Substation settlement monitoring method and system

Abstract

The invention provides a substation settlement monitoring method and system, and relates to the technical field of settlement monitoring, and the method comprises the steps: deploying a bionic root system mesh array sensor network which comprises a plurality of monitoring nodes; collecting current phase data of a transformer substation bus in real time to calculate magnetic field interference, and performing inverse compensation on an original displacement signal of the magnetostrictive displacement sensor; carrying out vibration monitoring through a triaxial accelerometer, and carrying out data weight adjustment on a plurality of monitoring nodes; and performing settlement prediction on the transformer substation according to the displacement compensation data in combination with the weight parameters of the plurality of nodes, performing backtracking calibration on the bionic root system mesh array sensor network, and performing closed-loop monitoring on the settlement of the transformer substation according to a calibration result. The technical problem of inaccurate settlement monitoring caused by interference in a high-intensity magnetic field environment of the transformer substation in the prior art is solved, and accurate prediction and closed-loop monitoring of settlement of the transformer substation are realized by deploying a plurality of intelligent sensors and performing magnetic field interference compensation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of settlement monitoring, and particularly relates to a substation settlement monitoring method and system. BACKGROUND

[0002] The substation usually has strong electromagnetic interference, especially around the high-voltage bus and cable, and the changes of electric field and magnetic field have a significant impact on the sensor. Strong electromagnetic interference will be coupled into the sensor signal, resulting in serious distortion and distortion of the output signal, making it extremely difficult to accurately measure the tiny settlement displacement, and ultimately significantly reducing the reliability of the monitoring data and directly affecting the effectiveness of the settlement trend prediction and early warning system. Although some methods try to reduce interference by hardware shielding or improving the sensor itself, such as choosing optical fiber sensors and other immune devices, although the interference can be avoided, the cost is high and there are still problems of signal distortion and inaccurate data, which further affect the accuracy of the settlement monitoring data.

[0003] In summary, the prior art has the technical problem that the strong magnetic field environment of the substation will cause interference, resulting in serious distortion of the measurement signal, which further affects the accuracy of the settlement monitoring data. SUMMARY

[0004] The purpose of the present application is to provide a substation settlement monitoring method and system to solve the technical problem in the prior art that the strong magnetic field environment of the substation will cause interference, resulting in serious distortion of the measurement signal, which further affects the accuracy of the settlement monitoring data.

[0005] In view of the above problems, the present application provides a substation settlement monitoring method and system.

[0006] In a first aspect, the present application provides a substation settlement monitoring method, which is realized by a substation settlement monitoring system, wherein the substation settlement monitoring method comprises: deploying a bionic root system network, the bionic root system network comprises a plurality of monitoring nodes, each monitoring node comprises a magnetostrictive displacement sensor and a three-axis accelerometer; calculating the magnetic field interference through the bionic root system network in real time to collect the current phase data of the substation bus, compensating the original displacement signal of the magnetostrictive displacement sensor according to the dynamic magnetic field interference distribution data to obtain displacement compensation data; monitoring the vibration of the plurality of monitoring nodes through the three-axis accelerometer, adjusting the data weight of the plurality of monitoring nodes, and generating a plurality of node weight parameters; predicting the settlement of the substation according to the displacement compensation data combined with the plurality of node weight parameters, calibrating the bionic root system network based on the prediction result, and monitoring the settlement of the substation in a closed loop according to the calibration result.

[0007] Optionally, the foundation structure characteristics and load distribution characteristics of the transformer substation are introduced to construct a bionic root system topology; sensor layout analysis is performed on a plurality of monitoring nodes according to the bionic root system topology to formulate a layout scheme; sensors are laid out in a settlement area of the transformer substation according to the layout scheme to generate a bionic root system mesh array; magnetostrictive displacement sensors and three-axis accelerometers are configured based on the bionic root system mesh array to traverse a plurality of monitoring nodes to construct integrated sensing units; the integrated sensing units are configured for communication to construct a data transmission link, and the integrated sensing units are integrated with the bionic root system mesh array according to the data transmission link to construct a bionic root system mesh array sensor network.

[0008] Optionally, the bionic root system mesh array sensor network collects current phase data of the transformer substation bus at multiple time periods to construct a current-magnetic field space-time correlation database; instantaneous magnetic field direction data and magnetic field intensity distribution data are obtained by analyzing the current-magnetic field space-time correlation database; the spatial three-dimensional layout parameters of the transformer substation are called to map the instantaneous magnetic field direction data and the magnetic field intensity distribution data to the spatial three-dimensional layout parameters for magnetic field interference simulation calculation to construct a magnetic field interference distribution field map; the original displacement signal is inversely compensated based on the magnetic field interference distribution field map to obtain displacement compensation data.

[0009] Optionally, the original displacement signal is matched with the magnetic field interference distribution field map in space-time to generate a space-time matching result; the space-time matching result is traced back to a plurality of monitoring nodes for interference identification to obtain interference characteristic quantities; a magnetic field-displacement mapping relationship curve is calibrated, and the interference characteristic quantities are mapped to the magnetic field-displacement mapping relationship curve for inverse compensation operation on the original displacement signal to generate the displacement compensation data.

[0010] Optionally, the interference characteristic quantities are used as indexes to traverse the magnetic field-displacement mapping relationship curve for matching query to generate displacement supplement quantities; vector superposition operation is performed on the original displacement signal according to the displacement supplement quantities to generate reverse compensation quantities; the original displacement signal is subjected to the reverse compensation quantities in the interference direction to generate initial compensation displacement data; filtering calculation is performed based on the initial compensation displacement data to obtain the displacement compensation data.

[0011] Optionally, the three-axis accelerometer is used to monitor the real-time vibration state of a plurality of monitoring nodes to collect vibration acceleration time series data in three-axis directions; vibration analysis is performed based on the vibration acceleration time series data in three-axis directions to construct vibration state parameters; the sensors are subjected to reliability evaluation based on the vibration state parameters to obtain reliability scores; the weight coefficients of a plurality of monitoring nodes are dynamically adjusted according to the reliability scores to generate node weight parameters.

[0012] Optionally, Fourier transform is performed on the vibration acceleration time series data in the three axes, and the dominant frequency component and harmonic features are extracted based on the transformation results; the root mean square value of the vibration acceleration time series data in the three axes is calculated, and a vibration amplitude quantization index is set; power spectral density analysis is performed on the dominant frequency component and the harmonic features according to the vibration amplitude quantization index to determine the frequency distribution characteristics; frequency domain energy concentration analysis is performed based on the frequency distribution characteristics to construct the vibration state parameters.

[0013] Optionally, vibration data from historical normal operating conditions are retrieved to establish a benchmark vibration feature library; the benchmark vibration feature library is trained and classified to determine normal vibration modes and abnormal vibration modes; fuzzy inference is performed based on the normal vibration modes and the abnormal vibration modes, and the sensor is evaluated in multiple dimensions according to the inference results, setting multiple confidence intervals; the vibration state parameters are matched with the multiple confidence intervals to generate the confidence score.

