A frozen soil area ground net soil environment measurement monitoring method, storage medium and system

Abstract

This invention provides a method, storage medium, and system for measuring and monitoring the soil environment of substation grounding grids in permafrost areas, relating to the field of substation grounding grid safety. The method includes: deploying local environmental monitoring units to collect soil environmental data in real time; acquiring a soil environmental parameter sample dataset; constructing an SSA-SVM model and iteratively training it using the soil environmental parameter sample dataset; training a BP neural network based on the prediction results of the SSA-SVM model to construct a combined prediction model for grounding grid corrosion rate; processing the real-time collected soil environmental data using the combined prediction model to obtain the predicted real-time corrosion rate; and generating an early warning signal based on the predicted real-time corrosion rate. The storage medium and system are used to implement the method. The beneficial effects of this invention are: obtaining the real-time corrosion rate based on real-time environmental parameters, achieving high prediction accuracy, and providing reliable measurement, monitoring, and early warning of the soil environment of substation grounding grids, especially in permafrost areas.

Description

Technical Field

[0001] This invention relates to the field of substation grounding grid safety, and in particular to a method, storage medium and system for measuring and monitoring the soil environment of grounding grids in permafrost areas. Background Technology

[0002] With the increasing promotion of high-voltage power transmission and transformation in my country, the performance of power transmission and transformation equipment is receiving more and more attention for the safe operation of the power grid. In extremely cold regions, water and salt in permafrost soil undergo two migration processes during freeze-thaw cycles. During freezing, water in the unfrozen layer continuously migrates to the frozen layer and freezes, while salt migrates along with the water. During thawing, the unmelted ice layer hinders the infiltration of water and salt, and the stagnant water in the upper layer migrates to the surface and evaporates, causing a sharp increase in surface salt content. This process accelerates the corrosion of the grounding grid in permafrost areas and has a serious impact on the safety and reliability of the grounding grid.

[0003] Soil parameters such as moisture content, salinity, and temperature are in a dynamic process, directly influenced by environmental factors. Variations in environmental parameters such as air temperature, precipitation, sunlight, and evaporation directly cause dynamic changes in soil parameters. In the vertical profile of a typical homogeneous soil, the influence of the external environment gradually weakens with increasing soil depth, and the range of variation in soil parameters gradually decreases. The resistivity of seasonally frozen soil is affected by soil moisture content and salinity. The freeze-thaw process of frozen soil causes changes in soil moisture migration, which in turn leads to changes in salinity.

[0004] Currently, most measurements of grounding grid environment in permafrost areas are conducted by directly measuring grounding resistance, but there is limited research on the changing patterns of other soil environmental factors such as moisture and salinity. Summary of the Invention

[0005] The purpose of this invention is to provide a method, storage medium, and system for measuring and monitoring the soil environment of grounding grids in permafrost areas, in order to monitor the impact of environmental parameters on the corrosion rate of grounding grids in real time. A method for measuring and monitoring the soil environment of grounding grids in permafrost areas mainly includes the following steps: S1. Deploy local environmental monitoring units to collect soil environmental data in real time; S2. Obtain a sample dataset of soil environmental parameters; S3. Construct the SSA-SVM model and iteratively train it using a soil environmental parameter sample dataset; S4. Train a BP neural network based on the prediction results of the SSA-SVM model to construct a combined prediction model for the corrosion rate of the grounding grid. S5. Use the grounding grid corrosion rate combined prediction model to process the real-time collected soil environmental data and obtain the predicted real-time corrosion rate. S6. Generate an early warning signal based on the predicted real-time corrosion rate.

[0006] Furthermore, the local environmental monitoring unit includes soil parameter sensors and measurement channels. The soil parameter sensors are deployed in the measurement channels to collect soil environmental data in real time. Soil parameter sensors include soil moisture sensors, soil salinity sensors, soil resistivity sensors, and soil temperature sensors.

[0007] Preferably, the type and number of soil parameter sensors are dynamically changed according to different permafrost environments.

