Intelligent Control Method and System for Tropical Harvesting Cables Based on Multi-Parameter Compensation Correction

By constructing a primary and backup heat tracing correlation model and multi-parameter compensation mapping learning, the problems of unstable heating and discontinuous data during fault switching of the heat tracing system were solved, and the safe and stable operation and monitoring accuracy of high-voltage electrical equipment were achieved.

CN123067313APending Publication Date: 2026-11-10国网山西省电力有限公司超高压变电分公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611258546.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-19
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing high-voltage electrical equipment heat tracing systems suffer from uneven heat supply transitions and discontinuous operational data during fault switching, leading to equipment safety risks and monitoring misjudgments, making it difficult to meet the high reliability operation and maintenance requirements of unattended substations.

Method used

By constructing a primary and backup heat tracing correlation model, generating a dynamic overlapping observation window, executing complementary power migration control, and performing multi-parameter compensation mapping learning, the system achieves uninterrupted and smooth switching of primary and backup heat tracing lines and data benchmark alignment, generating a continuous operating data sequence.

Benefits of technology

It achieves seamless and smooth switching of the heat tracing system, stably maintains SF6 gas pressure, avoids false alarms caused by data jumps, provides reliable operational data support, and improves the system's operational reliability and monitoring accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123067313A_ABST
    Figure CN123067313A_ABST
Patent Text Reader

Abstract

This invention discloses an intelligent control method and system for heat tracing cables based on multi-parameter compensation correction, belonging to the field of heat tracing control and operation data processing for high-voltage electrical equipment. The method includes: establishing a primary and backup heat tracing association model based on the configuration relationship of the primary and backup heat tracing cable devices corresponding to the high-voltage circuit breaker's gas chamber; acquiring and analyzing real-time detection data, generating an overlapping observation window when the primary heat tracing cable triggers a fault; activating complementary power migration control of the associated backup heat tracing cable within the overlapping observation window to acquire synchronous detection data of the primary and backup heat tracing cables; learning multi-parameter compensation mapping relationships from the detection data to obtain mapped detection data of the heat tracing cables; and performing continuous reconstruction based on the real-time detection data and the mapped detection data to obtain a heat tracing operation data sequence, which is then sent to the high-voltage circuit breaker fault detection processor. This application solves the technical problems of uneven heating transition and discontinuous operation data benchmarks during the fault switching process of existing primary and backup heat tracing cables, which easily lead to equipment safety risks and monitoring misjudgments.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of heat tracing control and operation data processing for high-voltage electrical equipment, specifically to an intelligent control method and system for heat tracing cables based on multi-parameter compensation and correction. Background Technology

[0002] In substations operating in frigid regions, tank-type SF6 circuit breakers rely on electric heating cables to continuously heat the outer wall of the gas chamber to prevent SF6 gas from liquefying in low temperatures, which could lead to a decline in insulation and breaking performance, and consequently cause power supply safety accidents such as low-pressure lockout and unplanned outages. To improve the fault tolerance of the heating system, existing solutions often configure a backup circuit for the main heating cable to achieve fault redundancy. However, the existing switching mechanism has significant defects: the commonly used hard switching mode of disconnecting before switching on can cause short-term temperature fluctuations in the tank due to the heating interruption during the switching gap and the difference in heating characteristics between the main and backup circuits. This poses a safety hazard of local SF6 liquefaction in extreme low temperatures. Furthermore, the independent data collection of the main and backup circuits can cause reference jumps and timing discontinuities in the monitoring sequence after switching, triggering false alarms from the upper-level high-voltage circuit breaker fault detection processor. At the same time, it disrupts the continuity of operating data, cannot support the analysis of equipment status trends, and is difficult to meet the high-reliability operation and maintenance requirements of unattended substations. Summary of the Invention

[0003] This application provides a method and system for intelligent control of heating cables based on multi-parameter compensation and correction, which aims to solve the technical problems of uneven heating transition and discontinuous operating data benchmarks during the fault switching process of existing main and backup heating cables, which can easily lead to equipment safety risks and monitoring misjudgments.

[0004] In view of the above problems, this application provides a method and system for intelligent control of tracing cables based on multi-parameter compensation correction.

[0005] The first aspect disclosed in this application provides a smart control method for tracing cables based on multi-parameter compensation correction, the method comprising: Based on the configuration relationship of the main and backup heat tracing devices corresponding to the high-voltage circuit breaker gas chamber, a main and backup heat tracing association model is established. The main and backup heat tracing devices include at least one main heat tracing cable and one backup heat tracing cable. Real-time detection data of the main heat tracing cable is acquired. By analyzing the real-time detection data of the main heat tracing cable, an overlapping observation window of the main and backup heat tracing cables is generated when the main heat tracing cable triggers a fault. According to the main and backup heat tracing association model, the associated backup heat tracing cable is simultaneously activated in the overlapping observation window of the main and backup heat tracing cables to perform complementary power migration control, and synchronous detection data of the main and backup heat tracing cables is acquired. Multi-parameter compensation mapping relationship learning is performed on the synchronous detection data of the main and backup heat tracing cables to acquire heat tracing cable mapping detection data. Based on the real-time detection data of the main heat tracing cable and the heat tracing cable mapping detection data, continuous reconstruction is performed to obtain the heat tracing operation data sequence, which is then sent to the high-voltage circuit breaker fault detection processor.