[0014] Optionally, the displacement compensation data and the weight parameters of the multiple nodes are spatiotemporally aligned and fused to construct a weighted displacement dataset; based on the weighted displacement dataset, settlement prediction is performed according to the settlement time sequence to generate a time-series settlement prediction result; the actual settlement monitoring value is retrieved through the biomimetic root mesh array sensor network, and the time-series settlement prediction result is sent back to the biomimetic root mesh array sensor network for deviation comparison with the actual settlement monitoring value. When the deviation value is greater than a preset deviation threshold, a backtracking calibration command is triggered; the sensing parameters of the biomimetic root mesh array sensor network are dynamically adjusted through the backtracking calibration command to construct a settlement monitoring closed-loop feedback mechanism.

[0015] Secondly, this application also provides a substation settlement monitoring system for performing a substation settlement monitoring method as described in the first aspect. The substation settlement monitoring system includes: a sensor deployment module for deploying a biomimetic root-like mesh array sensor network, the biomimetic root-like mesh array sensor network including multiple monitoring nodes, each monitoring node including a magnetostrictive displacement sensor and a triaxial accelerometer; an interference compensation module for real-time acquisition of current phase data of the substation busbar through the biomimetic root-like mesh array sensor network to calculate magnetic field interference, and inversely compensating the original displacement signal of the magnetostrictive displacement sensor based on dynamic magnetic field interference distribution data to obtain displacement compensation data; a weight adjustment module for vibration monitoring of multiple monitoring nodes through the triaxial accelerometer, adjusting the data weights of multiple monitoring nodes, and generating multiple node weight parameters; and a settlement prediction module for predicting substation settlement based on the displacement compensation data and the multiple node weight parameters, performing retrospective calibration of the biomimetic root-like mesh array sensor network based on the prediction results, and performing closed-loop monitoring of substation settlement according to the calibration results.

[0016] One or more technical solutions provided in this application have at least the following beneficial effects: By deploying a biomimetic root-like mesh array sensor network, which includes multiple monitoring nodes, each node comprising a magnetostrictive displacement sensor and a triaxial accelerometer, the network collects real-time current phase data of the substation busbars to calculate magnetic field interference. Based on the dynamic magnetic field interference distribution data, the original displacement signal from the magnetostrictive displacement sensor is inversely compensated to obtain displacement compensation data. Vibration monitoring is performed on the multiple monitoring nodes using the triaxial accelerometer, and data weighting is adjusted to generate multiple node weight parameters. Settlement prediction of the substation is performed based on the displacement compensation data and the multiple node weight parameters. The biomimetic root-like mesh array sensor network is then retrospectively calibrated based on the prediction results, and closed-loop monitoring of substation settlement is conducted according to the calibration results. In other words, by deploying multiple smart sensors to construct a biomimetic root-like mesh array sensor network, the current phase data of the substation busbar is collected in real time and magnetic field interference is compensated. Vibration monitoring of multiple monitoring nodes is carried out through triaxial accelerometers, the node data weights are adjusted, settlement is predicted based on displacement compensation data and node weight parameters, and the sensor network is backtracked and calibrated based on the prediction results, so as to achieve accurate prediction and closed-loop monitoring of substation settlement and improve the accuracy of settlement monitoring.

[0017] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating a substation settlement monitoring method according to this application.

[0020] Figure 2 This is a schematic diagram of the structure of a substation settlement monitoring system according to this application.

[0021] Figure labeling: Sensor deployment module 11, interference compensation module 12, weight adjustment module 13, settlement prediction module 14. Detailed Implementation

[0022] This application provides a substation settlement monitoring method and system, solving the technical problem in existing technologies where interference from the strong magnetic field environment of substations leads to severe distortion of measurement signals, further affecting the accuracy of settlement monitoring data. By deploying multiple intelligent sensors to construct a biomimetic root-like mesh array sensor network, real-time current phase data of the substation busbars is collected and magnetic field interference is compensated. Vibration monitoring of multiple monitoring nodes is performed using triaxial accelerometers, node data weights are adjusted, and settlement prediction is made based on displacement compensation data and node weight parameters. The sensor network is then retrospectively calibrated based on the prediction results, achieving accurate prediction and closed-loop monitoring of substation settlement, thus improving the accuracy of settlement monitoring.

[0023] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0024] Example 1, please refer to the appendix. Figure 1 This application provides a substation settlement monitoring method, wherein the substation settlement monitoring method is executed by a substation settlement monitoring system, and the substation settlement monitoring method specifically includes the following steps: A biomimetic root mesh array sensor network is deployed, which includes multiple monitoring nodes, each of which contains a magnetostrictive displacement sensor and a triaxial accelerometer.

[0025] Furthermore, this application also includes the following steps: incorporating the foundation structure characteristics and load distribution characteristics of the substation to construct a biomimetic root system topology; analyzing the sensor deployment of multiple monitoring nodes based on the biomimetic root system topology and formulating a deployment scheme; deploying sensors in the settlement area of ​​the substation according to the deployment scheme to generate a biomimetic root system mesh array; configuring magnetostrictive displacement sensors and triaxial accelerometers across multiple monitoring nodes based on the biomimetic root system mesh array to construct an integrated sensing unit; configuring the integrated sensing unit for communication to construct a data transmission link; and integrating the integrated sensing unit with the biomimetic root system mesh array according to the data transmission link to construct the biomimetic root system mesh array sensor network.

[0026] Specifically, the foundation structural characteristics of a substation refer to the engineering geological conditions of the substation foundation, including factors such as the bearing capacity, stability, and soil type, including parameters such as soil layer distribution, thickness, density, bearing capacity, and compression modulus. For example, the substation foundation might consist of 2.5m thick plain fill, 5m thick silty clay, and 10m thick silty clay, with stable bedrock underneath. Load distribution characteristics refer to the load distribution within and around the substation, such as the impact of the weight of the electrical equipment and surrounding traffic loads on the foundation. For example, a 300 MVA main transformer and its foundation weigh approximately 350 tons, generating a base pressure of approximately 120 kPa.

[0027] By analyzing the foundation stability and load effects of the substation, a biomimetic root system topology was constructed to simulate the distribution characteristics of plant roots. This allows for an adaptive sensor layout under conditions of uneven foundation settlement or varying loads in different areas, thereby improving the stability and flexibility of the monitoring network. The biomimetic root system topology plans the sensor network layout by mimicking the distribution morphology of tree roots. Its key feature is the deployment of denser sensor nodes in areas with heavy loads and weak geological conditions, such as the main root area and areas where lateral roots seek support, forming a three-dimensional monitoring network capable of sensitively sensing changes in foundation stress and strain. In other words, in critical areas with heavy loads or weak geology, such as below the main transformer foundation or in GIS areas, the behavior of the main root is simulated, planning more and denser nodes in the vertical depth to capture deep soil compression; in general areas with lighter loads but requiring monitoring of overall stability, the behavior of lateral roots is simulated, planning nodes with a wider horizontal range but slightly lower density to form a large sensing surface.