[0008] Furthermore, the SSA-SVM model is an SVM model optimized from the SSA algorithm, and the training process is as follows: S31. The sample data in the soil environmental parameter sample dataset are organized into feature vectors according to the categories of soil environmental parameters. Each dimension of the vector is normalized separately. Then, the dataset is divided into training and test sets according to the sample size. The mathematical formula for normalization is:

[0009] In the formula , These are the original and normalized values ​​of the soil environmental parameter characteristic value n, respectively. These represent the maximum and minimum values ​​of the soil environmental parameter characteristic value n, respectively. S32. Use the root mean square error of the SVM model output as the fitness function of the SSA algorithm. Initialize the parameters of the SSA algorithm, including the sparrow population size, the proportion of scout sparrows and the proportion of head sparrows, and the search range. Perform two iterations of training with the penalty coefficient and kernel coefficient as the position of individual sparrows, respectively. Output the penalty coefficient and kernel coefficient after training. S33. Assign the output penalty coefficient and kernel coefficient to the SVM model to obtain the SSA-SVM model.

[0010] Furthermore, step S32 specifically includes: S321. Initialize the sparrow population, calculate the fitness function of individual sparrows, and select the sparrows with smaller fitness values ​​as leader sparrows and the rest as followers. S322. Update the position of the leader and followers. The calculation formula is as follows: The formula for updating the position of the leader is:

[0011] In the formula, i represents the sparrow's serial number, and j represents the dimension in the soil environmental parameter feature vector, with each dimension corresponding to a different type of soil environmental parameter; and Let be the positions of sparrow i in the j-th dimension at the (h+1)-th and h-th iterations, respectively; θ is a random number with a value range of [0,1]; Q is a random number following a standard normal distribution; I is a unit row matrix; S T R1 is the safety value; R2 is the warning value. The follower position update formula is:

[0012] In the formula, B is the total population capacity of sparrows; This is the position with the worst fitness in the h-th iteration; Let A be the optimal predator position in the (h+1)th iteration; A is a row matrix with elements taking values ​​of 1 or -1. -1 ; S323. Select a group with higher fitness from the sparrow population after the position is updated as scout sparrows, update the position of the scout sparrows, compare the fitness value obtained in this iteration with the optimal fitness value, and update the optimal fitness value. The formula for updating the scout bird's position is as follows:

[0013] Where ζ is the step size correction parameter; r is a random number with a value range of (0,1); This is the position with the best fitness in the h-th iteration; Let i be the position of sparrow i; This is the best hunting spot; This is the worst predation spot; S324. Repeat steps S322-S323 until the iteration termination condition is met, and output the optimal values ​​of the penalty coefficient and kernel coefficient.

[0014] Furthermore, the fitness function of the SSA algorithm is:

[0015] In the formula, f represents the fitness function; The actual corrosion rate of the grounding grid k was obtained from laboratory environmental simulation based on soil environmental parameter sample data. Predict the corrosion rate of the grounding grid k obtained from the SVM model.

[0016] Furthermore, the specific steps for training a BP neural network are as follows: S41. Input the data from the soil environmental parameter sample dataset into the SSA-SVM model to obtain the predicted corrosion rate; S42. Calculate the deviation between the predicted corrosion rate and the actual corrosion rate obtained in step S41. S43. Set the input of the BP neural network to soil environmental parameters and the output to the deviation between the corrosion rate predicted by the SSA-SVM model and the actual corrosion rate. Input the soil environmental parameter sample dataset and the deviation calculated in step S42 into the BP neural network for training.

[0017] Furthermore, the combined prediction model for grounding grid corrosion rate is composed of a trained BP neural network and an SSA-SVM model. The BP neural network corrects the output of the SSA-SVM model to obtain the predicted grounding grid corrosion rate. The calculation formula is as follows:

[0018] In the formula, The corrosion rate is predicted by the combined model. The corrosion rate is predicted by the SSA-SVM model. This represents the error output value of the BP neural network.

[0019] Furthermore, the early warning signals include real-time corrosion early warning signals and cumulative corrosion early warning signals; When the predicted real-time corrosion rate exceeds the preset corrosion rate threshold, a real-time corrosion early warning signal is generated. The system records and summarizes the predicted real-time corrosion rate, forms a time-integrated real-time corrosion rate curve, and performs integration calculations. When the integrated value exceeds the preset corrosion degree threshold, a cumulative corrosion warning signal is generated.

[0020] A storage medium: The storage medium stores instructions and data to implement a ground network soil environment measurement and monitoring method for permafrost regions.

[0021] A ground-grid soil environment measurement and monitoring system for permafrost regions includes a local environmental monitoring unit and an environmental analysis unit, which communicate with each other via GPRS. The local environmental monitoring unit includes soil parameter sensors and measurement channels for real-time acquisition of soil environmental data; The environmental analysis unit includes a processor and a storage medium. The processor loads and executes instructions and data in the storage medium, obtains and predicts the real-time corrosion rate based on real-time collected soil environmental data, and generates an early warning signal.