[0006] Another aspect of this application discloses an intelligent control system for heat tracing cables based on multi-parameter compensation and correction, the system comprising: The system comprises the following modules: a configuration module for establishing a primary and backup heat tracing association model based on the configuration relationship of the primary and backup heat tracing devices corresponding to the high-voltage circuit breaker's gas chamber, wherein the primary and backup heat tracing devices include at least one primary heat tracing device and one backup heat tracing device; an acquisition module for acquiring real-time detection data of the primary heat tracing device and generating an overlapping observation window for the primary and backup heat tracing devices when the primary heat tracing device triggers a fault, by analyzing the real-time detection data of the primary heat tracing device; a migration module for simultaneously activating the associated backup heat tracing device to perform complementary power migration control within the overlapping observation window of the primary and backup heat tracing devices according to the primary and backup heat tracing association model, thereby acquiring synchronous detection data of the primary and backup heat tracing devices; a compensation module for learning a multi-parameter compensation mapping relationship from the synchronous detection data of the primary and backup heat tracing devices, thereby acquiring heat tracing device mapping detection data; and a reconstruction module for continuously reconstructing the data based on the real-time detection data of the primary heat tracing device and the heat tracing device mapping detection data, thereby obtaining a heat tracing operation data sequence, which is then sent to the high-voltage circuit breaker fault detection processor.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: By constructing a primary and backup heat tracing correlation model to clarify the loop mapping relationship, setting a dynamically convergent overlapping observation window to provide transition timing, executing complementary power migration control to maintain constant heating power, conducting multi-parameter compensation mapping learning to achieve data benchmark alignment, and completing continuous reconstruction to output a unified monitoring sequence, the system ultimately achieved a smooth and uninterrupted switching of the primary and backup heat tracing cables. This not only ensured the stability of the total heating power of the gas chamber throughout the switching process from the heating perspective, completely eliminating the risk of temperature fluctuations during the switching gap, reliably maintaining SF6 gas pressure, and strengthening the safety defense line for equipment operation, but also generated a continuous heat tracing operation data sequence without breaks or benchmark jumps from the data perspective. This allows the upper-level fault detection processor to conduct status analysis based on a unified data benchmark, effectively avoiding false alarms and missed alarms caused by data mutations. At the same time, the complete and continuous operation data can retain the trend characteristics of the equipment, providing reliable data support for SF6 liquefaction prediction, full life assessment of heat tracing cables, and energy consumption statistics, thereby improving the operational reliability and monitoring accuracy of the heat tracing system.

[0008] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0009] Figure 1 A flowchart illustrating the intelligent control method for heat tracing cables based on multi-parameter compensation and correction provided in this application embodiment; Figure 2 This is a schematic diagram of the structure of the intelligent control system for tracing cables based on multi-parameter compensation and correction provided in an embodiment of this application.

[0010] Explanation of reference numerals in the attached diagram: Configuration module 11, Acquisition module 12, Migration module 13, Compensation module 14, Reconstruction module 15. Detailed Implementation

[0011] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0012] The overall concept of the technical solution provided in this application is as follows: This application provides a method and system for intelligent control of tracing cables based on multi-parameter compensation and correction. By constructing a primary and backup heat tracing correlation model, the mapping relationship between the circuit and the measuring point is clarified. After the primary heat tracing cable is triggered by a fault, a dynamically convergent overlapping observation window is generated to provide a parallel transition timing. Complementary power migration control is executed to maintain a constant total heating power in the gas chamber. Multi-parameter compensation mapping learning is carried out to align the backup circuit data to the primary circuit data representation space. Finally, after continuous reconstruction, a heat tracing operation data sequence without any breaks or reference jumps is output and sent to the high-voltage circuit breaker fault detection processor. This achieves a smooth and uninterrupted switching between the primary and backup heat tracing cables from the heating perspective, completely eliminating the hidden dangers of temperature fluctuations during the switching gap, stabilizing the SF6 gas pressure, and strengthening the safety defense of the circuit breaker equipment in cold environments. It also achieves transparency of the primary and backup switching to the upper-level monitoring system from the data perspective, avoiding false alarms and missed alarms caused by data jumps. At the same time, the continuous and complete operation data sequence can fully retain the equipment operation trend characteristics, which can provide reliable data support for SF6 liquefaction trend prediction, full life cycle assessment of the heat tracing cable, and energy consumption statistical analysis, thereby improving the operational reliability and intelligent operation and maintenance level of the heat tracing system.

[0013] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0014] Example 1, as Figure 1 As shown in the embodiment of this application, a smart control method for tracing cables based on multi-parameter compensation and correction is provided. The method includes: S100: Based on the configuration relationship of the main and backup heat tracing devices corresponding to the high-voltage circuit breaker gas chamber, establish a main and backup heat tracing association model. The main and backup heat tracing devices include at least one main heat tracing device and one backup heat tracing device.

[0015] Specifically, based on the physical configuration and circuit topology of the high-voltage circuit breaker gas chamber at the substation site, a primary and backup heat tracing correlation model is constructed. The high-voltage circuit breaker gas chamber refers to the sealed, pressure-bearing metal cavity inside the tank-type circuit breaker filled with SF6 arc-quenching and insulating gas, which is the core protected object for heat tracing and insulation. The heat tracing cable, by tightly adhering to its outer wall, raises the overall temperature of the cavity through heat conduction, preventing the liquefaction of SF6 gas in frigid environments from causing a sharp drop in insulation strength and breaking capacity. The primary and backup heat tracing cable devices are multiple heat tracing units configured for the same gas chamber, including at least one primary unit responsible for the daily heating load. The system includes a heat tracing cable and a backup heat tracing cable for redundancy. The main heat tracing cable is the heat tracing unit that is given priority to be put into operation under normal low temperature conditions and undertakes the main heating load. The backup heat tracing cable is a redundant unit that is automatically put into operation when the main heat tracing cable fails or the heating capacity is insufficient. Both use graphene composite heat tracing cables with a rated power of 2000W and AC220V independent power supply. The power specifications are completely consistent to ensure that the heating capacity is not reduced after switching. The configuration follows the field standard of two main and two backup heat tracing cables for circuit breaker air chambers of voltage level 220kV and above, and two main and one backup heat tracing cables for circuit breaker air chambers of voltage level 110kV and below. Specifically, the first step is to complete the site registration and labeling. For each SF6 gas chamber, the installation location of the heating cable, the power supply circuit number, and the control channel address are recorded according to phase and chamber number. Simultaneously, corresponding control output elements and status acquisition elements are matched for each main and backup heating cable. Each heating cable circuit is connected to one solid-state relay output terminal and one current transmitter acquisition channel in the local control cabinet. The solid-state relay is a contactless power electronic switching device that relies on the conduction and cutoff of semiconductor devices to achieve circuit switching. The current transmitter can convert the AC current of the heating cable working circuit into a 4-20mA standard analog signal and send it to the PLC controller. It can accurately capture minute changes in the circuit current and is used to identify heating cable disconnections, loose connections, power deviations, etc. in real time. Operational anomalies are the core data acquisition element for fault diagnosis. Simultaneously, each main and backup heat tracing device is bound to all sensor point attributes corresponding to the gas chamber, including the PT100 surface-mount temperature sensor on the tank surface, the SF6 gas pressure transmitter, outdoor ambient temperature measurement points, and unique equipment identifiers for temperature measurement points inside the local control box. The PT100 is a platinum resistance temperature sensor with a nominal resistance of 100Ω at 0℃, and its resistance changes approximately linearly with temperature. It boasts high measurement accuracy and strong low-temperature stability, making it the core sensing element for acquiring tank and ambient temperatures in frigid environments. Its acquired data is the core input basis for the temperature control logic. The SF6 gas pressure transmitter can acquire the gas pressure inside the gas chamber in real time, indirectly reflecting the degree of gas liquefaction, and is used for temperature control compensation and calibration.