[0028] Based on a biomimetic root system topology, sensors are deployed at multiple monitoring nodes, and a reasonable monitoring node deployment scheme is developed. Considering the actual load distribution of the substation, appropriate sensor locations and quantities are selected to ensure comprehensive monitoring of settlement and displacement changes both inside and outside the substation. Developing the deployment scheme involves translating the biomimetic root system topology into a construction blueprint, specifying the three-dimensional coordinates, drilling depth, and sensor type of each monitoring node. Multiple monitoring nodes are individual deployment points in the sensor network, and each point will be equipped with an integrated sensing unit.

[0029] According to the deployment plan, sensors are deployed in the settlement area of ​​the substation to form a biomimetic root-like mesh array. Multiple monitoring nodes are distributed throughout the substation area to effectively capture settlement data from different regions. Based on the structural characteristics of the biomimetic root-like mesh array, multiple monitoring nodes are deployed, and through reasonable layout, it can be ensured that every important area can be monitored. Each monitoring node is equipped with a magnetostrictive displacement sensor and a triaxial accelerometer to achieve dual monitoring of displacement and vibration, constructing an integrated sensing unit. The magnetostrictive displacement sensor is a sensor that works based on the magnetostrictive effect principle; when the magnetic field changes, the length of the material changes slightly, used to accurately measure the displacement of an object. The triaxial accelerometer is a sensor that can simultaneously measure acceleration changes in three-dimensional space (X, Y, and Z axes), and is commonly used to monitor the vibration, tilt, or dynamic motion of objects.

[0030] The integrated sensing units are configured with communication modules, establishing a data transmission link. Each monitoring node possesses the ability to communicate with other nodes in real time, ensuring timely data transmission and synchronous updates. Following the data transmission link, the integrated sensing units are integrated with a biomimetic root-like mesh array to construct a complete biomimetic root-like mesh array sensor network, enabling real-time monitoring of substation settlement. By constructing this biomimetic root-like mesh array sensor network, full-coverage, high-precision settlement monitoring is achieved. The biomimetic root structure allows for highly adaptable sensor deployment, capable of handling different foundation conditions and load distributions, ensuring comprehensive monitoring of substation settlement.

[0031] The current phase data of the substation bus is collected in real time by the biomimetic root mesh array sensor network to calculate the magnetic field interference. The original displacement signal of the magnetostrictive displacement sensor is inversely compensated according to the dynamic magnetic field interference distribution data to obtain displacement compensation data.

[0032] Furthermore, this application also includes the following steps: acquiring the current phase of the substation busbar through the biomimetic root-like mesh array sensor network at multiple time periods to construct a current-magnetic field spatiotemporal correlation database; parsing the current-magnetic field spatiotemporal correlation database to obtain instantaneous magnetic field direction data and magnetic field intensity distribution data; retrieving the spatial three-dimensional layout parameters of the substation, mapping the instantaneous magnetic field direction data and the magnetic field intensity distribution data to the spatial three-dimensional layout parameters to perform magnetic field interference simulation calculations, and constructing a magnetic field interference distribution field map; and performing inverse compensation on the original displacement signal based on the magnetic field interference distribution field map to obtain displacement compensation data.

[0033] Furthermore, this application also includes the following steps: performing spatiotemporal alignment matching between the original displacement signal and the magnetic field interference distribution field spectrum to generate a spatiotemporal matching result; tracing the spatiotemporal matching result back to multiple monitoring nodes for interference identification to obtain interference feature quantities; calibrating the magnetic field-displacement mapping relationship curve, mapping the interference feature quantities to the magnetic field-displacement mapping relationship curve to perform inverse compensation operation on the original displacement signal to generate the displacement compensation data.

[0034] Furthermore, this application also includes the following steps: using the interference feature quantity as an index to traverse the magnetic field-displacement mapping relationship curve for matching and querying, and generating a displacement supplement quantity; performing vector superposition operation on the original displacement signal according to the displacement supplement quantity, and generating a reverse compensation quantity; applying the reverse compensation quantity to the original displacement signal according to the interference direction, and generating initial compensation displacement data; performing filtering calculation based on the initial compensation displacement data to obtain the displacement compensation data.

[0035] Specifically, a biomimetic root-like mesh array sensor network acquires the current phase of substation buses at different time points, obtaining current phase information for multiple time periods. This generates a current-magnetic field spatiotemporal correlation database, establishing a mapping relationship between time, space, and electromagnetic states. Each record includes a timestamp, the current amplitude and phase of each bus at that moment, and the theoretical magnetic field vector (direction and magnitude) of each point in the three-dimensional space of the substation calculated based on these currents. Specifically, by controlling each node in the biomimetic root-like mesh array sensor network to synchronize with a GPS or BeiDou clock, the network records the original displacement signal of its location at multiple consecutive time points. Simultaneously, the biomimetic root-like mesh array sensor network acquires the instantaneous current values ​​and phase angles of key buses within the substation with high precision through the secondary side of Rogowski coils or current transformers. By associating the current data at all time points with spatial location information, a current-magnetic field spatiotemporal correlation database is constructed.

[0036] The spatiotemporal correlation database of current and magnetic field is analyzed to obtain instantaneous magnetic field direction data and magnetic field intensity distribution data, which shows the distribution and intensity changes of the magnetic field at different locations and is crucial for analyzing magnetic field interference. Instantaneous magnetic field direction data is the direction of the magnetic field vector at a specific point in space at a particular moment, reflecting the instantaneous distribution of the magnetic field; magnetic field intensity distribution data is the intensity change data of the magnetic field at different spatial locations, helping to assess the intensity and range of magnetic field interference.

[0037] The three-dimensional spatial layout parameters of the substation are retrieved, including the spatial location and dimensions of all equipment and facilities within the substation. This includes the precise geometry, spatial location, and material properties of all magnetic conductors and fluid-carrying materials such as busbars, cables, steel frames, grounding grids, and equipment casings. The obtained instantaneous magnetic field direction and intensity distribution data are mapped to the three-dimensional spatial layout parameters for magnetic field interference simulation calculations. This simulates the magnetic field interference under this layout, ultimately generating a magnetic field interference distribution field map. The simulation calculations consider current phase, the geometric layout of equipment within the substation, and the laws of magnetic field propagation. By simulating the propagation and distribution of electromagnetic fields within the substation, the specific numerical values ​​and spatial distribution of magnetic field interference are obtained. The simulation result is a magnetic field interference distribution field map, showing the intensity and direction variations of magnetic field interference at different locations within the substation. Color or pattern changes in the map typically represent different magnetic field intensities, helping to identify the areas with the strongest magnetic field interference and the affected equipment within the substation. The magnetic field interference distribution field map clarifies which areas have excessively high magnetic field intensities, potentially affecting the normal operation of equipment or the accuracy of sensors within the substation.