[0022] The beneficial effects of the technical solution provided by this invention are as follows: This invention trains a combination of SSA-SVM model and BP neural network using soil environmental parameter sample data. The combined prediction model obtains the real-time corrosion rate based on real-time environmental parameters, which has a high prediction accuracy and provides reliable measurement, monitoring and early warning of soil environment in substation grounding grids, especially in permafrost areas. Attached Figure Description

[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 This is a flowchart of a soil environment measurement and monitoring method using a ground grid in a permafrost region, as described in an embodiment of the present invention. Figure 2 This is a schematic diagram of the local environment detection unit in an embodiment of the present invention; Figure 3 This is a schematic diagram of corrosion early warning signal determination in an embodiment of the present invention; Figure 4 This is a schematic diagram of a ground network soil environment measurement and monitoring system in a frozen soil area according to an embodiment of the present invention. Detailed Implementation

[0024] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0025] The embodiments of the present invention provide a method, storage medium and system for measuring and monitoring soil environment in permafrost areas using a ground grid.

[0026] Please refer to Figure 1 , Figure 1 This is a flowchart of a soil environment measurement and monitoring method using a ground grid in a permafrost region, as described in an embodiment of the present invention. The method specifically includes the following steps: The first step is to deploy local environmental monitoring units to collect soil environmental data in real time.

[0027] The local environmental monitoring unit includes soil parameter sensors and measurement channels. The soil parameter sensors are deployed in the measurement channels to collect soil environmental data in real time.

[0028] like Figure 2 As shown, the types of soil parameter sensors include soil moisture sensors, soil salinity sensors, soil resistivity sensors, and soil temperature sensors.

[0029] Preferably, the type and number of soil parameter sensors are dynamically changed according to different permafrost environments.

[0030] (1) The function of the soil moisture sensor is to monitor the moisture content of the frozen soil environment. The Meter Group's Teros 12 sensor is used, and its measurement range is: mineral soil calibration: 0.00~0.62m 3 / m 3 Non-soil medium calibration: 0.0~1.0m 3 / m 3 Resolution 0.001m 3 / m 3 ; (2) The function of the soil salinity sensor is to monitor the salinity of the permafrost environment. The Pico-Profile sensor manufactured by Imko is used. The moisture measurement range is 0~100% volumetric water content; the moisture accuracy is ±2%; and the repeatability is <0.1%. (3) The function of the resistivity sensor is to monitor the soil resistivity in the permafrost environment. The Stevens Hydra Probe II soil conductivity sensor is used. The range and accuracy are as follows: dielectric constant: 1 to 78, air = 1, distilled water = 78; ±1.5% or 0.2 (whichever is greater); soil moisture: dry to saturated; ±0.03; soil conductivity: 0.01 to 1.5 S / m; ±2.0% or 0.005 S / m (whichever is greater). (4) The function of the soil temperature sensor is to monitor the soil temperature in the frozen soil environment. It uses fiber Bragg grating for monitoring, with a monitoring range of -50℃ to 70℃.

[0031] The second step is to obtain a sample dataset of soil environmental parameters.

[0032] The third step is to construct an SSA-SVM model and iteratively train it using a soil environmental parameter sample dataset.

[0033] The SSA-SVM model is an SVM model optimized from the SSA algorithm. The training process is as follows: Step 1: The sample data in the soil environmental parameter sample dataset are organized into feature vectors according to the categories of soil environmental parameters. Each dimension of the vector is normalized. Then, the dataset is divided into training and test sets according to the sample size. The normalization mathematical formula is:

[0034] In the formula , These are the original and normalized values ​​of the soil environmental parameter characteristic value n, respectively. These represent the maximum and minimum values ​​of the soil environmental parameter characteristic value n, respectively. Step 2: Use the root mean square error of the SVM model output as the fitness function of the SSA algorithm, initialize the parameters of the SSA algorithm, including the sparrow population size, the proportion of scout sparrows and the proportion of head sparrows, the search range of the penalty coefficient and the kernel coefficient, and output the penalty coefficient and kernel coefficient after iterative training.

[0035] The fitness function of the SSA algorithm is:

[0036] In the formula, f represents the fitness function; The actual corrosion rate of the grounding grid k was obtained from laboratory environmental simulation based on soil environmental parameter sample data. Predict the corrosion rate of the grounding grid k obtained from the SVM model.