[0016] Based on this, a hierarchical primary and backup heat tracing association model is constructed. The model adopts a three-level topology structure with the gas chamber ID as the primary key. The first level is the gas chamber entity layer, where each entry stores the basic attributes of the corresponding gas chamber, including voltage level, tank material, wall thickness, heat dissipation area, SF6 rated working pressure, liquefaction temperature threshold, and target temperature control range, serving as the basic calculation parameters for the temperature control strategy. The second level is the primary and backup circuit mapping layer, which is the core structure of the model. A directional mapping relationship is established between the primary heat tracing circuit set and the backup heat tracing circuit set for each gas chamber. For a gas chamber with two primary and two backup configurations, the circuit set includes two primary heat tracing circuits and two backup heat tracing circuits. The model defines that the operating priority of the primary circuit is higher than that of the backup circuit. Under normal operating conditions, the primary circuits are activated in order of priority. When a primary circuit fails or all primary circuits are running at full power but still cannot reach the target tank temperature, the backup circuits are activated in order of priority. The model uses circuits and labels each circuit with callable fields such as circuit type, rated current, cumulative running time, and PLC channel address. For a gas chamber with a two-main-one-standby configuration, the model maps one standby heating cable to a common standby unit of two main heating cables. Mutually exclusive triggering logic is set, meaning that a fault in any main circuit can trigger the standby circuit to be put into operation, but only one faulty main circuit can be provided with backup power at the same time to avoid overloading a single standby circuit. The model also stipulates that all main and standby heating cables corresponding to the same gas chamber share the tank temperature measurement point and SF6 pressure measurement point of that gas chamber as the temperature control feedback benchmark, ensuring that the temperature control target and feedback benchmark are completely consistent before and after the main-standby switchover. The third level is the measurement point association layer, which binds all sensing points such as temperature, pressure, current, and switch status corresponding to each gas chamber to the circuit set, clarifying the data ownership and calling logic of each measurement point, and providing a unified data entry for subsequent control algorithms.

[0017] Finally, the model is pre-set with operational constraint rules, including a power consistency rule, which ensures that the rated power of the standby heating cable is fully matched with that of the main heating cable to avoid sudden changes in heating capacity after switching, which could cause temperature fluctuations in the tank. A periodic rotation rule is also set, where the model accumulates the running time of each heating cable in real time. When the running time of the main heating cable reaches the set rotation cycle threshold, the main and standby roles are automatically switched to balance the service life of each heating cable and reduce the risk of burnout caused by long-term operation. A fault mutual exclusion verification rule is also set, which means that the main and standby circuits in the same air chamber cannot be judged as faulty at the same time. If this state occurs, it is judged as an abnormality in the common power supply circuit or sensor, which directly triggers a hierarchical alarm in the upper-level system.

[0018] S200: Acquire real-time detection data of the main and accompanying tropical zones, and generate an overlapping observation window of the main and backup tropical zones when the main and accompanying tropical zones trigger a fault by analyzing the real-time detection data of the main and accompanying tropical zones.

[0019] Specifically, the PLC local controller, deployed in a double-insulated local control box, acts as the execution entity, continuously acquiring real-time detection data of the main heat tracing cable at a fixed scanning cycle. This real-time detection data is a multi-dimensional, synchronously collected dataset covering the electrical operating status of the heat tracing cable itself and the corresponding air chamber's thermal conditions. It includes four core acquisition data streams: First, the operating current data of the main heat tracing cable circuit, which is acquired in real-time by current transmitters connected in series in each main heat tracing cable power supply circuit and converted into a 4-20mA standard analog signal input to the PLC analog signal channel. The current transmitter is a sensing device that linearly converts the large circuit current to a standard low-voltage signal, accurately capturing milliampere-level fluctuations in the circuit current. It is the core data source for identifying hardware faults such as heat tracing cable disconnections and loose connections; Second, the operating current data of the main heat tracing cable circuit... The terminal voltage data is synchronously acquired from the power supply bus by the voltage acquisition module. This is used to distinguish between faults in the heat tracing cable itself and faults in the upstream power supply circuit, avoiding misjudgment. The third data is the surface temperature data of the tank corresponding to the gas chamber, which is acquired by a patch-type PT100 temperature sensor attached to the heat tracing cable installation position on the outer wall of the gas chamber. The PT100 is a platinum resistance thermometer with a nominal resistance of 100Ω at 0℃. Its resistance value changes with temperature in a highly approximately linear manner. It has high measurement accuracy and strong long-term stability in the extremely cold operating conditions from -40℃ to 100℃, and can directly reflect the actual heat conduction effect of the heat tracing cable on the metal tank. The fourth data is the SF6 gas pressure data in the gas chamber, which is output by the pressure transmitter matched with the gas chamber. This can indirectly reflect the degree of gas liquefaction and is used to verify the actual pressure holding effect of the heat tracing system from the perspective of the essential requirements of equipment operation.