[0038] In one example, assuming the three-dimensional layout parameters of the substation have been measured, including the spatial coordinates and layout of five main pieces of equipment within the substation: transformer location (10,15,5), busbar 1 location (12,15,5), busbar 2 location (8,15,5), switchgear location (11,17,5), and cable outlet location (10,18,5), the following magnetic field data were collected: Instantaneous magnetic field direction data: the magnetic field direction near the transformer is (1,0,0); magnetic field strength distribution data: the magnetic field strength near the transformer is 50μT, at busbar 1 it is 40μT, and near the cable outlet it is 30μT. Through magnetic field interference simulation calculation, a magnetic field interference distribution field spectrum is obtained, as shown below: the magnetic field strength near the transformer is 50μT, shown as the red area in the spectrum; the magnetic field strength in the busbar 1 area is 40μT, shown as the orange area; the magnetic field strength in the cable outlet area is low, only 30μT, shown as the blue area.

[0039] Based on the three-dimensional spatial layout parameters of the substation, the original displacement signal is unprocessed displacement data collected by a magnetostrictive displacement sensor. This data may be affected by factors such as magnetic field interference and vibration, thus requiring compensation processing. The original displacement signal is spatiotemporally aligned and matched with the magnetic field interference distribution field map to ensure accurate temporal and spatial alignment, enabling the correct identification of which parts of the displacement data are affected by magnetic field interference. The spatiotemporal matching results are traced back to multiple monitoring nodes for interference identification to obtain interference characteristic quantities. This involves identifying the magnetic field interference affecting the displacement signal, extracting interference characteristic quantities, and describing the nature of the interference, such as its intensity, frequency, or direction. Each node independently analyzes its own data pair (D_raw, B). Through signal analysis, such as coherence analysis, it is confirmed that the fluctuation of D_raw is highly correlated with the change of B, thereby identifying the existence of interference and extracting interference characteristic quantities from the magnetic field data B. For example, for a certain type of magnetostrictive sensor, its interference is mainly related to the vertical component of the magnetic field Bz; therefore, the interference characteristic quantity of this node is Bz(t).

[0040] Calibration is used to determine the magnetic field-displacement mapping curve. This curve, based on experimental or theoretical modeling, describes the relationship between changes in magnetic field and displacement, and is used to determine the specific impact of magnetic field interference on the displacement signal. For example, in a laboratory environment, a magnetostrictive displacement sensor is fixed to a zero-displacement platform and placed in a uniform magnetic field of precisely controllable strength generated by a Helmholtz coil. The output values ​​of the magnetostrictive displacement sensor under different magnetic field strengths are recorded, thus plotting the sensor's magnetic field-displacement mapping curve.

[0041] Using the interference characteristic as an index, the magnetic field-displacement mapping curve is traversed for matching and querying to obtain the corresponding displacement compensation amount under the given magnetic field strength, i.e., the reading deviation caused by the interference. The displacement compensation amount is then vector-superimposed with the original displacement signal to obtain a corrected displacement value, i.e., the reverse compensation amount. The vector superposition operation considers not only the magnitude of the displacement but also the direction and amplitude of the compensation amount to ensure that the compensated displacement data is consistent with the true displacement.

[0042] Interference direction is a signal processing concept used to indicate whether the interference affects the original signal positively or negatively. An inverse compensation amount is applied to the original displacement signal according to the interference direction to generate initial compensated displacement data. This initial compensated displacement data is a preliminary correction obtained after applying the inverse compensation amount and may still contain some errors. Filtering calculations are used to smooth the initial compensated displacement data, removing noise and obtaining the final displacement compensation data. Common methods such as low-pass filters and Gaussian filters can be used in the filtering process to remove high-frequency noise, ensuring the accuracy and stability of the data. Further smoothing and denoising processing of the initial compensated displacement data is performed through filtering calculations to remove potential high-frequency noise or other irrelevant signals, ultimately obtaining accurate displacement compensation data. For example, if the original displacement signal reading at monitoring node Node-15 is +1850μm, and the reading suddenly increases, indicating an anomaly, and the spatiotemporally matched magnetic field data is 15000μT, with interference identification and feature extraction also yielding 15000μT, a matching query in the magnetic field-displacement mapping curve yields the corresponding output value +1800μm. This is the calculated displacement compensation, meaning that even if the sensor does not experience actual displacement under this magnetic field, the reading will be 1800μm higher. Since the interference causes the reading to be higher, a reverse compensation of -1800μm is needed to compensate for it. Because the interference is determined to be positive, the reverse compensation is directly algebraically added to the original signal: 1850 - 1800 = 50μm, which is the initial compensated displacement data. The value of +50μm may contain various high-frequency noises. A low-pass filter is used to filter this time series. The filter will remove all fluctuations with frequencies higher than 1Hz and retain only the slowly changing low-frequency trend signal. The output value at this moment after filtering is +48μm. This +48μm is the final, clean, and highly reliable displacement compensation data, reflecting the cumulative settlement at this monitoring point.

[0043] The combination of spatiotemporal alignment matching and interference identification accurately identifies and compensates for errors caused by magnetic field variations, ensuring high precision and reliability of displacement data. Through the compensation process, the influence of magnetic field interference on the original displacement signal is eliminated, restoring the true settlement data.

[0044] Vibration monitoring is performed on multiple monitoring nodes using a triaxial accelerometer, and data weights are adjusted for these nodes to generate multiple node weight parameters.

[0045] Furthermore, this application also includes the following steps: real-time vibration state monitoring of multiple monitoring nodes using a triaxial accelerometer, collecting vibration acceleration time-series data in the three-axis direction; vibration analysis based on the vibration acceleration time-series data in the three-axis direction, constructing vibration state parameters; reliability assessment of the sensor based on the vibration state parameters, obtaining a reliability score; and dynamic adjustment of the weight coefficients of the multiple monitoring nodes according to the reliability score, generating the weight parameters of the multiple nodes.

[0046] Furthermore, this application also includes the following steps: performing Fourier transform on the vibration acceleration time series data in the three axes, and extracting the dominant frequency component and harmonic features based on the transformation results; calculating the root mean square value of the vibration acceleration time series data in the three axes, and setting a vibration amplitude quantization index; performing power spectral density analysis on the dominant frequency component and the harmonic features according to the vibration amplitude quantization index to determine the frequency distribution characteristics; and performing frequency domain energy concentration analysis based on the frequency distribution characteristics to construct the vibration state parameters.

[0047] Furthermore, this application also includes the following steps: retrieving vibration data from historical normal operating conditions and establishing a benchmark vibration feature library; training and classifying the benchmark vibration feature library to determine normal vibration modes and abnormal vibration modes; performing fuzzy inference based on the normal vibration modes and the abnormal vibration modes, and performing multi-dimensional evaluation of the sensor based on the inference results to set multiple confidence intervals; matching the vibration state parameters with the multiple confidence intervals to generate the confidence score.