[0037] In this embodiment, the maximum number of iterations is set to 200, the population size is 20, the proportion of head sparrows is 0.2, the proportion of scout sparrows is 0.1, and the search range is set to [0,100]. The penalty coefficient C and kernel coefficient γ of the SVM model are used as the position of individual sparrows for two iterations of training. After training, the penalty coefficient and kernel coefficient are output.

[0038] The specific steps of iterative training are as follows: Step 1: Initialize the sparrow population, calculate the fitness function of individual sparrows, and select the sparrows with smaller fitness values ​​as leader sparrows, and the rest as followers.

[0039] Step 2, update the position of the leader and followers, calculated using the following formula: The formula for updating the position of the leader is:

[0040] In the formula, i represents the sparrow's serial number, and j represents the dimension in the soil environmental parameter feature vector, with each dimension corresponding to a different type of soil environmental parameter; and Let be the positions of sparrow i in the j-th dimension at the (h+1)-th and h-th iterations, respectively; θ is a random number with a value range of [0,1]; Q is a random number following a standard normal distribution; I is a unit row matrix; S T R1 is the safety value; R2 is the warning value.

[0041] Sparrows are threatened by predators such as snakes, cats, and eagles in the wild. ST and R2 indicate whether the predation area is safe. If R2 T , indicating that the predation area is safe; conversely, R2≥S T When the larvae indicate that the hunting area is unsafe, the larvae will lead its followers to other locations and update their positions.

[0042] The follower position update formula is:

[0043] In the formula, B is the total population capacity of sparrows; This is the position with the worst fitness in the h-th iteration; Let A be the optimal predator position in the (h+1)th iteration; A is a row matrix with elements taking values ​​of 1 or -1. -1 When i > B / 2, the follower has poor fitness, cannot obtain food, and needs to update its position.

[0044] ​Step 3: Select a group with higher fitness from the sparrow population after the position is updated as scout sparrows, update the position of the scout sparrows, compare the fitness value obtained in this iteration with the optimal fitness value, and update the optimal fitness value.

[0045] It should be noted that sparrows are highly vigilant in order to adapt to their environment. They will notify the entire population to evacuate quickly once they detect any abnormalities. Therefore, the SSA algorithm includes a certain number of scout sparrows to ensure timely detection of anomalies. The formula for updating the scout sparrows' positions is as follows:

[0046] Where ζ is the step size correction parameter; r is a random number with a value range of (0,1); This is the position with the best fitness in the h-th iteration; Let i be the position of sparrow i; This is the best hunting spot; This is the worst predation point; when > When this occurs, it indicates that sparrows have poor adaptability, live far from food, and are vulnerable to predators; when... At this point, the sparrow's fitness level is at its lowest, indicating it is very close to predators and needs to retreat immediately to a new location. During this process, the fitness level iterates along with the sparrow's location.

[0047] Step 4: Use the penalty coefficient and kernel coefficient as the iteration objects of the SSA algorithm, repeat steps 2-3 until the iteration termination condition is met, and output the optimal values ​​of the penalty coefficient and kernel coefficient.

[0048] Step 3: Assign the output penalty coefficient C and kernel coefficient γ to the SVM model to obtain the SSA-SVM model.

[0049] The fourth step is to train a BP neural network based on the prediction results of the SSA-SVM model to construct a combined prediction model for the corrosion rate of the grounding grid.

[0050] The combined prediction model for grounding grid corrosion rate consists of a trained BP neural network and an SSA-SVM model. The BP neural network corrects the output of the SSA-SVM model to obtain the predicted grounding grid corrosion rate. The calculation formula is as follows:

[0051] In the formula, The corrosion rate is predicted by the combined model. The corrosion rate is predicted by the SSA-SVM model. This represents the error output value of the BP neural network.

[0052] The specific steps for training a BP neural network are as follows: Step 1: Input the data from the soil environmental parameter sample dataset into the SSA-SVM model to obtain the predicted corrosion rate; Step 2: Calculate the deviation between the predicted corrosion rate and the actual corrosion rate obtained in step S41. Step 3: Set the input of the BP neural network to soil environmental parameters, and the output to the deviation between the corrosion rate predicted by the SSA-SVM model and the actual corrosion rate. The deviation between the predicted and actual values ​​of the SSA-SVM corrosion rate prediction model is then used as the basis for this step. As output, it forms the new training samples. The data is then input into a BP neural network for training.

