Method for cathode protection of natural gas pipeline

Abstract

The invention discloses a method for cathode protection of a natural gas pipeline, and the method achieves the high-precision reasoning of the state of a target monitoring point through the spatial sequence coding and multi-scale hierarchical interaction of the data of a plurality of monitoring points along the line. Further calculating a proportionality coefficient between a reasoning value and a measured value to dynamically quantify the confidence coefficient of the input data; then, the input data confidence coefficient serves as a weight to directly participate in PID parameter modulation, so that a PID controller can conduct self-adaptive adjustment according to the current input data quality; and finally, the fine-tuned PID is used for generating an output instruction of the potentiostat, so that the optimal cathode protection effect and the optimal energy consumption efficiency are both considered. Through the mode, the capability of responding to noise interference and sudden abnormal events under on-site complex working conditions is remarkably enhanced, so that a more scientific, reasonable, energy-saving and efficient current control instruction is provided for the output of the potentiostat, and the optimal protection effect and the optimal energy consumption balance are realized.

Description

Technical Field

[0001] The present application relates to the field of intelligent control, and more particularly, to a method for cathodic protection of natural gas pipelines. Background Art

[0002] The establishment of a cathodic protection system is of irreplaceable importance in the long-term safe operation of natural gas pipelines. Cathodic protection solutions are not only a core measure for preventing corrosion and extending the service life of underground pipelines, but also a key technical path for ensuring energy transmission safety, reducing accident risks, and minimizing economic losses. As the scale of natural gas pipeline networks continues to expand and the operating environment becomes increasingly complex, the traditional approach of relying on manual inspections and empirical adjustments can no longer meet the urgent need for efficient and intelligent protection methods. Therefore, the establishment of a scientific, automated, and dynamically optimized cathodic protection solution is an important foundation for improving the level of pipeline management throughout its entire life cycle and supporting the security of the national energy strategy.

[0003] Currently, mainstream cathodic protection solutions for natural gas pipelines primarily include automatic or semi-automatic control systems based on devices such as constant potentiostats (CP systems) or constant current sources. These systems typically collect data from key monitoring points using fixed-point data collection and implement closed-loop control using preset parameters. However, due to the complex and ever-changing field environment, real-time data is susceptible to noise and sensor drift. Existing control algorithms, such as PID, often overlook the confidence level of input data. This makes it impossible to adaptively adjust parameters when input data quality deteriorates, leading to over- or under-protection, which not only compromises corrosion protection effectiveness but also wastes energy.

[0004] Therefore, an optimized solution for cathodic protection of natural gas pipelines is expected. Summary of the Invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a method for cathodic protection of natural gas pipelines, which realizes high-precision reasoning of the state of the target monitoring point by spatial sequence encoding and multi-scale hierarchical interaction of data from multiple monitoring points along the line, and further calculates the proportional coefficient between the inference value and the measured value to dynamically quantify the input data confidence; then, the input data confidence is directly involved in the PID parameter modulation as a weight, so that the PID controller can be adaptively adjusted according to the current input data quality; finally, the fine-tuned PID is used to generate the output instruction of the constant potential instrument, so as to achieve the best cathodic protection effect and the best energy efficiency. In this way, the response capability to noise interference and sudden abnormal events under complex on-site working conditions is significantly enhanced, thereby providing a more scientific, reasonable, energy-saving and efficient current control instruction for the output of the constant potential instrument, and achieving the best protection effect and the best energy consumption balance.

[0006] According to one aspect of the present application, a method for cathodic protection of a natural gas pipeline is provided, comprising: Obtain real-time pipe-to-ground potential from multiple key monitoring points along the pipeline; Extract the real-time pipe-to-ground potential of the target key monitoring point from the real-time pipe-to-ground potential of multiple key monitoring points as the data item to be verified; Based on the global distribution of the real-time pipe-to-ground potential of the multiple key monitoring points, performing input data confidence calculation on the data item to be verified to obtain input data confidence; Modulating the initial parameters of the PID controller based on the data confidence to obtain a fine-tuned PID controller; The real-time pipe-to-ground potential and the preset target pipe-to-ground potential of the target key monitoring point are input into the fine-tuning PID controller to obtain the output current control instruction of the constant potentiostat sent to the target key monitoring point.

[0007] Compared with the existing technology, the present application provides a method for cathodic protection of natural gas pipelines. It realizes high-precision reasoning of the state of the target monitoring point by spatial sequence encoding and multi-scale hierarchical interaction of data from multiple monitoring points along the line, and further calculates the proportional coefficient between the inference value and the measured value to dynamically quantify the input data confidence. Subsequently, the input data confidence is directly involved in the PID parameter modulation as a weight, so that the PID controller can be adaptively adjusted according to the current input data quality. Finally, the fine-tuned PID is used to generate the output instruction of the constant potential instrument to achieve the best cathodic protection effect and the best energy efficiency. In this way, the response capability to noise interference and sudden abnormal events under complex working conditions on site is significantly enhanced, thereby providing a more scientific, reasonable, energy-saving and efficient current control instruction for the output of the constant potential instrument, achieving the best protection effect and the best energy consumption balance. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0009] Figure 1 is a flow chart of a method for cathodic protection of a natural gas pipeline according to an embodiment of the present application; Figure 2 Schematic diagram of data flow for a method for cathodic protection of a natural gas pipeline according to an embodiment of the present application; Figure 3This is a flowchart of sub-step S3 of the method for cathodic protection of a natural gas pipeline according to an embodiment of the present application; Figure 4 4 is a flowchart of sub-step S32 of the method for cathodic protection of a natural gas pipeline according to an embodiment of the present application. DETAILED DESCRIPTION

[0010] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0011] As used in this application and the claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not intended to refer to the singular but may include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0012] Although the present application makes various references to certain modules in the system according to embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are illustrative only, and different aspects of the system and method can use different modules.