[0020] During the data acquisition process, the PLC performs third-order moving average filtering on the raw data to eliminate instantaneous data jumps caused by electromagnetic interference from the switchgear on site, ensuring the accuracy of subsequent fault diagnosis. Based on this, the PLC performs multi-dimensional cross-analysis of real-time detection data through its built-in fault judgment logic model. This fault judgment model adopts a two-layer progressive structure of single-parameter threshold triggering and multi-parameter interference verification. The first layer is the single-parameter threshold triggering layer, which sets three independent fault judgment threshold rules: the open circuit fault threshold is determined by the circuit current being less than 5% of the rated current for three consecutive sampling cycles, indicating that the circuit is completely disconnected; the loose connection fault threshold is determined by the circuit current fluctuation exceeding ±20% of the rated value for five consecutive sampling cycles, indicating abnormal power fluctuation caused by poor circuit contact; and the heating failure threshold is determined by the tank temperature showing a monotonous downward trend for 10 minutes and the SF6 pressure decreasing synchronously when the main heating cable is running at full power, indicating that the heating cable itself has failed to heat up. The second layer is the multi-parameter interference verification layer. When any fault threshold is triggered by a single-parameter rule, the model will simultaneously compare the bus voltage data with the outdoor ambient temperature data to eliminate external interference factors such as power supply voltage drops and normal temperature drops caused by extreme cold waves. Only after confirming that the fault is caused by the heating cable itself will the main heating cable be formally judged as having triggered a fault.

[0021] The moment the fault diagnosis model confirms a fault in the main heat tracing cable, the system immediately invokes the backup circuit mapping relationship of the corresponding air chamber in the main and backup heat tracing cable association model to generate an overlapping observation window for the main and backup heat tracing cables. This overlapping observation window is a manually set time interval between the confirmation of the main heat tracing cable fault and its complete shutdown, during which both the main and backup heat tracing cables are simultaneously energized and their operational data is collected synchronously. Its core value is to provide a heating transition buffer and mapping learning sample carrier for the main / backup switching process. The window adopts a time-series structure with a fixed starting point and a dynamically converged ending point. The overlapping observation starting point is fixed at the time node when the main heat tracing cable fault is officially triggered, i.e., the first PLC scan cycle when the fault diagnosis model outputs a valid fault signal. The overlapping observation ending point is initially set to the starting point... The preset duration node of 30 minutes after the point is used as the default length of the initial observation window. During the window operation, the system will synchronously start the associated backup heat tracing cable through solid-state relays to perform complementary power migration control. At the same time, it will synchronously align and sample the full parameter operation data of the main and backup heat tracing cables at a first sampling frequency that is 4 times higher than the normal sampling frequency. The first sampling frequency is set to complete the full parameter synchronous acquisition once every 250ms to obtain a sufficient density of time-series samples to support the subsequent learning of multi-parameter compensation mapping relationship. The end point of the window will not be fixed at the initial preset value. It will be dynamically adjusted backward according to the real-time learning error of the subsequent multi-parameter compensation mapping until the mapping error meets the preset error threshold before locking as the final end point of the window and ending the overlapping observation state.

[0022] S300: According to the main and backup heat tracing correlation model, the associated backup heat tracing cable is simultaneously activated in the overlapping observation window of the main and backup heat tracing cables to perform complementary power migration control and obtain synchronous detection data of the main and backup heat tracing cables.

[0023] Specifically, firstly, based on the high-voltage circuit breaker chamber ID to which the faulty main heat tracing cable belongs, the corresponding associated backup heat tracing cable is retrieved in the circuit mapping layer of the main and backup heat tracing association model. The associated backup heat tracing cable refers to an independent backup heating circuit that belongs to the same circuit breaker chamber as the faulty main heat tracing cable, has the same rated power, and is pre-bound with a redundant mapping relationship in the association model. Its power supply circuit and control channel are physically isolated from each other, and it can completely take over the heating load when the main circuit fails. During the retrieval process, the system synchronously retrieves the output channel address of the solid-state relay and the acquisition channel address of the current transmitter corresponding to the backup heat tracing cable, providing hardware addressing basis for subsequent control output and data acquisition. The solid-state relay is a contactless switching element that uses semiconductor devices to realize circuit switching. It supports millisecond-level response and high-frequency duty cycle power adjustment, and has no mechanical wear or arc interference. It is the core hardware carrier for realizing fine and smooth power adjustment. After replacing the traditional contactor, it can adapt to the temperature control requirements of frequent power adjustment.

[0024] Subsequently, the system officially started complementary power migration control. Complementary power migration control is a smooth switching control strategy with the core objective of keeping the total heating power of the gas chamber constant. Its core logic is to gradually reduce the output power of the faulty main heating cable, which is still in the energized state, according to the set slope during the parallel operation cycle of the overlapping observation window. At the same time, the associated backup heating cable synchronously increases the output power according to the corresponding slope. Throughout the process, the total heating power of the main and backup heating cables in the same gas chamber is completely consistent with the main circuit operating power before the fault. In specific implementation, the PLC achieves fine power adjustment by adjusting the duty cycle of the solid-state relay. It adopts a step rate of adjusting 10% of the rated power every 10 seconds to avoid thermal shock to the tank and current surge in the circuit caused by power change. This not only prevents the tank temperature from dropping suddenly and SF6 gas from liquefaction due to insufficient heating during the switching gap, but also avoids energy waste and device stress loss caused by power surge.

[0025] Throughout the parallel operation of the power migration, the system performs full-parameter synchronous acquisition at the first sampling frequency preset in the overlapping observation window to obtain synchronous detection data of the main and backup heat tracing cables. The first sampling frequency is a high-density sampling mode of 250ms / time, which is four times higher than the conventional operating sampling frequency. All measurement points use the unified system time base of the PLC for synchronous triggering sampling to ensure that each set of acquired data has a completely consistent timestamp and achieve strict time sequence alignment. The synchronous detection data of the main and backup heat tracing cables specifically includes seven core parameters at the same sampling time: the operating current of the main heat tracing cable, the terminal voltage of the main heat tracing cable, the operating current of the associated backup heat tracing cable, the terminal voltage of the associated backup heat tracing cable, the surface temperature of the corresponding gas chamber, the SF6 gas pressure, and the outdoor ambient temperature. All data are stored as structured samples paired by timestamp and continuously accumulated to form a complete synchronous detection dataset within the entire overlapping observation window. During the acquisition process, the system synchronously performs third-order moving average filtering and outlier removal on the raw data to eliminate data jumps caused by electromagnetic interference from the switchgear on site and ensure the accuracy and validity of the sample data.