[0053] The fifth step involves using a combined prediction model of grounding grid corrosion rate to process the real-time collected soil environmental data and obtain the predicted real-time corrosion rate.

[0054] The sixth step is to generate an early warning signal based on the predicted real-time corrosion rate.

[0055] Early warning signals include real-time corrosion early warning signals and cumulative corrosion early warning signals, such as... Figure 3 As shown: When the predicted real-time corrosion rate exceeds the preset corrosion rate threshold, a real-time corrosion early warning signal is generated. The system records and summarizes the predicted real-time corrosion rate, forms a time-integrated real-time corrosion rate curve, and performs integration calculations. When the integrated value exceeds the preset corrosion degree threshold, a cumulative corrosion warning signal is generated.

[0056] Please refer to Figure 4 , Figure 4 This is a schematic diagram of a ground network soil environment measurement and monitoring system in a frozen soil area according to an embodiment of the present invention.

[0057] A storage medium 404: The storage medium 404 stores instructions and data for implementing a ground network soil environment measurement and monitoring method in permafrost areas.

[0058] A ground-grid soil environment measurement and monitoring system for permafrost regions includes a local environmental monitoring unit 401 and an environmental analysis unit 402, which communicate via GPRS. The environmental data collected by the local environmental monitoring unit is returned to the remote analysis and monitoring unit via GPRS and processed into a data format that can be input into the local environmental monitoring unit.

[0059] The local environmental monitoring unit 401 includes soil parameter sensors and measurement channels for real-time acquisition of soil environmental data; The environmental analysis unit 402 includes a processor 403 and a storage medium 404; the processor loads and executes the instructions and data in the storage medium 404, obtains the predicted real-time corrosion rate based on the real-time collected soil environmental data, and generates an early warning signal.

[0060] The beneficial effects of this invention are: This invention trains a combination of SSA-SVM model and BP neural network using soil environmental parameter sample data. The combined prediction model obtains the real-time corrosion rate based on real-time environmental parameters, which has a high prediction accuracy and provides reliable measurement, monitoring and early warning of soil environment in substation grounding grids, especially in permafrost areas.

[0061] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for measuring and monitoring the soil environment in permafrost regions using a ground-grid system, characterized in that, The specific steps include: S1. Deploy local environmental monitoring units to collect soil environmental data in real time; S2. Obtain a sample dataset of soil environmental parameters; S3. Construct the SSA-SVM model and iteratively train it using a soil environmental parameter sample dataset; S4. Train a BP neural network based on the prediction results of the SSA-SVM model to construct a combined prediction model for the corrosion rate of the grounding grid. S5. Use the grounding grid corrosion rate combined prediction model to process the real-time collected soil environmental data and obtain the predicted real-time corrosion rate. S6. Generate an early warning signal based on the predicted real-time corrosion rate.

2. The ground-grid soil environment measurement and monitoring system for permafrost regions as described in claim 1, characterized in that, The local environment detection unit includes a soil parameter sensor and a measurement channel. The soil parameter sensor is deployed in the measurement channel to collect soil environmental data in real time. Soil parameter sensors include soil moisture sensors, soil salinity sensors, soil resistivity sensors, and soil temperature sensors.

3. The method for measuring and monitoring soil environment in permafrost areas using a ground grid as described in claim 2, characterized in that, The SSA-SVM model is an SVM model optimized using the SSA algorithm. The training process is as follows: S31. The sample data in the soil environmental parameter sample dataset are organized into feature vectors according to the categories of soil environmental parameters. Each dimension of the vector is normalized separately. Then, the dataset is divided into training and test sets according to the sample size. The mathematical formula for normalization is: In the formula , These are the original and normalized values ​​of the soil environmental parameter characteristic value n, respectively. These represent the maximum and minimum values ​​of the soil environmental parameter characteristic value n, respectively. S32. Use the root mean square error of the SVM model output as the fitness function of the SSA algorithm. Initialize the parameters of the SSA algorithm, including the sparrow population size, the proportion of scout sparrows and the proportion of head sparrows, and the search range. Perform two iterations of training with the penalty coefficient and kernel coefficient as the position of individual sparrows, respectively. Output the penalty coefficient and kernel coefficient after training. S33. Assign the output penalty coefficient and kernel coefficient to the SVM model to obtain the SSA-SVM model.