[0048] Specifically, triaxial accelerometers are used to monitor the vibration status of multiple monitoring nodes in real time. Each monitoring node is equipped with a triaxial accelerometer, which simultaneously collects acceleration data in three directions, obtaining time-series vibration acceleration data in three axes. This data comprehensively describes the changes in the vibration state of multiple monitoring nodes in space over time. For example, the X-axis measures vibration along the horizontal direction, the Y-axis measures vibration along the vertical direction, and the Z-axis measures vibration perpendicular to the ground. The triaxial accelerometers continuously acquire vibration acceleration values ​​at different time points, converting them into time-series data, which is then transmitted to the data acquisition system as digital signals, recording the vibration acceleration at each moment. By analyzing the time-series data, information such as vibration amplitude, frequency, and acceleration fluctuations is obtained.

[0049] Performing a Fourier transform on the time-series vibration acceleration data in three axes converts the time-domain vibration acceleration data into frequency-domain data, revealing the intensity of each frequency component in the vibration signal. Extracting the dominant frequency component and harmonic characteristics from the signal through Fourier transform is crucial in vibration analysis. The dominant frequency component corresponds to the operating frequency of equipment within the substation, while harmonic components are typically related to the nonlinear characteristics or fault modes of the equipment, such as imbalance, looseness, and wear. The dominant frequency component is the main frequency component of the signal in the frequency domain, usually corresponding to the strongest vibration frequency in the signal, and in many cases represents the operating frequency or main vibration mode of the equipment within the substation. Harmonic characteristics are frequency components in the signal whose frequencies are integer multiples of the dominant frequency, usually caused by nonlinear or periodic faults in the equipment within the substation. In vibration analysis, harmonic components can reflect problems such as wear, looseness, or imbalance in the equipment within the substation.

[0050] The root mean square (RMS) value of vibration acceleration time-series data in three axes is calculated to measure the effective value of the signal or the intensity of the vibration. The RMS value represents the average level of signal energy and is commonly used in vibration analysis to represent the amplitude of vibration. By calculating the RMS value, the average intensity of the vibration signal can be obtained, and a vibration amplitude quantification index can be set based on this value. Vibration amplitude quantification indices include average intensity, maximum intensity, or other standardized indicators. Vibration amplitude quantification indices are typically based on one or more threshold values ​​set according to the RMS value, used to convert continuous vibration amplitude values ​​into discrete, physically meaningful levels, such as slight, moderate, and severe.

[0051] Power spectral density analysis is performed on the dominant frequency components and harmonic characteristics based on the vibration amplitude quantification index. The power spectral density of the signal is calculated, and the frequency distribution characteristics are determined accordingly. Frequency distribution characteristics refer to the distribution of the signal in the frequency domain, i.e., the intensity of each frequency component. By analyzing the frequency distribution characteristics, we can understand the different frequency components of the vibration and their contribution to the overall vibration state.

[0052] Frequency domain energy concentration analysis is performed based on frequency distribution characteristics. By analyzing the degree of energy concentration of the vibration signal in the frequency domain, it can be determined whether the frequency distribution of the signal is concentrated in certain specific frequency bands. If the energy of the vibration signal is concentrated in a certain frequency band, it may indicate that the equipment in the substation is in a specific vibration mode, such as the operating frequency or fault frequency. Vibration state parameters are parameters reflecting the vibration state of equipment or structure, obtained through time-domain and frequency-domain analysis of the vibration signal. These parameters include vibration amplitude, frequency, and harmonic components.

[0053] Vibration data from historical normal operating conditions is retrieved; that is, vibration data collected during the normal operation of the substation represents the vibration characteristics under healthy conditions. Based on this, a benchmark vibration characteristic library is established. The benchmark vibration characteristic library is a database containing vibration data characteristics from historical normal operating conditions. It stores the vibration modes of the equipment during normal operation, including data characteristics such as acceleration, frequency, harmonic components, and vibration amplitude in various directions.

[0054] An unsupervised learning algorithm is used to train and classify a baseline vibration feature database to identify normal and abnormal vibration modes. Normal vibration modes represent typical vibration data of equipment when there is no fault, and usually have fixed frequency, amplitude and other characteristics; while abnormal vibration modes may include different vibration characteristics caused by equipment faults, such as looseness, wear, imbalance, etc., and are usually accompanied by large vibration amplitude, frequency shift or harmonic components.

[0055] Fuzzy reasoning is based on normal and abnormal vibration modes. Fuzzy reasoning is a computational method for handling uncertainty and fuzzy concepts, simulating the human brain's decision-making process. It uses membership degrees to replace binary yes / no judgments, making it suitable for scenarios with unclear boundaries, such as vibration assessment. It compares real-time vibration data with established normal and abnormal vibration modes to determine whether the current equipment is in a normal or abnormal state.

[0056] The sensors are evaluated in multiple dimensions using fuzzy inference results, considering various factors at each monitoring node, such as signal stability, amplitude, and frequency response, to ensure normal sensor operation. The evaluation results are used to define multiple confidence intervals based on the sensor's reliability and performance, with each interval representing a different level of sensor confidence. The confidence score, generated based on fuzzy inference and confidence intervals, represents the degree of matching between the current vibration state and historical normal operating conditions, typically used to determine whether the equipment is operating normally or experiencing an anomaly. Based on the confidence intervals, the system determines whether each sensor is functioning correctly, avoiding erroneous judgments caused by sensor malfunctions or data noise. Vibration state parameters are matched with multiple confidence intervals to determine their corresponding confidence scores, representing the degree of matching between the current vibration data and normal operating conditions, usually expressed numerically. For example, a higher confidence score indicates that the current vibration state is closer to historical normal operating conditions, and the equipment is in a normal state; conversely, a lower score indicates an anomaly or malfunction.

[0057] The weight coefficients of multiple monitoring nodes are dynamically adjusted based on their reliability scores. Different monitoring nodes may have different monitoring accuracy and importance. By dynamically adjusting the weight coefficients of each node, the influence of the node can be optimized according to its reliability score. For example, if a node has a low reliability score, it means that the data from that node is unreliable, and its weight needs to be reduced. Finally, multiple node weight parameters are generated based on the weight coefficients of each node. For example, in real-time data analysis, suppose the vibration state parameters of monitoring node 1 are: X-axis frequency 49.8Hz, RMS value 0.04g. Through fuzzy inference, it is determined that the vibration characteristics of this node are close to normal, and the reliability score is 0.85. The weight coefficient of monitoring node 1 is dynamically adjusted to increase its influence in the system. If the vibration state parameters of another monitoring node 2 are: X-axis frequency 47Hz, RMS value 0.05g; after fuzzy inference, the reliability score of this node is 0.65. Therefore, the weight coefficient of node 2 will be reduced accordingly to decrease its impact on the overall result.