[0026] S400: Perform multi-parameter compensation mapping relationship learning on the synchronous detection data of the main and backup tracing cables to obtain tracing cable mapping detection data.

[0027] Specifically, the system first performs time-series consistency verification and effective sample screening on the synchronous detection data of the main and backup heat tracing cables. The synchronous detection data of the main and backup heat tracing cables refers to the paired datasets collected synchronously with a high sampling frequency of 250ms within the overlapping observation window and based on the time base of the unified PLC system. Each set of samples carries a completely identical timestamp and contains six core parameters at the same time: the operating current of the main heat tracing cable, the operating current of the associated backup heat tracing cable, the bus terminal voltage, the tank surface temperature, the SF6 gas pressure, and the outdoor ambient temperature. During the screening process, the system adds the corresponding complementary power migration control operation stage identifier to all samples. The operation stage identifier is used to distinguish the power rise / fall transition segment and the power stable operation segment to which the data belongs. The system automatically filters out non-steady-state samples in the power dynamic adjustment stage and retains only effective samples in which the power of both the main and backup cables is in the steady-state range, thus constructing a main-backup mapping learning sample set. The main-backup mapping learning sample set is a supervised sample set used to train the mapping relationship. The corresponding parameters of the backup heat tracing cable are the model input features, and the corresponding parameters of the main heat tracing cable are the mapping target ground truth values. This ensures that the learning process always revolves around the characteristic differences under steady-state conditions and avoids mapping deviations caused by dynamic power fluctuations.

[0028] After completing the sample construction, the system performs multi-parameter compensation correction on the parameters of each dimension in the sample set. Multi-parameter compensation correction refers to the process of calibrating the original detection values ​​by introducing corresponding correction amounts for external influencing factors such as differences in the installation position of the heating cable, bus voltage fluctuations, and environmental heat dissipation interference. Specifically, for the current parameter, the ratio of the real-time bus terminal voltage to the rated voltage is used for conversion to compensate the actual working current into the equivalent heating current under the rated voltage, eliminating the power deviation caused by grid voltage fluctuations. For the tank temperature parameter, the tank heat transfer mathematical model based on Fourier's law of thermal conductivity is corrected to offset the influence of cold penetration at the temperature measurement point on the tank surface due to low ambient temperature. At the same time, the temperature difference caused by the relative position deviation between the temperature measurement point and the installation position of the heating cable is corrected to obtain the equivalent temperature rise data of the tank contributed only by the heating of the heating cable. For the SF6 gas pressure parameter, dynamic hysteresis compensation is performed in combination with the heat capacity parameter of the gas chamber metal tank to correct the pressure value deviation caused by the time difference of temperature conduction to the gas. The main and backup detection data after compensation are then obtained. It can eliminate systematic errors caused by differences in external operating conditions and installation, retaining only the inherent differences in the heating characteristics of the heating cable itself. On this basis, the system performs multi-parameter compensation mapping relationship learning. The multi-parameter compensation mapping relationship is a set of multi-dimensional nonlinear mapping functions. Its function is to project the operating parameters of the backup heating cable onto the data representation space of the main heating cable, so that the two have consistent numerical performance under the same heating effect. The learning process uses the compensated backup heating cable detection data as the input feature vector and the compensated main heating cable detection data at the same time point as the target output vector. The weighted least squares fitting method is used to iteratively solve the mapping coefficients of each parameter dimension. The mapping error is calculated once after each round of sample iteration. The mapping error refers to the root mean square deviation between the calculated value output by the mapping model and the actual compensated value of the main heating cable. It is used to measure the fitting accuracy of the mapping relationship. The system continuously uses the newly added samples accumulated in the overlapping observation window to iteratively optimize the mapping coefficients until the mapping error is less than or equal to the preset error threshold. The system is then judged to have converged and the model parameters are solidified.

[0029] Once the mapping model converges, the system inputs the real-time detection data of the backup tracing heat line into the trained multi-parameter compensation mapping model. After calculation and transformation, the system outputs the tracing heat line mapping detection data. The tracing heat line mapping detection data is virtual detection data equivalent to the main tracing heat line data benchmark. Its numerical magnitude, trend of change, and fluctuation characteristics are highly consistent with the data characteristics of the main tracing heat line during normal operation. It can directly continue the historical operating data of the main tracing heat line without numerical jumps.

[0030] S500: Based on the real-time detection data of the main heat tracing cable and the mapping detection data of the heat tracing cable, a continuous reconstruction is performed to obtain the heat tracing operation data sequence, which is then sent to the high-voltage circuit breaker fault detection processor.

[0031] Specifically, the system first extracts all real-time detection data of the main heat trough before the overlapping observation start point as the pre-reconstruction reference segment. This data is a measured dataset calibrated using third-order moving average filtering and parameter compensation, covering all parameters including the main heat trough's operating current, the surface temperature of the corresponding gas chamber tank, SF6 gas pressure, and outdoor ambient temperature. Each data point carries a precise timestamp, completely preserving the equipment's operating baseline and trend characteristics before the fault. Then, using the overlapping observation start point as the time-series connection node, the system sequentially appends the heat trough mapping detection data generated periodically within the overlapping observation window to the reference segment according to the timestamp order. This heat trough mapping detection data is obtained through a multi-parameter compensation mapping model. Equivalent data, having completed the alignment of the data representation space, has its parameter dimensions, numerical range, and change sensitivity all on the same benchmark as the measured data of the main and accompanying heat lines. It can directly inherit the data characteristics of the main and accompanying heat lines. To avoid minor numerical jumps at the connection points, the system performs continuous reconstruction processing. Continuous reconstruction refers to the process of using a unified time axis as a benchmark to splice the operating parameters of different collection objects and different data sources into a continuous time series without breaks or benchmark jumps through representation matching and weighted transition processing. Specifically, at the connection point, data from 5 sampling periods before and after the connection point are taken and time-weighted smoothing processing is performed, so that the weight of the preceding measured data gradually decreases over time, while the weight of the mapped data gradually increases synchronously, achieving a seamless transition between the two data segments.