[0013] Flowcharts are used in this application to illustrate the operations performed by the systems according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the various steps may be processed in reverse order or simultaneously, as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0014] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0015] In the technical solution of the present application, a method for cathodic protection of a natural gas pipeline is proposed. Figure 1 The figure is a flow chart of a method for cathodic protection of a natural gas pipeline according to an embodiment of the present application. Figure 2 Schematic diagram of data flow for the method for cathodic protection of natural gas pipelines according to an embodiment of the present application. Figure 1 and Figure 2As shown, the method for cathodic protection of a natural gas pipeline according to an embodiment of the present application includes the following steps: S1, obtaining real-time pipe-to-ground potentials from multiple key monitoring points along the pipeline; S2, extracting the real-time pipe-to-ground potential of a target key monitoring point from the real-time pipe-to-ground potentials of multiple key monitoring points as a data item to be verified; S3, based on the global distribution of the real-time pipe-to-ground potentials of the multiple key monitoring points, performing input data confidence calculation on the data item to be verified to obtain input data confidence; S4, modulating the initial parameters of the PID controller based on the data confidence to obtain a fine-tuning PID controller; S5, inputting the real-time pipe-to-ground potential of the target key monitoring point and a preset target pipe-to-ground potential into the fine-tuning PID controller to obtain an output current control instruction of the constant potentiostat sent to the target key monitoring point.

[0016] In particular, S1 acquires real-time pipe-to-ground potential from multiple key monitoring points along the pipeline. Because natural gas pipelines are typically buried below the surface and traverse a variety of soil environments and complex geological structures, factors such as external interference, changes in soil resistivity, and stray currents can lead to significant differences in corrosion risk at different locations. Therefore, relying solely on single-point or small amounts of monitoring data makes it difficult to fully reflect the overall protection status of the pipeline, nor can it timely capture local anomalies and potential risks. By scientifically deploying multiple key monitoring points along the pipeline and collecting real-time pipe-to-ground potential at these points, the protection effect of each area can be dynamically perceived, achieving precise control of the global status of the cathodic protection system.

[0017] During the specific implementation process, it is first necessary to rationally select several key monitoring points based on factors such as the pipeline direction, geological environment, and the distribution of electrochemical interference sources. High-precision potential acquisition devices (such as reference electrodes and signal acquisition terminals) should be deployed at each monitoring point, and the real-time pipe-to-ground potential data collected should be uploaded to the central processing platform via wired or wireless communication. This platform can achieve synchronous reception and time-series archiving of all monitoring data, laying a solid foundation for subsequent data processing and model reasoning. By obtaining real-time pipe-to-ground potentials from multiple key monitoring points along the line, not only is the information perception capability of the cathodic protection system significantly improved, but it also lays a reliable data foundation for subsequent adaptive regulation based on spatial feature interaction and confidence, enabling the entire solution to achieve a more scientific, efficient, energy-saving, and safer dynamic optimization protection effect.

[0018] In particular, S2 extracts the real-time pipe-to-ground potential of a target key monitoring point from the real-time pipe-to-ground potential of multiple key monitoring points as a data item to be verified. It should be understood that the pipe-to-ground potential data collected in real time from multiple key monitoring points reflects the spatial heterogeneity of the pipeline's protected status under different geographical locations and environmental conditions. In order to achieve accurate assessment and dynamic regulation of corrosion risks in specific areas or key areas, the technical solution of the present application extracts the real-time pipe-to-ground potential of a target key monitoring point from the real-time pipe-to-ground potential of multiple key monitoring points as a data item to be verified. By extracting the real-time pipe-to-ground potential of the target key monitoring point, it can serve as an important reference for subsequent spatial feature encoding and inference model verification, and through fusion with global information, a more accurate and confident dynamic judgment of the state of the point can be achieved. In this way, not only can the system's sensitivity to local risk changes be improved, but it can also ensure that the most realistic data is used as the basis for subsequent control strategy adjustments, thereby enhancing the protection effect and energy consumption optimization capabilities of the entire cathodic protection system under complex working conditions.

[0019] In particular, the S3, based on the global distribution of the real-time pipe-to-ground potential of the multiple key monitoring points, performs input data confidence calculation on the data item to be verified to obtain the input data confidence. In a specific example of the present application, Figure 3 As shown, the S3 includes: S31, based on the serial position of the data item to be verified in the real-time pipe-to-ground potential of the multiple key monitoring points, bidirectionally serially encoding the real-time pipe-to-ground potential of the multiple key monitoring points to obtain a real-time pipe-to-ground potential space forward propagation encoding vector and a real-time pipe-to-ground potential space backward propagation encoding vector; S32, performing bidirectional convergence reasoning on the real-time pipe-to-ground potential space forward propagation encoding vector and the real-time pipe-to-ground potential space backward propagation encoding vector to obtain the inferred real-time pipe-to-ground potential of the target key monitoring point; S33, calculating the proportional coefficient between the inferred real-time pipe-to-ground potential of the target key monitoring point and the real-time pipe-to-ground potential of the target key monitoring point as the input data confidence.

[0020] Specifically, in S31, based on the sequence position of the data item to be verified in the real-time pipeline-to-ground potential at the multiple key monitoring points, bidirectional sequence encoding is performed on the real-time pipeline-to-ground potential at the multiple key monitoring points to obtain a forward propagation encoding vector and a backward propagation encoding vector. In an embodiment of the present application, the real-time pipeline-to-ground potential at the multiple key monitoring points is first divided into two groups based on the position of the data item to be verified in the real-time pipeline-to-ground potential at the multiple key monitoring points to obtain forward and backward real-time pipeline-to-ground potential groups. It should be understood that as a typical linear structure, the electrochemical state at any location on a natural gas pipeline is affected by multiple factors, including the soil environment, current distribution, and stray interference in the adjacent sections. Relying solely on data from a single point cannot accurately reflect its true protection status. However, using data before the target point as the "forward group" and data after the target point as the "backward group" can systematically capture the flow of information and its temporal and spatial evolution along the pipeline. In the specific implementation process, it is first necessary to clearly define the index position of the target key monitoring point in the overall acquisition sequence. Using this index as the boundary, all real-time pipe-to-ground potential data before (or including itself) is grouped into the forward group, and data after it is grouped into the backward group. This processing flow not only improves the model's ability to perceive abnormal signals and sudden interference in complex scenarios, but also provides a solid information foundation for subsequent multi-scale hierarchical interaction, inference decoding, and confidence assessment.

