfill monitoring node 204 and the concrete panel monitoring node 206

[0086] According to the deformation simulation result data of the rockfill monitoring node 204 and the strain simulation result data of the concrete panel monitoring node 206 , establish a mapping relationship between the rockfill monitoring node 204 and the concrete panel monitoring node 206 , as shown below, the mapping relationship By transferring the function tensor express:

[0087] ;

[0088] Mapping relationship The role of the simulation data is to reveal the intrinsic connection between the rockfill deformation monitoring node and the concrete panel strain monitoring node, provide a basis for the subsequent construction of more complex relationships, and build a bridge between the two from the perspective of simulation data.

[0089] (2) Establishing a mapping relationship between the rockfill monitoring node 204 and the concrete panel virtual node 207

[0090] According to the deformation simulation result data of the rockfill monitoring node 204 and the strain simulation result data of the concrete panel virtual node 207 , establish a mapping relationship between the rockfill monitoring node 204 and the concrete panel virtual node 207 , as shown below, the mapping relationship By transferring the function tensor express:

[0091] ;

[0092] Mapping relationship It is mainly used to establish the relationship between the rockfill deformation monitoring node and the calculation results of the concrete panel strain virtual node (corresponding to the concrete panel strain virtual sensor 106), and use the monitoring information of the rockfill body to infer the strain of the concrete panel strain virtual sensor 106, thereby expanding the coverage of strain monitoring.

[0093] (3) Establishing a mapping relationship between the actual monitoring data of the rockfill monitoring node 204 and the concrete panel monitoring node 206

[0094] According to the monitoring data extracted from the rockfill deformation sensor 104 and monitoring data extracted from the concrete panel strain sensor 105 , establish the actual monitoring data mapping relationship between the rockfill monitoring node 204 and the concrete panel monitoring node 206 , as shown below, the mapping relationship By transferring the function tensor express:

[0095] ; ; ;

[0096] Where: represents monitoring data provided by the concrete panel strain sensor at the concrete panel monitoring node; Indicates monitoring data provided by rockfill deformation sensors at rockfill monitoring nodes; and are all time series vectors.

[0097] Mapping relationship It reflects the actual correlation between the monitoring data of the rockfill deformation monitoring node 204 and the concrete panel strain monitoring node 206, organically combines the actual monitoring data, and makes the relationship more consistent with the actual monitoring situation.

[0098] Next, integrate the mapping relationship and Building a mapping model , the mapping model It is used to combine the simulation result data with the actual monitoring data, as shown in the following formula: By transferring the function tensor express:

[0099] ; ; ;

[0100] Where: for The generalized inverse matrix of ;

[0101] The integration of the relationship between simulation result data and actual monitoring data can make the virtual strain monitoring results more consistent with the actual situation and effectively improve the accuracy of monitoring.

[0102] Based on the mapping relationship and mapping models Building a mapping model , the mapping model The virtual monitoring data for calculating the concrete panel virtual node 207 is shown in the following equation: By transferring the function tensor express:

[0103] ; ;

[0104] The monitoring data of the rockfill deformation sensor 104 Input to the final mapping model , calculate the monitoring data of the concrete panel strain virtual sensor 106 , as shown below:

[0105] ; ;

[0106] Where: It represents the monitoring data provided by the concrete panel strain virtual sensor at the concrete panel virtual node 207 .

[0107] In summary, this invention combines finite element simulation technology with actual monitoring data to provide an efficient and low-cost method for monitoring strain in rockfill dam concrete panels. This method not only enables real-time monitoring of concrete panel strain but also provides important data support for safety assessment and maintenance of rockfill dams, promising broad application prospects.

[0108] The following further illustrates the implementation of the present invention through a specific example.

[0109] A concrete-faced rockfill dam project has a height of 120 meters and a crest length of 250 meters. The thickness of the concrete face varies from 0.3 to 0.72 meters, and the rockfill is made of limestone gravel. Rockfill deformation sensors 104 and concrete face strain sensors 105 were installed to monitor the deformation and strain of the dam.

[0110] Based on the rockfill dam's geometric structure, material properties, and boundary conditions, a finite element simulation model 201 was established using ANSYS software. The finite element simulation model 201 includes 20,000 rockfill body calculation nodes and 500 concrete panel calculation nodes. The monitoring nodes include a rockfill body monitoring node 204 and a concrete panel monitoring node 206, whose locations correspond one-to-one to the actual locations of the rockfill body deformation sensor 104 and the concrete panel strain sensor 105, respectively. The virtual nodes include a rockfill body virtual node 205 and a concrete panel virtual node 207.