[0058] By establishing a benchmark vibration feature library and training classification, accurate differentiation between normal and abnormal equipment states can be achieved. The introduction of fuzzy reasoning enables effective judgment under uncertain or fuzzy data conditions, improving the robustness of fault diagnosis. Dynamically adjusting node weight coefficients further enhances the reliability of monitoring data.

[0059] The substation settlement is predicted based on the displacement compensation data and the weight parameters of the multiple nodes. The biomimetic root mesh array sensor network is back-calibrated based on the prediction results. The substation settlement is then monitored in a closed loop according to the calibration results.

[0060] Furthermore, this application also includes the following steps: spatiotemporally aligning and fusing the displacement compensation data with the weight parameters of the multiple nodes to construct a weighted displacement dataset; performing settlement prediction based on the weighted displacement dataset according to the settlement time sequence to generate a time-series settlement prediction result; retrieving actual settlement monitoring values ​​through the biomimetic root mesh array sensor network, and comparing the time-series settlement prediction result with the actual settlement monitoring value by sending it back to the biomimetic root mesh array sensor network; triggering a backtracking calibration command when the deviation value is greater than a preset deviation threshold; and dynamically adjusting the sensing parameters of the biomimetic root mesh array sensor network through the backtracking calibration command to construct a settlement monitoring closed-loop feedback mechanism.

[0061] Specifically, displacement compensation data is spatiotemporally aligned and fused with the weight parameters of multiple nodes. The displacement compensation data for each monitoring node is obtained based on its measurement results and corrections for external disturbances. By spatiotemporally aligning the displacement compensation data, it is ensured that displacement data from different nodes at the same time and spatial location can be fused together. The displacement data of each node is weighted according to its weight parameters, making the more reliable node data have a greater impact on the final result.

[0062] By combining displacement compensation data and weight parameters of multiple nodes for weighted calculation, a weighted displacement dataset is obtained. The displacement data of each node is weighted according to its weight, thereby optimizing the data fusion result. The weighted displacement dataset is no longer the original data set of each node, but a unified and standardized dataset representing the overall settlement trend of the entire monitoring area after reliability weighting.

[0063] Based on a weighted displacement dataset, settlement is predicted according to the settlement time series. The settlement time series is a sequence of settlement changes of a substation over time, recording the changes in the amount of settlement of the structure over time, and is used to predict future settlement trends. The time series settlement prediction result is a prediction of the settlement trend and amount of settlement over a future period of time, and the output is usually a sequence of predicted values ​​and confidence intervals.

[0064] A biomimetic root-like mesh array sensor network is used to retrieve actual settlement monitoring values, i.e., directly collected real settlement data. The time-series settlement prediction results are compared with the actual settlement monitoring values ​​to calculate the difference, i.e., the deviation value. If the deviation value is greater than a preset deviation threshold, it indicates that there is an error in the prediction, requiring adjustment and triggering a backtracking calibration command. The preset deviation threshold is a pre-defined deviation tolerance range, typically a numerical interval. The backtracking calibration command is an automatically generated command signal used to trigger the calibration process of the sensor network model.

[0065] By using backtracking calibration commands, the sensing parameters of the biomimetic root-based mesh array sensor network are adjusted, forming a closed-loop feedback mechanism for settlement monitoring. This can involve recalibrating sensor sensitivity, increasing sensor sampling frequency, or adjusting the weights of specific nodes. The closed-loop feedback mechanism continuously improves the performance of the biomimetic root-based mesh array sensor network and ensures the reliability of long-term monitoring data. The purpose of backtracking calibration is to adjust key parameters in the sensor network, including sensor sensitivity, sampling frequency, and node weights, based on the difference between predicted and actual settlement data. Backtracking calibration is typically triggered when the deviation between predicted and actual values ​​is large, adjusting sensor parameters to reduce future prediction errors and thus improve the overall accuracy of the monitoring system.

[0066] Based on the results of retrospective calibration, closed-loop monitoring of substation settlement is implemented, and the settlement prediction model is continuously optimized based on each calibration result. In closed-loop monitoring, predicted and actual settlement data are continuously compared, and self-adjustment is performed based on each error to ensure the long-term effectiveness and accuracy of settlement monitoring. Through retrospective calibration and closed-loop monitoring, automatic adjustments are made based on the deviation between predicted and actual settlement data, ensuring the accuracy and reliability of long-term monitoring. The combination of weighted displacement datasets and retrospective calibration enables continuous optimization of prediction results under different operating conditions, reducing prediction errors. Through the closed-loop feedback mechanism, each calibration and adjustment improves the accuracy of settlement prediction, avoiding data deviations caused by sensor errors or environmental interference. The settlement monitoring closed-loop feedback mechanism dynamically adjusts the configuration of the biomimetic root-like mesh array sensor network, helping to identify potential settlement risks in advance and ensuring the safe operation of the substation.

[0067] In summary, the substation settlement monitoring method provided in this application has the following beneficial effects: By deploying a biomimetic root-like mesh array sensor network, which includes multiple monitoring nodes, each node comprising a magnetostrictive displacement sensor and a triaxial accelerometer, the network collects real-time current phase data of the substation busbars to calculate magnetic field interference. Based on the dynamic magnetic field interference distribution data, the original displacement signal from the magnetostrictive displacement sensor is inversely compensated to obtain displacement compensation data. Vibration monitoring is performed on the multiple monitoring nodes using the triaxial accelerometer, and data weighting is adjusted to generate multiple node weight parameters. Settlement prediction of the substation is performed based on the displacement compensation data and the multiple node weight parameters. The biomimetic root-like mesh array sensor network is then retrospectively calibrated based on the prediction results, and closed-loop monitoring of substation settlement is conducted according to the calibration results. In other words, by deploying multiple smart sensors to construct a biomimetic root-like mesh array sensor network, the current phase data of the substation busbar is collected in real time and magnetic field interference is compensated. Vibration monitoring of multiple monitoring nodes is carried out through triaxial accelerometers, the node data weights are adjusted, settlement is predicted based on displacement compensation data and node weight parameters, and the sensor network is backtracked and calibrated based on the prediction results, so as to achieve accurate prediction and closed-loop monitoring of substation settlement and improve the accuracy of settlement monitoring.

[0068] Example 2: Based on the same inventive concept as the substation settlement monitoring method in Example 1, this application also provides a substation settlement monitoring system. Please refer to the appendix. Figure 2 The substation settlement monitoring system includes: The sensor deployment module 11 is used to deploy a biomimetic root-like mesh array sensor network, which includes multiple monitoring nodes, each of which contains a magnetostrictive displacement sensor and a triaxial accelerometer. The interference compensation module 12 is used to collect current phase data of the substation busbar in real time through the biomimetic root-like mesh array sensor network to calculate magnetic field interference, and to perform inverse compensation on the original displacement signal of the magnetostrictive displacement sensor based on the dynamic magnetic field interference distribution data to obtain displacement compensation data. The weight adjustment module 13 is used to perform vibration monitoring on multiple monitoring nodes through the triaxial accelerometer, and to adjust the data weights of multiple monitoring nodes to generate multiple node weight parameters. The settlement prediction module 14 is used to predict the settlement of the substation based on the displacement compensation data and the multiple node weight parameters, to perform backtracking calibration on the biomimetic root-like mesh array sensor network based on the prediction results, and to perform closed-loop monitoring of the substation settlement according to the calibration results.