[0021] Furthermore, it should be understood that within the cathodic protection system of a natural gas pipeline, changes in pipe-to-ground potential exhibit spatial continuity. The potential state at a target monitoring point is influenced not only by its own factors but also by the forward conduction of potential changes at upstream (power-side) monitoring points. Traditional single-point monitoring cannot capture this potential decay pattern along the pipeline (e.g., gradient changes due to soil resistivity differences or uneven anti-corrosion coatings), while environmental noise and sensor drift further obscure the true trend. Therefore, in the proposed technical solution, the monitoring sequence upstream of the target point is independently modeled to quantify the spatial propagation effect of upstream potential fluctuations on the target point. Specifically, the forward real-time pipe-to-ground potential group is passed through a forward LSTM-based sequence encoder to obtain a spatial forward propagation encoding vector for the real-time pipe-to-ground potential. Here, the forward LSTM encoder learns the nonlinear mapping relationship (e.g., distance-dependent exponential decay) from the upstream potential to the target point, identifying potential gradient changes caused by pipeline physical properties (e.g., pipe diameter, coating thickness). Furthermore, a gating mechanism filters out single-point data anomalies (e.g., transient sensor drift) and extracts spatially consistent trend features, improving the reliability of the input data.

[0022] Similarly, it should be understood that in the cathodic protection system of a natural gas pipeline, the distribution of the pipe-to-ground potential does not exist in isolation, but rather exhibits continuous changes along the pipeline space. The potential state of the target monitoring point is not only affected by the upstream (forward) monitoring point, but is also affected by the reverse propagation of the potential changes at the downstream (backward) monitoring point. Traditional single-point or unidirectional monitoring cannot capture this bidirectional spatial correlation. Especially in complex pipeline environments (such as branches, elbows, or areas with sudden changes in soil resistivity), abnormal fluctuations in the downstream potential (such as stray current interference or coating damage) may be reversely conducted to the target point through the electrolyte. Therefore, in the technical solution of the present application, the monitoring data downstream of the target point (backward group) is independently modeled to fully reflect the bidirectional dynamic propagation characteristics of the spatial potential field. Specifically, the backward real-time pipe-to-ground potential group is passed through a sequence encoder based on a backward LSTM to obtain a real-time pipe-to-ground potential spatial backward propagation encoding vector. The backward encoding vector captures the temporal patterns of downstream potential changes (such as attenuation gradients and abrupt phase changes), complementing the forward encoding. This allows the model to identify the propagation direction of potential anomalies (such as over- or under-protection). For example, a sudden drop in downstream potential may indicate an impending under-protection risk at a target point. Furthermore, by modeling the sequential dependencies of the backward groups through a reverse LSTM, single-point sensor noise or transient interference (such as lightning pulses) can be suppressed, allowing for the extraction of more statistically significant spatial trend features.

[0023] Specifically, the S32 performs bidirectional convergence reasoning on the real-time pipe-to-ground potential space forward propagation coding vector and the real-time pipe-to-ground potential space backward propagation coding vector to obtain the inferred real-time pipe-to-ground potential of the target key monitoring point. In a specific example of the present application, Figure 4 As shown, the S32 includes: S321, performing spatial bidirectional propagation multi-scale hierarchical interaction on the real-time pipe-ground potential spatial forward propagation coding vector and the real-time pipe-ground potential spatial backward propagation coding vector to obtain the real-time pipe-ground potential spatial bidirectional convergence propagation coding vector; S322, obtaining the inferred real-time pipe-ground potential of the target key monitoring point based on the real-time pipe-ground potential spatial bidirectional convergence propagation coding vector.

[0024] More specifically, in S321, the real-time pipeline-ground potential spatial forward propagation encoding vector and the real-time pipeline-ground potential spatial backward propagation encoding vector are subjected to spatial bidirectional propagation multi-scale hierarchical interaction to obtain the real-time pipeline-ground potential spatial bidirectional convergence propagation encoding vector. It should be understood that during natural gas pipeline operation, the environment is complex and changeable (such as soil moisture and stray current interference). Data from single or limited monitoring points are susceptible to noise and sensor drift, making it difficult to fully reflect the overall potential distribution along the pipeline. This can lead to delayed or inaccurate response of the PID control strategy, causing over-protection (energy waste) or under-protection (corrosion risk). Therefore, in the technical solution of the present application, the real-time pipeline-ground potential spatial forward propagation encoding vector and the real-time pipeline-ground potential spatial backward propagation encoding vector are subjected to spatial bidirectional propagation multi-scale hierarchical interaction to obtain the real-time pipeline-ground potential spatial bidirectional convergence propagation encoding vector. In other words, bidirectional encoding captures the dynamic characteristics of pipeline spatial propagation (e.g., the forward direction represents upstream potential propagation, and the backward direction represents downstream feedback) to cope with environmental noise and local anomalies, ensuring that the input features can more robustly represent the overall pipeline status. Specifically, the forward propagation encoding vector characterizes the cumulative effect of potential changes upstream of the target monitoring point, while the backward propagation encoding vector reflects the reverse effect of the downstream region on the target point. The bidirectional sequence encoding of the two forms a bidirectional dynamic model of potential spatial propagation. Through multi-scale hierarchical interaction, the model simultaneously leverages low-level detailed features (such as local potential mutations), mid-level structural features (such as potential gradients between adjacent monitoring points), and deep-level semantic features (such as the overall potential distribution pattern), thereby extracting robust global potential propagation patterns from noisy data. The generated real-time pipeline-to-ground potential spatial bidirectional convergent propagation encoding vector enables more accurate inference of the true potential state of the target monitoring point, reducing the risk of misjudgment due to local data anomalies. This approach not only improves the calculation accuracy of the input data confidence level but also provides a reliable basis for the subsequent dynamic parameter adjustment of the PID controller. Ultimately, precise control of the cathodic protection current is achieved, avoiding the over- or under-protection issues caused by data quality fluctuations in traditional methods, significantly improving pipeline corrosion prevention effectiveness and energy efficiency.