[0111] Monitoring data is extracted from the rockfill deformation sensor 104 and the concrete panel strain sensor 105 and recorded as a time series vector. Using a genetic algorithm, the material parameters in the finite element simulation model are adjusted to minimize the error between the simulation results and the actual monitoring data. After multiple iterations, a calibrated finite element simulation model 201 is ultimately obtained.

[0112] After parameter calibration, environmental loads (such as water pressure and temperature changes) from actual projects are applied to the finite element simulation model 201 to perform finite element simulation calculations. Deformation simulation results for the rockfill calculation nodes and strain simulation results for the concrete slab calculation nodes are extracted.

[0113] The actual monitoring data is matched with the simulation data, and a transfer function tensor is used to establish a mapping model between the monitoring data of the rockfill deformation sensor and the concrete panel virtual node 207. The transfer function tensor is solved by the least squares method to obtain the mapping model.

[0114] The monitoring data of the rockfill deformation sensor 104 is input, and the mapping model is used to calculate the strain virtual monitoring data of the concrete panel virtual node 207. The virtual monitoring data of the concrete panel strain is output as the monitoring data of the concrete panel strain virtual sensor 106.

[0115] The present invention achieves soft sensing of rockfill dam concrete panel strain by establishing a finite element simulation model 201 and combining it with actual monitoring data. This method can effectively reduce monitoring costs and improve monitoring efficiency, providing a new technical means for safety monitoring of rockfill dams.

[0116] Example 2:

[0117] An embodiment of the present invention further provides a soft sensing system for strain of a rockfill dam concrete panel, comprising:

[0118] A simulation and parameter calibration module is used to establish a finite element simulation model 201 based on an actual rockfill dam project, and to perform inverse calibration of material parameters in the finite element simulation model 201 using the actual collected rockfill deformation monitoring data and concrete panel strain monitoring data;

[0119] The environmental load simulation module is used to simulate the deformation of the rockfill body and the strain of the concrete panel under the real environmental load through the finite element simulation model 201 after parameter calibration, and obtain simulation result data;

[0120] a mapping model establishment module for establishing a mapping relationship between rockfill deformation and concrete panel strain using actually collected rockfill deformation monitoring data, concrete panel strain monitoring data, and simulation result data, and constructing a final mapping model based on the mapping relationship;

[0121] The virtual monitoring data calculation module is used to calculate the monitoring data of the concrete panel strain virtual sensor using the final mapping model. The actual collected rockfill deformation monitoring data includes the monitoring data of the rockfill deformation sensor. The actual collected concrete panel strain monitoring data includes the monitoring data of the concrete panel strain sensor ;

[0122] Y ̃ = [ y 1 ̃ y 2 ̃ ⋮ y p ̃ ] ; S ̃ = [ s 1 ̃ s 2 ̃ ⋮ s q ̃ ] ;

[0123] Where: Indicates the number The monitoring data of the rockfill deformation sensor, , Indicates the total number of rockfill deformation sensors; Indicates the number Monitoring data of concrete panel strain sensors, , represents the total number of concrete panel strain sensors; and are all time series vectors.

[0124] The above is only a preferred embodiment of the present invention. It should be pointed out that for virtual technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A soft sensing method for strain of rockfill dam concrete panels, characterized in that: include: Obtain monitoring data from the rockfill deformation sensor, input it into a pre-built mapping model, and calculate monitoring data from the concrete panel strain virtual sensor; The construction of the mapping model includes: Establish a finite element simulation model based on the actual rockfill dam project; Using the actual collected rockfill deformation monitoring data and concrete panel strain monitoring data, the material parameters in the finite element simulation model are inversely calibrated to obtain the finite element simulation model after parameter calibration; Using the finite element simulation model after the parameter calibration, the deformation of the rockfill body and the strain of the concrete panel under the real environmental load are simulated to obtain simulation result data; The mapping relationship between rockfill deformation and concrete panel strain is established by using the actual collected rockfill deformation monitoring data, concrete panel strain monitoring data and simulation result data. Based on the mapping relationship, a final mapping model is constructed.

2. The soft sensing method for rockfill dam concrete panel strain according to claim 1 is characterized in that: The actual collected rockfill deformation monitoring data includes monitoring data of the rockfill deformation sensor. The actual collected concrete panel strain monitoring data includes the monitoring data of the concrete panel strain sensor ; ； ； Where: Indicates the number The monitoring data of the rockfill deformation sensor, , Indicates the total number of rockfill deformation sensors; Indicates the number Monitoring data of concrete panel strain sensors, , represents the total number of concrete panel strain sensors; and are all time series vectors.