[0069] Furthermore, the sensor deployment module 11 in the substation settlement monitoring system is also used for: introducing the substation's foundation structure characteristics and load distribution characteristics to construct a biomimetic root system topology; performing sensor deployment analysis on multiple monitoring nodes based on the biomimetic root system topology to formulate a deployment scheme; deploying sensors in the substation's settlement area according to the deployment scheme to generate a biomimetic root system mesh array; configuring magnetostrictive displacement sensors and triaxial accelerometers across multiple monitoring nodes based on the biomimetic root system mesh array to construct an integrated sensing unit; configuring the integrated sensing unit for communication to construct a data transmission link; and integrating the integrated sensing unit with the biomimetic root system mesh array according to the data transmission link to construct the biomimetic root system mesh array sensor network.

[0070] Furthermore, the interference compensation module 12 in the substation settlement monitoring system is also used for: acquiring the current phase of the substation busbar through the biomimetic root-like mesh array sensor network at multiple time periods to construct a current-magnetic field spatiotemporal correlation database; parsing the current-magnetic field spatiotemporal correlation database to obtain instantaneous magnetic field direction data and magnetic field intensity distribution data; retrieving the spatial three-dimensional layout parameters of the substation, mapping the instantaneous magnetic field direction data and the magnetic field intensity distribution data to the spatial three-dimensional layout parameters to perform magnetic field interference simulation calculations and construct a magnetic field interference distribution field map; and performing inverse compensation on the original displacement signal based on the magnetic field interference distribution field map to obtain displacement compensation data.

[0071] Furthermore, the interference compensation module 12 in the substation settlement monitoring system is also used for: performing spatiotemporal alignment matching between the original displacement signal and the magnetic field interference distribution field map to generate a spatiotemporal matching result; tracing the spatiotemporal matching result back to multiple monitoring nodes for interference identification to obtain interference characteristic quantities; calibrating the magnetic field-displacement mapping relationship curve, mapping the interference characteristic quantities to the magnetic field-displacement mapping relationship curve to perform inverse compensation operation on the original displacement signal to generate the displacement compensation data.

[0072] Furthermore, the interference compensation module 12 in the substation settlement monitoring system is also used to: use the interference feature quantity as an index to traverse the magnetic field-displacement mapping curve for matching and querying, and generate a displacement supplement quantity; perform vector superposition operation on the original displacement signal according to the displacement supplement quantity to generate a reverse compensation quantity; apply the reverse compensation quantity to the original displacement signal according to the interference direction to generate initial compensation displacement data; and perform filtering calculation based on the initial compensation displacement data to obtain the displacement compensation data.

[0073] Furthermore, the weight adjustment module 13 in the substation settlement monitoring system is also used for: real-time vibration state monitoring of multiple monitoring nodes using a triaxial accelerometer, and collecting vibration acceleration time-series data in the three-axis direction; performing vibration analysis based on the vibration acceleration time-series data in the three-axis direction to construct vibration state parameters; evaluating the reliability of the sensors based on the vibration state parameters to obtain a reliability score; and dynamically adjusting the weight coefficients of the multiple monitoring nodes according to the reliability score to generate the weight parameters of the multiple nodes.

[0074] Furthermore, the weight adjustment module 13 in the substation settlement monitoring system is also used for: performing Fourier transform on the vibration acceleration time series data in the three axes, extracting the dominant frequency component and harmonic characteristics based on the transformation results; calculating the root mean square value of the vibration acceleration time series data in the three axes, and setting a vibration amplitude quantization index; performing power spectral density analysis on the dominant frequency component and the harmonic characteristics according to the vibration amplitude quantization index to determine the frequency distribution characteristics; and performing frequency domain energy concentration analysis based on the frequency distribution characteristics to construct the vibration state parameters.

[0075] Furthermore, the weight adjustment module 13 in the substation settlement monitoring system is also used for: retrieving vibration data from historical normal operating conditions and establishing a benchmark vibration feature library; training and classifying the benchmark vibration feature library to determine normal vibration modes and abnormal vibration modes; performing fuzzy inference based on the normal vibration modes and the abnormal vibration modes, performing multi-dimensional evaluation of the sensors based on the inference results, and setting multiple confidence intervals; matching the vibration state parameters with the multiple confidence intervals to generate the confidence score.

[0076] Furthermore, the settlement prediction module 14 in the substation settlement monitoring system is also used for: spatiotemporally aligning and fusing the displacement compensation data with the weight parameters of the multiple nodes to construct a weighted displacement dataset; performing settlement prediction based on the weighted displacement dataset according to the settlement time sequence to generate a time-series settlement prediction result; retrieving the actual settlement monitoring value through the biomimetic root mesh array sensor network, and sending the time-series settlement prediction result back to the biomimetic root mesh array sensor network for deviation comparison with the actual settlement monitoring value; triggering a backtracking calibration command when the deviation value is greater than a preset deviation threshold; and dynamically adjusting the sensing parameters of the biomimetic root mesh array sensor network through the backtracking calibration command to construct a settlement monitoring closed-loop feedback mechanism.

[0077] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Figure 1 The substation settlement monitoring method and specific example in Embodiment 1 are also applicable to the substation settlement monitoring system in this embodiment. Through the foregoing detailed description of the substation settlement monitoring method, those skilled in the art can clearly understand the substation settlement monitoring system in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0078] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0079] Obviously, those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for monitoring substation settlement, characterized in that, include: A biomimetic root mesh array sensor network is deployed, the biomimetic root mesh array sensor network including multiple monitoring nodes, each monitoring node including a magnetostrictive displacement sensor and a triaxial accelerometer; The current phase data of the substation bus is collected in real time by the biomimetic root mesh array sensor network to calculate the magnetic field interference. The original displacement signal of the magnetostrictive displacement sensor is inversely compensated according to the dynamic magnetic field interference distribution data to obtain displacement compensation data. Vibration monitoring is performed on multiple monitoring nodes using a triaxial accelerometer, and data weights are adjusted for multiple monitoring nodes to generate multiple node weight parameters. The substation settlement is predicted based on the displacement compensation data and the weight parameters of the multiple nodes. The biomimetic root mesh array sensor network is back-calibrated based on the prediction results. The substation settlement is then monitored in a closed loop according to the calibration results.