[0025] Specifically, a deep nonlinear transformation is first performed on the forward propagation encoding vector and the backward propagation encoding vector of the real-time pipeline-ground potential space to obtain the intermediate-layer implicit feature encoding vector, the intermediate-layer implicit feature encoding vector, the deep-layer implicit feature encoding vector, and the deep-layer implicit feature encoding vector. It is understood that the spatial propagation characteristics of the potential along the pipeline are affected by multiple factors, including soil resistivity distribution, stray current interference, and changes in ambient temperature and humidity. A single-scale feature representation is unable to fully capture the dynamic evolution of the potential. Traditional methods rely on direct input of raw potential data into the control module, but issues such as environmental noise and sensor drift can easily lead to fragile feature representations. Therefore, in order to explore spatial correlation patterns at different levels of abstraction, in the technical solution of the present application, multi-level implicit feature extraction is performed on the real-time pipe-ground potential space forward propagation coding vector and the real-time pipe-ground potential space backward propagation coding vector to obtain the real-time pipe-ground potential space forward propagation middle-layer implicit feature coding vector, the real-time pipe-ground potential space backward propagation middle-layer implicit feature coding vector, the real-time pipe-ground potential space forward propagation deep-layer implicit feature coding vector and the real-time pipe-ground potential space backward propagation deep-layer implicit feature coding vector.

[0026] Specifically, by leveraging the hierarchical abstraction capabilities of deep neural networks, feature representations with varying semantic granularity are extracted from the bidirectionally propagating potential sequence, thereby constructing a composite feature system that covers both local details and global correlations, providing a robust data foundation for subsequent multi-scale interactions. Specifically, the forward propagation encoding vector carries the historical accumulation of potential changes upstream of the target monitoring point, while the backward propagation encoding vector encodes the feedback effects of the downstream region. The mid-level implicit features of both capture structural connections between adjacent monitoring points (such as the spatial transmission pattern of potential gradients) through nonlinear transformations, while the deep-level implicit features extract long-range dependencies (such as the macroscopic trend of cross-regional potential balance). This hierarchical feature decoupling decomposes the complex potential propagation phenomenon into physically meaningful representations at different levels of abstraction. For example, mid-level features can characterize the potential distortion caused by sudden changes in local soil resistivity, while deep-level features can reflect the coordinated evolution of the polarization state across the entire pipeline. By separating features at different levels, the model can effectively suppress the interference of sensor noise on low-level details while preserving the stability of high-level semantics.

[0027] In a specific example of the present application, the real-time pipe-ground potential space forward propagation coding vector and the real-time pipe-ground potential space backward propagation coding vector are respectively subjected to deep nonlinear transformation using the following formula to obtain the real-time pipe-ground potential space forward propagation middle-layer implicit feature coding vector, the real-time pipe-ground potential space backward propagation middle-layer implicit feature coding vector, the real-time pipe-ground potential space forward propagation deep-layer implicit feature coding vector, and the real-time pipe-ground potential space backward propagation deep-layer implicit feature coding vector; wherein, the formula is: , , , , in, is the real-time tube-ground potential space forward propagation encoding vector, is the real-time tube-ground potential space back-propagation coding vector, represents the weight matrix of mid-level feature extraction, represents the bias term for mid-level feature extraction, express function, is the implicit feature encoding vector of the pressure-velocity coordinated deviation, is the target pressure gauge pressure mid-layer temporal implicit feature coding vector, represents the weight matrix for deep feature extraction, represents the bias term for deep feature extraction, is the deep implicit feature encoding vector of the pressure-flow rate collaborative deviation, is the deep temporal implicit feature encoding vector of the target pressure gauge pressure.

[0028] Next, a multi-level bidirectional fusion of the real-time pipeline-ground potential space forward propagation code vector and the real-time pipeline-ground potential space backward propagation code vector is performed to obtain a real-time pipeline-ground potential space low-level temporal fusion feature code vector, a real-time pipeline-ground potential space mid-level temporal fusion feature code vector, and a real-time pipeline-ground potential space deep-level temporal fusion feature code vector. In an embodiment of the present application, a low-level feature fusion of the real-time pipeline-ground potential space forward propagation code vector and the real-time pipeline-ground potential space backward propagation code vector is first performed to obtain a real-time pipeline-ground potential space low-level fusion feature code vector. It should be understood that potential fluctuations along a pipeline often exhibit complex spatiotemporal correlations, encompassing both local potential anomalies caused by coating damage or soil corrosion, and the global dynamic balance formed by current distribution among multiple monitoring points. Traditional methods directly use raw monitoring point data for control decisions, but ignore the detailed correlation modeling of high-frequency details in the bidirectional propagation potential sequence, resulting in a delayed system response to transient disturbances or local anomalies. Therefore, in order to establish an early interaction channel between bidirectional potential propagation paths and retain the original pattern of the high-frequency components of the potential fluctuation and the time-space correlation, in the technical solution of the present application, the real-time pipe-ground potential space forward propagation coding vector and the real-time pipe-ground potential space backward propagation coding vector are subjected to real-time pipe-ground potential space low-level feature fusion to obtain the real-time pipe-ground potential space low-level fusion feature coding vector.

[0029] That is, by directly fusing the two, the transient details and precise correspondences of the potential along the pipeline that are not filtered by the deep network abstraction layer during spatial propagation are captured, providing an undistorted underlying interaction foundation for subsequent multi-level fusion. Specifically, the forward propagation encoding vector records the temporal trajectory of the potential changes upstream of the target monitoring point, while the backward propagation encoding vector encodes the feedback effect of the downstream area. The low-level fusion of the two directly mines the synchronization of potential changes, phase differences, and other fine correlations between adjacent monitoring points through feature cross-operation. The fused encoding vector generated by low-level feature fusion can preserve the timestamp alignment features and spatial gradient details at the sensor level during the potential propagation process.

[0030] In a specific example of the present application, the real-time pipe-ground potential space forward propagation coding vector and the real-time pipe-ground potential space backward propagation coding vector are fused with the real-time pipe-ground potential space low-level feature to obtain the real-time pipe-ground potential space low-level fusion feature coding vector; wherein, the formula is: , in, It means adding by position. represents a multilayer perceptron, A low-level temporal fusion feature encoding vector of the target pressure gauge main behavior-cooperative behavior is obtained.