[0032] After the time sequence splicing is completed, the system performs timestamp continuity verification and linear interpolation to fill in missing values ​​for the entire data segment, ensuring that each sampling time point corresponds to a complete set of full-dimensional parameters. This results in a heat tracing operation data sequence with a continuous time axis, unified data benchmark, and smooth trend. This sequence is equivalent to the monitoring data generated by the continuous and stable operation of the main heat tracing cable, and does not show any trace of main / backup switching. Subsequently, the system transmits this sequence to the high-voltage circuit breaker fault detection processor in real time via industrial Ethernet or wireless data transmission radio link. The high-voltage circuit breaker fault detection processor is a core processing unit deployed in the substation monitoring backend, responsible for real-time analysis of operational risks such as abnormal SF6 gas chamber pressure of tank circuit breakers, heat tracing system failure, and decreased equipment insulation performance. Its built-in fault analysis algorithm relies on a continuous and stable operating data sequence. Data gaps or numerical jumps can easily trigger false alarms or missed alarms.

[0033] Furthermore, in the method provided in this application embodiment, when the main accompanying tropical trough triggers a fault, an overlapping observation window of the main and backup accompanying tropical troughs is generated. The overlapping observation window of the main and backup accompanying tropical troughs includes an overlapping observation start point and an overlapping observation end point. The overlapping observation start point is the time node when the main accompanying tropical trough triggers a fault. The overlapping observation end point is the time node corresponding to satisfying a preset end point condition. The preset end point condition is the time node corresponding to when the mapping error of the multi-parameter compensation mapping relationship is less than or equal to a preset error threshold.

[0034] Specifically, the overlapping observation window for the main and backup heat tracing cables refers to a time interval artificially set from the triggering of a main heat tracing cable fault to its complete shutdown, during which the main and backup heat tracing cables are simultaneously energized and data is collected synchronously. Its purpose is to balance the smooth transition of heating power and the acquisition of training samples for the mapping model. The window includes two key time-series anchor points: the start point and the end point of the overlapping observation. First, the start point of the overlapping observation is strictly set to the time node when the main heat tracing cable fault is triggered. Here, the triggering of the main heat tracing cable fault is the first PLC scan cycle moment when a valid fault signal is officially output after single-parameter threshold triggering, multi-parameter interference verification, double-layer fault judgment model confirmation, elimination of external misjudgment factors such as bus voltage fluctuations and on-site electromagnetic interference. This moment is strictly aligned with the actual occurrence of the fault, ensuring that subsequent complementary power migration control and data acquisition are started from the first moment the fault occurs, avoiding a drop in tank temperature due to heating lag.

[0035] In the initial stage of window generation, the system sets the initial observation duration based on historical mapping learning experience of circuit breaker chambers of the same voltage level. Typically, 30 minutes after the start point is used as the initial overlapping observation endpoint, forming an observation window of initial size. This provides an initial sample accumulation interval for training the multi-parameter compensation mapping relationship. During the window operation, the system performs full-parameter synchronous sampling of the operating data of the main heat tracing cable under fault conditions and the associated standby heat tracing cable at a first sampling frequency. The first sampling frequency is a high-density sampling mode that is much higher than the conventional operating sampling frequency, initially set to 250ms / time. All sampled data are based on the time base aligned with the timestamp of the PLC unified system. After multi-dimensional parameter compensation calibration, they form the main and standby mapping learning samples, which are used to iteratively train the multi-parameter compensation mapping relationship. After each round of full-sample iterative training, the system calculates the mapping error of the current mapping model. The mapping error refers to the root mean square deviation between the standby equivalent data output by the mapping model and the actual compensation detection data of the main heat tracing cable. It is used to quantify the degree of fit of the mapping relationship to the heating characteristics and installation position differences of the main and standby heat tracing cables. The smaller the value, the higher the consistency between the mapped data and the original data benchmark of the main circuit.

[0036] The system pre-sets a corresponding preset error threshold as the criterion for judging the preset endpoint condition. This threshold is the maximum allowable mapping deviation set according to the accuracy requirements of the high-voltage circuit breaker fault detection processor for data continuity. It is usually set to 2% of the rated parameter value. When the mapping error is less than or equal to this threshold, it is determined that the accuracy of the mapping model has met the requirements for subsequent data continuity reconstruction. The corresponding time node at this time is the final overlapping observation endpoint. If the accumulated samples in the current window are insufficient to allow the mapping error to converge to within the threshold, the system will prioritize increasing the first sampling frequency within the sampling frequency limit range. By increasing the sample density per unit time, the model convergence speed is accelerated. The sampling frequency limit is the highest sampling rate supported by the PLC analog signal acquisition channel hardware, which is usually 100ms / time. When the sampling frequency has reached the hardware limit and still cannot meet the error requirements, the system automatically postpones the time node for initializing the overlapping observation endpoint, extends the overlapping observation duration to accumulate more training samples, and continues iterative training until the mapping error converges to within the preset error threshold, locking the final overlapping observation endpoint.

[0037] Furthermore, in the method provided in this application embodiment, the overlapping observation endpoint is a time node corresponding to a preset endpoint condition. The method includes: initializing the overlapping observation endpoint based on the overlapping observation start point, and generating an initialization primary and backup tracing heat wave overlapping observation window; sampling data of the primary tracing heat wave and the associated backup tracing heat wave at a first sampling frequency within the initialization primary and backup tracing heat wave overlapping observation window to obtain primary tracing heat wave fault sampling data and associated backup tracing heat wave sampling data; performing multi-parameter compensation mapping relationship learning based on the primary tracing heat wave fault sampling data and associated backup tracing heat wave sampling data, and calculating the mapping error until convergence; when the mapping error is greater than the preset error threshold, optimizing and adjusting the time node of the initialization overlapping observation endpoint until the mapping error is less than or equal to the preset error threshold.