2. The substation settlement monitoring method as described in claim 1, characterized in that, A biomimetic root-like mesh array sensor network is deployed, comprising multiple monitoring nodes, each node including a magnetostrictive displacement sensor and a triaxial accelerometer, including: By incorporating the foundation structure characteristics and load distribution characteristics of substations, a biomimetic root system topology is constructed. Based on the biomimetic root system topology, sensor deployment analysis was conducted on multiple monitoring nodes, and a deployment plan was formulated. Sensors are deployed in the settlement area of ​​the substation according to the deployment scheme to generate a biomimetic root network array; Based on the biomimetic root mesh array, a magnetostrictive displacement sensor and a triaxial accelerometer are configured to traverse multiple monitoring nodes to construct an integrated sensing unit; The integrated sensing unit is configured for communication to construct a data transmission link. According to the data transmission link, the integrated sensing unit is integrated with the biomimetic root mesh array to construct the biomimetic root mesh array sensor network.

3. The substation settlement monitoring method as described in claim 1, characterized in that, The current phase data of the substation busbar is collected in real time by the biomimetic root-like mesh array sensor network to calculate magnetic field interference. Based on the dynamic magnetic field interference distribution data, the original displacement signal of the magnetostrictive displacement sensor is inversely compensated to obtain displacement compensation data, including: The current phase of the substation busbar is acquired in multiple time periods through the biomimetic root mesh array sensor network to construct a current-magnetic field spatiotemporal correlation database. Based on the current-magnetic field spatiotemporal correlation database, instantaneous magnetic field direction data and magnetic field intensity distribution data are obtained through analysis. The spatial three-dimensional layout parameters of the substation are retrieved, and the instantaneous magnetic field direction data and the magnetic field intensity distribution data are mapped to the spatial three-dimensional layout parameters to perform magnetic field interference simulation calculations and construct a magnetic field interference distribution field map. Based on the magnetic field interference distribution field map, the original displacement signal is inversely compensated to obtain displacement compensation data.

4. The substation settlement monitoring method as described in claim 3, characterized in that, Based on the magnetic field interference distribution field map, inverse compensation is performed on the original displacement signal to obtain displacement compensation data, including: The original displacement signal is spatiotemporally aligned and matched with the magnetic field interference distribution field map to generate a spatiotemporal matching result. The spatiotemporal matching results are traced back to multiple monitoring nodes for interference identification to obtain interference feature quantities; The magnetic field-displacement mapping curve is calibrated, and the interference characteristic quantity is mapped to the magnetic field-displacement mapping curve to perform inverse compensation operation on the original displacement signal to generate the displacement compensation data.

5. The substation settlement monitoring method as described in claim 4, characterized in that, Mapping the interference features to the magnetic field-displacement mapping curve and performing inverse compensation on the original displacement signal to generate the displacement compensation data includes: The interference feature quantity is used as an index to traverse the magnetic field-displacement mapping curve for matching and querying, and a displacement supplement quantity is generated. The original displacement signal is vector superimposed according to the displacement compensation amount to generate a reverse compensation amount; The reverse compensation amount is applied to the original displacement signal according to the direction of interference to generate initial compensation displacement data; The displacement compensation data is obtained by filtering and calculating based on the initial compensation displacement data.

6. The substation settlement monitoring method as described in claim 1, characterized in that, Vibration monitoring was performed on multiple monitoring nodes using a triaxial accelerometer. Data weighting was adjusted for these nodes to generate multiple node weight parameters, including: Real-time vibration status monitoring of multiple monitoring nodes is performed using a triaxial accelerometer, and time-series data of vibration acceleration in three axes are collected. Vibration analysis is performed based on the time-series data of vibration acceleration in the three axes to construct vibration state parameters; The reliability of the sensor is evaluated based on the vibration state parameters to obtain a reliability score; The weight coefficients of multiple monitoring nodes are dynamically adjusted according to the credibility score to generate the weight parameters of the multiple nodes.

7. A substation settlement monitoring method as described in claim 6, characterized in that, Vibration analysis is performed based on the time-series data of vibration acceleration in the three axes to construct vibration state parameters, including: Fourier transform is performed on the vibration acceleration time series data in the three axes, and the main frequency component and harmonic features are extracted based on the transformation results; Calculate the root mean square value of the vibration acceleration time series data in the three axes, and set the vibration amplitude quantization index; Power spectral density analysis is performed on the dominant frequency component and the harmonic characteristics according to the vibration amplitude quantification index to determine the frequency distribution characteristics; Based on the frequency distribution characteristics, frequency domain energy concentration analysis is performed to construct the vibration state parameters.

8. A substation settlement monitoring method as described in claim 6, characterized in that, The reliability of the sensor is evaluated based on the vibration state parameters to obtain a reliability score, including: Retrieve vibration data from historical normal operating conditions and establish a benchmark vibration characteristic library; The reference vibration feature library is trained and classified to determine normal vibration modes and abnormal vibration modes; Fuzzy reasoning is performed based on the normal vibration mode and the abnormal vibration mode. The sensor is then evaluated in multiple dimensions based on the reasoning results, and multiple confidence intervals are set. The vibration state parameters are matched with the multiple confidence intervals to generate the confidence score.

9. A substation settlement monitoring method as described in claim 1, characterized in that, Settlement prediction for the substation is performed based on the displacement compensation data and the weight parameters of the multiple nodes. Backward calibration of the biomimetic root-like mesh array sensor network is then performed based on the prediction results. Closed-loop monitoring of substation settlement is then conducted according to the calibration results, including: The displacement compensation data is spatiotemporally aligned and fused with the weight parameters of the multiple nodes to construct a weighted displacement dataset. Based on the weighted displacement dataset, settlement prediction is performed according to the settlement time sequence to generate time-series settlement prediction results. The actual settlement monitoring value is retrieved through the biomimetic root network array sensor network. The time-series settlement prediction result is sent back to the biomimetic root network array sensor network and compared with the actual settlement monitoring value. When the deviation value is greater than the preset deviation threshold, a backtracking calibration command is triggered. The sensing parameters of the biomimetic root mesh array sensor network are dynamically adjusted by the backtracking calibration command to construct a closed-loop feedback mechanism for settlement monitoring.

10. A substation settlement monitoring system, characterized in that, The steps for implementing the substation settlement monitoring method according to any one of claims 1 to 9, wherein the substation settlement monitoring system comprises: A sensor deployment module is used to deploy a biomimetic root mesh array sensor network, which includes multiple monitoring nodes, each monitoring node containing a magnetostrictive displacement sensor and a triaxial accelerometer. The interference compensation module is used to collect the current phase data of the substation bus in real time through the biomimetic root mesh array sensor network to calculate the magnetic field interference, and to perform inverse compensation on the original displacement signal of the magnetostrictive displacement sensor according to the dynamic magnetic field interference distribution data to obtain displacement compensation data. The weight adjustment module is used to monitor the vibration of multiple monitoring nodes through a triaxial accelerometer, adjust the data weight of multiple monitoring nodes, and generate multiple node weight parameters. The settlement prediction module is used to predict the settlement of the substation based on the displacement compensation data and the weight parameters of the multiple nodes, perform retrospective calibration of the biomimetic root mesh array sensor network based on the prediction results, and perform closed-loop monitoring of the substation settlement according to the calibration results.

Images