[0031] Next, the mid-level features of the real-time pipe-ground potential space forward propagation implicit feature coding vector and the mid-level implicit feature coding vector in the real-time pipe-ground potential space backward propagation are fused to obtain the mid-level fused feature coding vector in the real-time pipe-ground potential space. It should be understood that the propagation characteristics of the potential along the pipeline are affected by both the local soil resistance differences and the current distribution between adjacent areas, forming a complex spatial correlation. Traditional methods rely on feature analysis in a single direction or a single level, making it difficult to distinguish between true corrosion signals and potential fluctuations caused by environmental noise. Therefore, in order to establish a structural correlation model of the bidirectional potential propagation path, in the technical solution of the present application, the mid-level features of the real-time pipe-ground potential space forward propagation implicit feature coding vector and the mid-level implicit feature coding vector in the real-time pipe-ground potential space backward propagation are fused to obtain the mid-level fused feature coding vector in the real-time pipe-ground potential space.

[0032] That is, by integrating the abstract structural information that has been filtered through noise during the bidirectional propagation process, the potential laws of cross-regional potential balance are explored, thereby overcoming the risk of local misjudgment of single-directional features in complex electromagnetic environments. Specifically, the implicit features in the forward propagation layer may encode the abnormal transmission of potential gradients caused by upstream coating damage, while the implicit features in the backward propagation layer reflect the polarization compensation effect of the downstream sacrificial anode. The fusion of the two identifies the dynamic balance relationship of potential across regions through feature cross-attention or weight sharing mechanisms. The encoding vector generated by the fusion of mid-level features can reveal the systematic imbalance characteristics of the potential along the pipeline; at the same time, the mid-level fusion provides more physically meaningful intermediate features for data confidence calculation by filtering high-frequency noise and retaining stable spatial correlation patterns, supporting the robust decision-making of the control strategy under complex working conditions.

[0033] In a specific example of the present application, the following formula is used to perform mid-level feature fusion on the mid-level implicit feature coding vector in the forward propagation of the real-time pipe-ground potential space and the mid-level implicit feature coding vector in the backward propagation of the real-time pipe-ground potential space to obtain the mid-level fusion feature coding vector in the real-time pipe-ground potential space; wherein, the formula is: , , in, Indicates point multiplication by position, represents the attention mechanism, express function, represents the characteristic dimension of the vector, A hierarchical temporal fusion feature encoding vector for the target pressure gauge main behavior-cooperative behavior.

[0034] Furthermore, the real-time pipe-to-ground potential space forward propagation deep implicit feature coding vector and the real-time pipe-to-ground potential space backward propagation deep implicit feature coding vector are subjected to real-time pipe-to-ground potential space deep-level feature fusion to obtain the real-time pipe-to-ground potential space deep-level fusion feature coding vector. It should be understood that the global stability of the potential along the pipeline not only depends on the real-time data of the local monitoring points, but also requires grasping the synergistic relationship of the long-distance current distribution from a macroscopic level. The traditional method is limited to the shallow feature analysis of a single monitoring point and it is difficult to identify the potential risk of cross-regional potential imbalance. Therefore, in order to establish a global decision-making basis for the dynamic balance of pipeline potential, in the technical solution of the present application, the real-time pipe-to-ground potential space forward propagation deep implicit feature coding vector and the real-time pipe-to-ground potential space backward propagation deep implicit feature coding vector are subjected to real-time pipe-to-ground potential space deep-level feature fusion to obtain the real-time pipe-to-ground potential space deep-level fusion feature coding vector. Specifically, forward deep implicit features may encode the cumulative impact of soil resistivity changes within tens of kilometers upstream on the potential of the target point, while backward deep implicit features reflect the coordinated adjustment capabilities of the downstream constant potential instrument cluster. The fusion of the two reveals the critical threshold of the overall polarization state of the pipeline through semantic alignment and cross-domain reasoning. The encoding vector generated by deep-level feature fusion can predict the systematic imbalance trend of the potential along the pipeline. At the same time, deep fusion effectively filters local interference such as sensor drift by extracting highly abstract semantic features, providing a global verification basis for data confidence calculation, ensuring that PID parameter modulation responds to local anomalies while taking into account system stability, and ultimately achieving a global optimal balance between anti-corrosion effectiveness and energy efficiency.

[0035] In a specific example of the present application, the real-time pipe-ground potential space deep-level feature fusion is performed on the real-time pipe-ground potential space forward propagation deep implicit feature coding vector and the real-time pipe-ground potential space backward propagation deep implicit feature coding vector using the following formula to obtain the real-time pipe-ground potential space deep-level fusion feature coding vector; wherein, the formula is: , , in, represents the weight matrix of deep fusion, Representation layer normalization, express function, and is the low-rank projection matrix, A deep-level temporal fusion feature encoding vector of the target pressure gauge main behavior-cooperative behavior is provided.

[0036] Furthermore, a cross-level preliminary interactive fusion of the low-level fusion feature encoding vectors and the mid-level fusion feature encoding vectors in the real-time pipeline-ground potential space is performed based on order parameter phase complementarity to obtain a cross-level preliminary fusion feature encoding vector. It should be understood that pipeline potential data is affected by factors such as uneven soil resistivity and stray current interference. This leads to a complex spatiotemporal coupling between low-level features (such as instantaneous potential fluctuations at a single monitoring point) and mid-level features (such as regional potential gradient changes). Simple linear fusion ignores phase differences in information transmission between levels (such as the conflict between local abnormal fluctuations and overall trends), resulting in a fusion result that fails to accurately represent the pipeline's true corrosion status. Therefore, to dynamically coordinate the complementarity between local details and regional structure, in a preferred embodiment of the present application, a cross-level preliminary interactive fusion of the low-level fusion feature encoding vectors and the mid-level fusion feature encoding vectors in the real-time pipeline-ground potential space is performed based on order parameter phase complementarity to obtain a cross-level preliminary fusion feature encoding vector in the real-time pipeline-ground potential space.