4. The method for measuring and monitoring soil environment in permafrost areas using a ground grid as described in claim 3, characterized in that, Step S32 is as follows: S321. Initialize the sparrow population, calculate the fitness function of individual sparrows, and select the sparrows with smaller fitness values ​​as leader sparrows and the rest as followers. S322. Update the position of the leader and followers. The calculation formula is as follows: The formula for updating the position of the leader is: In the formula, i represents the sparrow's serial number, and j represents the dimension in the soil environmental parameter feature vector, with each dimension corresponding to a different type of soil environmental parameter; and Let be the positions of sparrow i in the j-th dimension at the (h+1)-th and h-th iterations, respectively; θ is a random number with a value range of [0,1]; Q is a random number following a standard normal distribution; I is a unit row matrix; S T R1 is the safety value; R2 is the warning value. The follower position update formula is: In the formula, B is the total population capacity of sparrows; This is the position with the worst fitness in the h-th iteration; Let A be the optimal predator position in the (h+1)th iteration; A is a row matrix with elements taking values ​​of 1 or -1. -1 ; S323. Select a group with higher fitness from the sparrow population after the position is updated as scout sparrows, update the position of the scout sparrows, compare the fitness value obtained in this iteration with the optimal fitness value, and update the optimal fitness value. The formula for updating the scout bird's position is as follows: Where ζ is the step size correction parameter; r is a random number with a value range of (0,1); This is the position with the best fitness in the h-th iteration; Let i be the position of sparrow i; This is the best hunting spot; This is the worst predation spot; S324. Repeat steps S322-S323 until the iteration termination condition is met, and output the optimal values ​​of the penalty coefficient and kernel coefficient.

5. The method for measuring and monitoring soil environment in permafrost areas using a ground grid as described in claim 4, characterized in that, The fitness function of the SSA algorithm is: In the formula, f represents the fitness function; The actual corrosion rate of the grounding grid k was obtained from laboratory environmental simulation based on soil environmental parameter sample data. Predict the corrosion rate of the grounding grid k obtained from the SVM model.

6. The method for measuring and monitoring soil environment in permafrost areas using a ground grid as described in claim 5, characterized in that, The specific steps for training the BP neural network are as follows: S41. Input the data from the soil environmental parameter sample dataset into the SSA-SVM model to obtain the predicted corrosion rate; S42. Calculate the deviation between the predicted corrosion rate and the actual corrosion rate obtained in step S41. S43. Set the input of the BP neural network to soil environmental parameters and the output to the deviation between the corrosion rate predicted by the SSA-SVM model and the actual corrosion rate. Input the soil environmental parameter sample dataset and the deviation calculated in step S42 into the BP neural network for training.

7. The method for measuring and monitoring soil environment in permafrost areas using a ground grid as described in claim 6, characterized in that... The grounding grid corrosion rate prediction model is composed of a trained BP neural network and an SSA-SVM model. The BP neural network corrects the output of the SSA-SVM model to obtain the predicted grounding grid corrosion rate. The calculation formula is as follows: In the formula, The corrosion rate is predicted by the combined model. The corrosion rate is predicted by the SSA-SVM model. This represents the error output value of the BP neural network.

8. The method for measuring and monitoring soil environment in permafrost areas using a ground grid as described in claim 7, characterized in that, The warning signals include real-time corrosion warning signals and cumulative corrosion warning signals; When the predicted real-time corrosion rate exceeds the preset corrosion rate threshold, a real-time corrosion early warning signal is generated. The system records and summarizes the predicted real-time corrosion rate, forms a time-integrated real-time corrosion rate curve, and performs integration calculations. When the integrated value exceeds the preset corrosion degree threshold, a cumulative corrosion warning signal is generated.

9. A storage medium: the storage medium stores instructions and data for implementing the ground network soil environment measurement and monitoring method for permafrost areas as described in any one of claims 1 to 8.

10. A ground grid soil environment measurement and monitoring system for permafrost regions, used to implement the ground grid soil environment measurement and monitoring method for permafrost regions as described in any one of claims 1 to 8, characterized in that, It includes a local environmental monitoring unit and an environmental analysis unit, which communicate with each other via GPRS; The local environmental monitoring unit is used to collect soil environmental data in real time; The environmental analysis unit includes a processor and a storage medium; the processor loads and executes instructions and data in the storage medium, obtains a predicted real-time corrosion rate based on real-time collected soil environmental data, and generates an early warning signal.

Images