3. The soft sensing method for strain of rockfill dam concrete panels according to claim 2 is characterized in that: The finite element simulation model after parameter calibration is used to simulate the deformation of the rockfill body and the strain of the concrete panel under real environmental loads, and the simulation result data are obtained, including: The finite element simulation model after parameter calibration includes several rockfill calculation nodes and concrete panel calculation nodes, and calculates and extracts the deformation simulation result data of the rockfill calculation nodes. and the strain simulation results of the concrete panel calculation nodes ; ； ； Where: Indicates the number The deformation simulation results at the calculation nodes of the rockfill body are as follows: , Indicates the total number of rockfill calculation nodes; Indicates the number The strain simulation results at the calculation nodes of the concrete panel are as follows: , Indicates the total number of calculation nodes of the concrete panel; and are all time series vectors.

4. The soft sensing method for strain of rockfill dam concrete panels according to claim 3 is characterized in that: The computing nodes are divided into monitoring nodes and virtual nodes. The monitoring nodes include rockfill monitoring nodes and concrete panel monitoring nodes, and their positions correspond one-to-one to the actual positions of the rockfill deformation sensor and the concrete panel strain sensor respectively. The virtual nodes include rockfill virtual nodes and concrete panel virtual nodes. The deformation simulation result data and strain simulation result data of the rockfill monitoring node and the concrete panel monitoring node are recorded as and The deformation simulation result data and strain simulation result data of the virtual node of the rockfill body and the virtual node of the concrete panel are recorded as and , and is subject to the following relations: ； ； 。 5. The soft sensing method for strain of rockfill dam concrete panels according to claim 4 is characterized in that: The mapping relationship between rockfill deformation and concrete slab strain is established using the actual collected rockfill deformation monitoring data, concrete slab strain monitoring data, and simulation result data, including: According to the deformation simulation result data of the rockfill monitoring node and strain simulation results of concrete panel monitoring nodes , establish the mapping relationship between the rockfill monitoring node and the concrete panel monitoring node , as shown below, the mapping relationship By transferring the function tensor express: ； According to the deformation simulation result data of the rockfill monitoring node and strain simulation results of virtual nodes of concrete panels , establish the mapping relationship between the rockfill monitoring node and the concrete panel virtual node , as shown below, the mapping relationship By transferring the function tensor express: ； According to the monitoring data of the rockfill deformation sensor and monitoring data from concrete panel strain sensors , establish the actual monitoring data mapping relationship between the rockfill monitoring node and the concrete panel monitoring node , as shown below, the mapping relationship By transferring the function tensor express: ； ； ； Where: represents monitoring data provided by the concrete panel strain sensor at the concrete panel monitoring node; Indicates monitoring data provided by rockfill deformation sensors at rockfill monitoring nodes; and are all time series vectors.

6. The soft sensing method for strain of rockfill dam concrete panels according to claim 5 is characterized in that: Based on the mapping relationship, a final mapping model is constructed, including: Based on the mapping relationship and Building a mapping model , the mapping model It is used to combine the simulation result data with the actual monitoring data, as shown in the following formula: By transferring the function tensor express: ； ； ； Where: for The generalized inverse matrix of ; Based on the mapping relationship and mapping models and build the final mapping model , as shown below, the final mapping model By transferring the function tensor express: ； ； The monitoring data of the rockfill deformation sensor Input to the final mapping model , calculate the monitoring data of the concrete panel strain virtual sensor , as shown below: ； ； Where: Represents the monitoring data provided by the concrete panel strain virtual sensor at the virtual node of the concrete panel.

7. A soft sensing system for strain of rockfill dam concrete panels, characterized by: include: The simulation and parameter calibration module is used to establish a finite element simulation model based on the actual rockfill dam project and to perform inverse calibration of the material parameters in the finite element simulation model using the actual collected rockfill deformation monitoring data and concrete panel strain monitoring data; The environmental load simulation module is used to simulate the deformation of the rockfill body and the strain of the concrete panel under real environmental loads through the finite element simulation model after parameter calibration, and obtain simulation result data; a mapping model establishment module for establishing a mapping relationship between rockfill deformation and concrete panel strain using actually collected rockfill deformation monitoring data, concrete panel strain monitoring data, and simulation result data, and constructing a final mapping model based on the mapping relationship; The virtual monitoring data calculation module is used to calculate the monitoring data of the concrete panel strain virtual sensor using the final mapping model.

8. The rockfill dam concrete panel strain soft sensing system according to claim 7 is characterized in that: The actual collected rockfill deformation monitoring data includes monitoring data of the rockfill deformation sensor. The actual collected concrete panel strain monitoring data includes the monitoring data of the concrete panel strain sensor ; ； ； Where: Indicates the number The monitoring data of the rockfill deformation sensor, , Indicates the total number of rockfill deformation sensors; Indicates the number Monitoring data of concrete panel strain sensors, , represents the total number of concrete panel strain sensors; and are all time series vectors.

Images