[0037] Specifically, in pipeline monitoring scenarios, low-level features capture high-frequency noise or transient anomalies at the sensor level (such as current pulse interference), while mid-level features encode the potential propagation patterns between adjacent monitoring points (such as the polarization influence of upstream on downstream). By generating gated order parameters and utilizing a phase complementation mechanism (e.g., aligning the fluctuation features in the time and space domains), the contribution weights of the two types of features are adaptively adjusted. For example, when transient noise occurs at a monitoring point, the gating mechanism can reduce the interference of its low-level fluctuations on the mid-level regional model, preserving effective structural information and forming a more stable preliminary cross-level fusion feature of the real-time pipeline-ground potential space, laying the foundation for subsequent in-depth analysis.

[0038] In this example, the low-level fusion feature coding vector of the real-time pipe-ground potential space and the mid-level fusion feature coding vector of the real-time pipe-ground potential space are subjected to cross-level preliminary interactive fusion based on order parameter phase complementarity to obtain the cross-level preliminary fusion feature coding vector of the real-time pipe-ground potential space; wherein, the formula is: , in, represents the gating weight matrix, express function, represents the gating order parameter, A semantic feature encoding vector is preliminarily fused across levels for the target pressure gauge main behavior-collaborative behavior.

[0039] Subsequently, a deep interactive analysis of the real-time pipe-ground potential space's cross-level preliminary fusion feature encoding vectors and the real-time pipe-ground potential space's deep-level fusion feature encoding vectors, based on order parameter field constraints and path integral optimization, is performed to obtain a bidirectionally convergent and propagated encoding vector for the real-time pipe-ground potential space. It should be understood that when deep-level features (such as global protection effectiveness) interact with the preliminary fusion results, traditional methods are prone to suppressing the transmission of key information due to the accumulation of non-integrability in the order parameter field (constraint structure instability), thereby reducing the robustness of the encoding vectors. In a preferred embodiment of the present application, a deep interactive analysis of the real-time pipe-ground potential space's cross-level preliminary fusion feature encoding vectors and the real-time pipe-ground potential space's deep-level fusion feature encoding vectors, based on order parameter field constraints and path integral optimization, is performed to obtain a bidirectionally convergent and propagated encoding vector for the real-time pipe-ground potential space. In other words, order parameter field constraints ensure the stability of global semantic fusion.

[0040] Here, the deep-level features contain abstract semantics of the overall protection status of the pipeline (such as whether the cathodic protection of the entire line meets the standards). However, when directly interacting with the preliminary fused features, the non-ideal constraint structure of the weight matrix may lead to geometric phase accumulation in the path integral (distortion of feature interaction). This step suppresses the non-integrability of the constraint connection by decomposing the field path integral, and maintains the balance of the eigenvalue distribution with the help of the field uniform alignment loss function. It is worth mentioning that the preliminary interactive fusion suppresses the contamination of the regional model by local noise through the phase complementation mechanism (for example, filtering out abnormal gradients caused by single-point sensor drift), making the preliminary fused features more consistent with the physical propagation laws of the pipeline (such as the potential decay trend). Deep interactive analysis stabilizes the global semantic fusion through path integral optimization and field constraints, ensuring that deep features (such as corrosion status assessment) can be losslessly transferred to the final encoding vector.

[0041] In this example, the following formula is used to perform a deep interactive analysis of the real-time pipe-ground potential space based on order parameter field constraints and path integral optimization on the cross-level preliminary fusion feature coding vector and the deep-level fusion feature coding vector of the real-time pipe-ground potential space to obtain a bidirectional convergence propagation coding vector of the real-time pipe-ground potential space; wherein, the formula is: , , , , , in, represents the query weight matrix, represents the key weight matrix, represents the value weight matrix, represents the nuclear norm of the matrix, that is, the sum of the eigenvalues of the matrix, To scale hyperparameters, is the field uniform alignment loss function, represents the optimized query weight matrix, represents the optimized key weight matrix, represents the optimized value weight matrix, Indicates a query, represents the key vector, represents a value vector, A joint encoding vector for the target pressure gauge main behavior-cooperative behavior time series.

[0042] More specifically, S322 obtains the inferred real-time pipe-to-ground potential at the target key monitoring point based on the real-time pipe-to-ground potential spatial bidirectional convergence propagation code vector. Furthermore, due to factors such as uneven soil resistivity along the pipeline and interference from stray currents, the real-time pipe-to-ground potential at the target monitoring point is often contaminated by noise or sensor drift. While the bidirectional convergence propagation code vector generated in the previous step incorporates multi-scale spatial propagation patterns (such as the upstream polarization effect of forward propagation and the downstream feedback of backward propagation), it remains an abstract high-dimensional feature representation and cannot be directly used for current control decisions. Therefore, to achieve a precise mapping from the spatial propagation model to the local potential value, the technical solution of this application passes the real-time pipe-to-ground potential spatial bidirectional convergence propagation code vector through a decoder-based real-time pipe-to-ground potential model to obtain a decoded value. This decoded value is used to represent the inferred real-time pipe-to-ground potential at the target key monitoring point.

[0043] It is worth noting that a decoder usually refers to a neural network structure that can restore high-dimensional abstract representations to specific output results. Its typical working principle is that after receiving the output of the upstream encoder, it performs nonlinear transformation and reconstruction on the input features through a series of fully connected layers or recursive layers, and finally outputs a low-dimensional data item with clear physical meaning. In the specific implementation process, the real-time pipe-to-ground potential space bidirectional convergence propagation coding vector is first input into the trained decoding network. The network learns the implicit mapping relationship between "features-potentials" based on historical data, and through the adjustment of activation functions and weight parameters, it can achieve accurate inference of the current true pipe-to-ground potential value of the target key monitoring point. In this way, not only the system's perception of the actual operating status and anomaly detection capabilities are significantly improved, but also a solid foundation is laid for subsequent confidence calculation and PID parameter adaptive adjustment.

[0044] Specifically, the S33 calculates the proportional coefficient between the inferred real-time pipe-to-ground potential of the target key monitoring point and the real-time pipe-to-ground potential of the target key monitoring point as the input data confidence. It should be understood that the real-time pipe-to-ground potential data collected at multiple key monitoring points along the line are often affected by various uncertain factors such as changes in the on-site environment, sensor drift, noise interference, etc., which may cause some monitoring data to be abnormal or lack confidence. If these unscreened data are directly input into the PID controller for parameter adjustment, it is easy to cause problems such as overprotection or underprotection, thereby affecting the pipeline anti-corrosion effect and causing energy waste. Therefore, in the technical solution of the present application, the proportional coefficient between the inferred real-time pipe-to-ground potential of the target key monitoring point and the real-time pipe-to-ground potential of the target key monitoring point is calculated as the input data confidence, and the initial parameters of the PID controller are modulated based on the data confidence to obtain a fine-tuning PID controller.