[0038] Specifically, the system first uses the anchored starting point of the overlapping observation, i.e. the time node when the main heat tracing cable fault is officially triggered, as the time reference. Combined with the historical mapping training experience of the gas chamber of the tank circuit breaker of the same voltage level, a fixed parallel duration of 30 minutes is preset to complete the initial setting of the overlapping observation endpoint. Then, the system generates the initial main and backup heat tracing cable overlapping observation window. The initial main and backup heat tracing cable overlapping observation window is the initial main and backup parallel operation sequence interval preset by the system. It is used to provide the basic sample accumulation time for the first round of training of the multi-parameter compensation mapping model. During the window operation, the faulty main heat tracing cable remains energized and the associated backup heat tracing cable gradually increases its power according to the slope. The two provide parallel heating to avoid the risk of SF6 liquefaction caused by a sudden drop in tank temperature.

[0039] Subsequently, during the entire initialization window, the system synchronously triggers sampling of all dimensions of the main and backup circuits at the first sampling frequency. The first sampling frequency is a high-frequency acquisition mode that is much higher than the sampling density of normal operation, usually set to 250ms / time. All measurement points are triggered for sampling based on the millisecond-level unified system time base inside the PLC, ensuring that the timestamps of the two sets of data are strictly aligned. Finally, the main heat tracing cable fault sampling data and the associated backup heat tracing cable sampling data are output respectively. The main heat tracing cable fault sampling data covers the operating current of the main circuit after the fault, the bus terminal voltage, and the surface temperature of the corresponding PT100 tank of the gas chamber, SF6 gas pressure and other core parameters, which completely preserve the heating characteristics and thermal conduction characteristics of the main circuit under fault conditions. The associated backup heat tracing cable sampling data corresponds to the same operating parameters of the backup circuit at the same time. The two are paired one by one according to the timestamp to form a supervised learning sample set.

[0040] After completing the first round of sample accumulation, the system first performs multi-dimensional pre-compensation processing on the two sets of sampled data, including bus voltage fluctuation compensation, temperature measurement point position deviation correction, and outdoor environmental heat dissipation interference cancellation, to eliminate systematic errors caused by differences in external operating conditions and installation. Then, it conducts multi-parameter compensation mapping relationship learning, that is, iteratively solves the coefficients of the multi-dimensional nonlinear mapping function through a weighted least squares fitting algorithm to establish a transformation relationship from the parameter representation space of the backup heat tracing cable to the parameter representation space of the main heat tracing cable. After each round of full-sample iterative training, the mapping error is calculated. The mapping error refers to the root mean square deviation between the equivalent main circuit data output by the mapping model and the actual sampled data of the main heat tracing cable. It is used to quantify the fitting accuracy of the mapping relationship to the heating characteristics of the main and backup cables and the differences in installation positions. The smaller the value, the better the continuity of subsequent data reconstruction.

[0041] Subsequently, the real-time calculated mapping error is compared with the preset error threshold. The preset error threshold is the maximum allowable deviation pre-set based on the accuracy requirements of the high-voltage circuit breaker fault detection processor for data continuity. It is usually 2% of the rated parameter value. If the current mapping error is greater than this threshold, it means that the sample size in the initial window is insufficient to support the convergence of the mapping model to the accuracy requirements. The system then postpones the time node of the initialization of the overlapping observation endpoint, extends the window running time to continuously accumulate new sampling samples, and iterates and updates the mapping model coefficients and recalculates the mapping error for each new pair of samples. The closed-loop process of sampling, learning, and verification is executed in a loop until the mapping error is less than or equal to the preset error threshold. At this point, the current moment is immediately locked as the final overlapping observation endpoint, and the dynamic adjustment process of the window ends.

[0042] Furthermore, in the method provided in this application embodiment, when the mapping error is greater than the preset error threshold, the first sampling frequency is preferentially optimized under the sampling frequency limit; until the optimized sampling frequency is the sampling frequency limit, the time node of the initial overlapping observation endpoint is optimized and adjusted.

[0043] Specifically, when the calculated mapping error exceeds the preset error threshold, optimization is carried out following the control logic of prioritizing frequency increase and using delay as a fallback. The preset error threshold is the maximum allowable mapping deviation pre-set based on the accuracy requirements of the high-voltage circuit breaker fault detection processor for data continuity; it is typically set to 2% of the rated operating parameter value and is the core criterion for determining whether the mapping model meets the accuracy requirements of subsequent data reconstruction. During optimization, the system prioritizes frequency optimization of the first sampling frequency. This first sampling frequency refers to the high-density sampling rate used to collect synchronous operation data of the main and backup heat exchangers within the overlapping observation window. The initial default setting is 250ms / sample, significantly higher than the sampling density under normal operating conditions, used to obtain sufficient paired training samples within a limited window duration to accelerate model convergence. Frequency optimization uses a 50ms sampling interval. The step value gradually shortens the sampling period and increases the number of sampling points per unit time. After each frequency adjustment, the system re-accumulates master-slave paired samples based on the new sampling density, iteratively trains the multi-parameter compensation mapping relationship, and recalculates the mapping error to verify whether the accuracy requirements are met. The entire frequency increase process is strictly constrained by the sampling frequency limit. The sampling frequency limit is the maximum allowable sampling rate determined by the hardware AD conversion rate of the XD-E8AD-A analog expansion module matched with the PLC, the signal response time of the Acrel BA05AI / I type current transmitter, and the thermal inertia of the surface-mount PT100 temperature sensor. In this solution, the limit is 100ms / time. Exceeding this limit will cause a surge in sampling data noise and distortion of temperature measurement data tracking, which will reduce the accuracy of mapping learning. Therefore, the system will not exceed this hardware performance limit.

[0044] Subsequently, when the first sampling frequency has been optimized and increased to the sampling frequency limit, further increasing the frequency would damage the data quality, and the mapping error still has not met the preset requirements, the system starts the second level of optimization. The time node for initializing the overlapping observation endpoint is postponed, and the total duration of the overlapping observation window is extended in fixed steps of 5 minutes. More samples of the synchronous steady-state operation of the main and backup heat tracing cables are continuously accumulated. For each step extension, the mapping model coefficients are updated using the newly added samples and the error is re-verified. The closed-loop process of sampling, learning, and verification is executed in a loop until the mapping error is less than or equal to the preset error threshold, at which point the adjustment stops and the final overlapping observation endpoint is locked.