[0045] That is, the confidence level of the input data is quantified by calculating the proportionality coefficient between the two. The closer the proportionality coefficient is to 1, the more consistent the measured value is with the model inference, indicating a high degree of data confidence. Conversely, a lower proportionality coefficient indicates that the measured value may have been subject to abnormal interference or deviated from normal operating conditions, resulting in a lower confidence level. This allows for dynamic determination of the reliability of the input signal, enabling the system to adaptively adjust the PID controller parameters based on the current data quality. Specifically, when the confidence level is high, the initial PID parameters can be maintained to fully utilize high-quality signals for precise control. When the confidence level decreases, the PID parameters are automatically fine-tuned (such as increasing the integral time and decreasing the proportional gain), enhancing the system's robustness to abnormal signals and low-confidence input scenarios, thereby preventing a single abnormality from causing the entire system to lose control. This not only improves the cathodic protection system's adaptability to complex operating conditions and emergencies, but also lays the foundation for achieving optimal protection effectiveness and energy consumption.

[0046] Specifically, S4 and S5 modulate the initial parameters of the PID controller based on data confidence to produce a fine-tuned PID controller. The real-time pipe-to-ground potential and the preset target pipe-to-ground potential at the target key monitoring point are then input into the fine-tuned PID controller to generate an output current control instruction for the potentiostat at the target key monitoring point. In other words, using the PID controller, which has undergone adaptive confidence optimization, the actual potential at the target key monitoring point is compared with the ideal protection target value pre-set in the engineering project. Based on the deviation between the two, the optimal adjustment amount is automatically calculated, and a specific, executable current control instruction for the potentiostat is output.

[0047] During the specific implementation process, the system first synchronously inputs the data currently collected at the target key monitoring points and the preset standard values into the fine-tuned PID control algorithm module. This module performs a comprehensive calculation of the output signal based on the real-time error (i.e., the difference between the actual value and the target value), combined with the proportional, integral, and differential parameters that have been dynamically adjusted based on the data confidence level. This generates a highly targeted and adaptive constant potential instrument output current instruction. This instruction is directly issued to the on-site constant potential instrument to achieve precise control of parameters such as the protection current size and action duration, ensuring that the pipeline surface is always in the optimal cathodic protection state. This not only greatly improves the robustness of the cathodic protection system to abnormal fluctuations and emergencies in complex operating environments, but also significantly improves energy utilization efficiency and the level of intelligent operation and maintenance, providing a solid technical guarantee for the safe and efficient operation of natural gas pipelines throughout their life cycle.

[0048] In summary, the method for cathodic protection of natural gas pipelines according to the embodiment of the present application is explained, which realizes high-precision reasoning of the state of the target monitoring point by spatial sequence encoding and multi-scale hierarchical interaction of the data of multiple monitoring points along the line, and further calculates the proportional coefficient between the inference value and the measured value to dynamically quantify the input data confidence; then, the input data confidence is directly involved in the PID parameter modulation as a weight, so that the PID controller can be adaptively adjusted according to the current input data quality; finally, the fine-tuned PID is used to generate the output instruction of the constant potential instrument to achieve the best cathodic protection effect and the best energy efficiency. In this way, the response capability to noise interference and sudden abnormal events under complex working conditions on site is significantly enhanced, thereby providing a more scientific, reasonable, energy-saving and efficient current control instruction for the output of the constant potential instrument, achieving the best protection effect and the best energy consumption balance.

[0049] While various embodiments of the present disclosure have been described above, the above descriptions are illustrative, non-exhaustive, and not intended to be limiting of the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is selected to best explain the principles of the embodiments, their practical applications, or improvements to existing technologies, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for cathodic protection of a natural gas pipeline, characterized in that: include: Obtain real-time pipe-to-ground potential from multiple key monitoring points along the pipeline; Extract the real-time pipe-to-ground potential of the target key monitoring point from the real-time pipe-to-ground potential of multiple key monitoring points as the data item to be verified; Based on the global distribution of the real-time pipe-to-ground potential of the multiple key monitoring points, performing input data confidence calculation on the data item to be verified to obtain input data confidence; Modulating the initial parameters of the PID controller based on the data confidence to obtain a fine-tuned PID controller; The real-time pipe-to-ground potential and the preset target pipe-to-ground potential of the target key monitoring point are input into the fine-tuning PID controller to obtain the output current control instruction of the constant potentiostat sent to the target key monitoring point.

2. The method for cathodic protection of a natural gas pipeline according to claim 1, characterized in that: Based on the global distribution of the real-time pipe-to-ground potentials of the multiple key monitoring points, input data confidence calculation is performed on the data item to be verified to obtain input data confidence, including: Based on the sequence position of the to-be-verified data item in the real-time pipe-to-ground potentials of the multiple key monitoring points, bidirectional sequence encoding is performed on the real-time pipe-to-ground potentials of the multiple key monitoring points to obtain a real-time pipe-to-ground potential space forward propagation encoding vector and a real-time pipe-to-ground potential space backward propagation encoding vector; Performing bidirectional convergence reasoning on the real-time pipe-to-ground potential space forward propagation coding vector and the real-time pipe-to-ground potential space backward propagation coding vector to obtain the inferred real-time pipe-to-ground potential of the target key monitoring point; A proportionality coefficient between the inferred real-time pipe-to-ground potential of the target key monitoring point and the real-time pipe-to-ground potential of the target key monitoring point is calculated as the input data confidence.

3. The method for cathodic protection of a natural gas pipeline according to claim 2, characterized in that: Based on the sequence position of the to-be-verified data item in the real-time pipe-to-ground potentials of the multiple key monitoring points, bidirectional sequence encoding is performed on the real-time pipe-to-ground potentials of the multiple key monitoring points to obtain a real-time pipe-to-ground potential space forward propagation encoding vector and a real-time pipe-to-ground potential space backward propagation encoding vector, including: Based on the positions of the real-time pipe-to-ground potentials of the multiple key monitoring points of the data item to be verified, the real-time pipe-to-ground potentials of the multiple key monitoring points are divided into two groups to obtain a forward real-time pipe-to-ground potential group and a backward real-time pipe-to-ground potential group; Performing forward sequence coding on the forward real-time tube-ground potential group to obtain a real-time tube-ground potential space forward propagation coding vector; The backward sequence coding is performed on the backward real-time tube-ground potential group to obtain the real-time tube-ground potential space backward propagation coding vector.

4. The method for cathodic protection of a natural gas pipeline according to claim 2, characterized in that: Performing bidirectional convergence reasoning on the real-time pipe-to-ground potential space forward propagation coding vector and the real-time pipe-to-ground potential space backward propagation coding vector to obtain the inferred real-time pipe-to-ground potential of the target key monitoring point, including: Performing spatial bidirectional propagation multi-scale hierarchical interaction on the real-time pipe-ground potential space forward propagation coding vector and the real-time pipe-ground potential space backward propagation coding vector to obtain the real-time pipe-ground potential space bidirectional convergent propagation coding vector; Based on the real-time pipe-ground potential spatial bidirectional convergence propagation coding vector, the inferred real-time pipe-ground potential of the target key monitoring point is obtained.

5. The method for cathodic protection of a natural gas pipeline according to claim 4, characterized in that: Performing spatial bidirectional propagation multi-scale hierarchical interaction on the real-time pipe-ground potential space forward propagation coding vector and the real-time pipe-ground potential space backward propagation coding vector to obtain the real-time pipe-ground potential space bidirectional convergent propagation coding vector, including: Performing deep nonlinear transformation on the real-time pipe-ground potential space forward propagation coding vector and the real-time pipe-ground potential space backward propagation coding vector respectively to obtain the real-time pipe-ground potential space forward propagation middle-layer implicit feature coding vector, the real-time pipe-ground potential space backward propagation middle-layer implicit feature coding vector, the real-time pipe-ground potential space forward propagation deep-layer implicit feature coding vector and the real-time pipe-ground potential space backward propagation deep-layer implicit feature coding vector; Performing real-time pipe-ground potential space multi-level bidirectional fusion on the forward propagation coding vector and the backward propagation coding vector of the real-time pipe-ground potential space to obtain a real-time pipe-ground potential space low-level temporal fusion feature coding vector, a real-time pipe-ground potential space middle-level temporal fusion feature coding vector, and a real-time pipe-ground potential space deep-level temporal fusion feature coding vector; The low-level temporal fusion feature coding vector of the real-time pipe-ground potential space, the medium-level temporal fusion feature coding vector of the real-time pipe-ground potential space and the deep-level temporal fusion feature coding vector of the real-time pipe-ground potential space are subjected to multi-scale hierarchical progressive complementary perception fusion in the real-time pipe-ground potential space to obtain the bidirectional convergence propagation coding vector of the real-time pipe-ground potential space.

6. The method for cathodic protection of a natural gas pipeline according to claim 4, characterized in that: Based on the real-time pipe-ground potential spatial bidirectional convergence propagation coding vector, the inferred real-time pipe-ground potential of the target key monitoring point is obtained, including: The real-time pipe-to-ground potential spatial bidirectional convergence propagation coding vector is passed through a real-time pipe-to-ground potential model based on a decoder to obtain a decoded value, which is used to represent the inferred real-time pipe-to-ground potential of the target key monitoring point.

7. The method for cathodic protection of a natural gas pipeline according to claim 5, characterized in that: The low-level temporal fusion feature coding vector of the real-time pipe-ground potential space, the mid-level temporal fusion feature coding vector of the real-time pipe-ground potential space, and the deep-level temporal fusion feature coding vector of the real-time pipe-ground potential space are subjected to multi-scale hierarchical progressive complementary perception fusion in the real-time pipe-ground potential space to obtain a bidirectional convergent propagation coding vector in the real-time pipe-ground potential space, including: Performing cross-level preliminary interactive fusion based on order parameter phase complementarity on the low-level fusion feature coding vector of the real-time pipe-ground potential space and the middle-level fusion feature coding vector of the real-time pipe-ground potential space to obtain the cross-level preliminary fusion feature coding vector of the real-time pipe-ground potential space; A deep interactive analysis of the real-time pipe-ground potential space cross-level preliminary fusion feature coding vector and the real-time pipe-ground potential space deep-level fusion feature coding vector is performed based on order parameter field constraints and path integral optimization to obtain the real-time pipe-ground potential space bidirectional convergence propagation coding vector.

8. The method for cathodic protection of a natural gas pipeline according to claim 1, characterized in that: The initial parameters of the PID controller are modulated based on the data confidence to obtain a fine-tuned PID controller, including: Extracting a proportional coefficient, an integral coefficient, and a differential coefficient from initial parameters of the PID controller; The proportional coefficient, the integral coefficient, and the differential coefficient are fine-tuned based on the data confidence to obtain an updated proportional coefficient, an updated integral coefficient, and an updated differential coefficient.

9. The method for cathodic protection of a natural gas pipeline according to claim 8, characterized in that: Fine-tuning the proportional coefficient, the integral coefficient, and the differential coefficient based on the data confidence to obtain an updated proportional coefficient, an updated integral coefficient, and an updated differential coefficient, comprising: The proportional coefficient, the integral coefficient, and the differential coefficient are fine-tuned based on the data confidence using the following formula: in, 、 and are the proportional coefficient, integral coefficient and differential coefficient, The confidence level of the data.

10. The method for cathodic protection of a natural gas pipeline according to claim 1, characterized in that: The real-time pipe-to-ground potential and the preset target pipe-to-ground potential of the target key monitoring point are input into the fine-tuning PID controller to obtain the output current control instruction of the constant potential instrument sent to the target key monitoring point, including: Calculate the error between the real-time pipe-to-ground potential of the target key monitoring point and the preset target pipe-to-ground potential; The error is input into the fine-tuning PID controller to obtain an output current control instruction of the potentiostat sent to the target key monitoring point.

Images