[0045] Furthermore, in the method provided in this application embodiment, multi-parameter compensation mapping relationship learning is performed on the synchronous detection data of the primary and backup tracing cables to obtain tracing cable mapping detection data. The method includes: performing time-series alignment on the primary tracing cable fault detection data and associated backup tracing cable detection data in the synchronous detection data of the primary and backup tracing cables to generate a primary-backup mapping learning sample set; obtaining multi-parameter detection sample data corresponding to the same sampling time node; compensating the primary-backup mapping learning sample set according to the multi-parameter detection sample data to obtain primary tracing cable fault compensation detection data and associated backup tracing cable compensation detection data; using the associated backup tracing cable compensation detection data as mapping input and the primary tracing cable fault compensation detection data at the same sampling time node as mapping target, multi-parameter compensation mapping relationship learning is performed to obtain tracing cable mapping detection data.

[0046] Specifically, the first step is to perform time-series alignment processing on the main and backup heating cable synchronous detection data, specifically the main heating cable fault detection data and the associated backup heating cable detection data. The main and backup heating cable synchronous detection data consists of two independent datasets acquired within a window using a high-frequency sampling mode. The main heating cable fault detection data includes full-dimensional operating parameters such as the main circuit's operating current, bus terminal voltage, and the tank surface temperature and SF6 gas pressure of the corresponding gas chamber after a fault. The associated backup heating cable detection data consists of parameters from the same source collected by the backup circuit within the same time period. Time-series alignment refers to matching the data from the two different acquisition channels one by one according to the same timestamp, using the unified millisecond-level system time base within the PLC as the benchmark, and eliminating invalid samples with time misalignment or missing sampling. Finally, a point-by-point paired main and backup mapping learning sample set is generated. The main and backup mapping learning sample set is a structured supervised sample set used to train the mapping relationship. Each set of samples corresponds to the exact same operating time, ensuring that the learning process is based on a consistent external environment and heating conditions.

[0047] After completing the sample set construction, the system extracts multi-parameter detection sample data corresponding to the same sampling time node. The multi-parameter detection sample data covers two major dimensions: electrical operation and thermal conduction. Specifically, it includes five core parameters: the operating current of the main and backup heat tracing cables, the bus terminal voltage, the tank surface temperature collected by the patch PT100, the gas chamber pressure value output by the SF6 gas pressure transmitter, and the outdoor ambient temperature. Among them, the patch PT100 is a platinum resistance temperature measuring element attached to the heat tracing area on the outer wall of the tank. At 0℃, its nominal resistance is 100Ω and the resistance changes approximately linearly with temperature. It has strong measurement stability in the extremely cold operating conditions from -40℃ to 100℃ and is the core carrier for collecting the thermal state of the tank. Its temperature measurement data is easily affected by the penetration of low temperature in the environment and the deviation of the installation position, resulting in numerical deviation. It is the key calibration target in the parameter compensation process.

[0048] Subsequently, the system performs item-by-item compensation and correction on the bilateral data in the primary and backup mapping learning sample sets based on multi-parameter detection sample data, obtaining primary heating cable fault compensation detection data and associated backup heating cable compensation detection data. The compensation and correction are carried out one by one for three types of systematic deviations: First, voltage fluctuation compensation, which combines the ratio of real-time bus voltage to rated voltage to convert the actual operating current into the equivalent heating current under rated voltage, eliminating the power calculation deviation caused by grid voltage fluctuations; Second, temperature measurement position compensation, which corrects the relative position deviation between the temperature measurement point and the heating cable installation position and the temperature measurement deviation caused by the infiltration of cold air from the outdoor environment based on the tank heat transfer mathematical model constructed based on Fourier's law of thermal conductivity, obtaining the tank equivalent temperature rise data contributed only by the heating cable; Third, pressure lag compensation, which corrects the response lag of SF6 gas pressure to temperature changes by combining the tank metal heat capacity parameter, so that the pressure data and instantaneous heating power achieve precise time-series matching. The two sets of data after compensation can eliminate the common deviations caused by external operating conditions and installation differences. To eliminate inherent differences in the heating characteristics and heat transfer efficiency of the main and backup heat tracing cables, the system compensates and corrects these differences. Then, it uses the associated backup heat tracing cable compensation detection data as the mapping input feature vector and the main heat tracing cable fault compensation detection data from the same sampling time point as the mapping target true value. This process involves learning a multi-parameter compensation mapping relationship, a set of multi-dimensional nonlinear mapping functions. The function is to project the operating parameters of the backup heat tracing cable onto the data representation space of the main heat tracing cable. The learning process uses a weighted least squares fitting method to iteratively solve the mapping coefficients for each parameter dimension. After each full-sample iteration, the mapping error is calculated and the coefficient values ​​are optimized until the error converges to a preset threshold range. Finally, the real-time compensation data from the backup heat tracing cable is substituted into the trained mapping model to output the heat tracing cable mapping detection data. This data is highly consistent with the normal operating data of the main heat tracing cable in terms of numerical magnitude, trend, and fluctuation characteristics, and is equivalent to the detection data generated by the continuous and stable operation of the main circuit.

[0049] Furthermore, in the method provided in this application embodiment, after generating the primary-backup mapping learning sample set, the method further includes: adding a running stage identifier to the primary-backup mapping learning sample set according to the power migration stage corresponding to the complementary power migration control; and filtering out the primary-backup mapping learning sample set according to the running stage identifier to obtain a valid primary-backup mapping learning sample set.

[0050] Specifically, the power migration stage is first divided based on the execution logic of complementary power migration control. Complementary power migration control refers to a control strategy that uses solid-state relay duty cycle adjustment to achieve a smooth transition between main and backup power at a step rate of 10% of rated power every 10 seconds. Throughout the process, the total heating power of the gas chamber is kept consistent with the operating power of the main circuit before the fault. The entire parallel operation process is divided into two categories according to the power state: dynamic power migration stage and steady-state parallel operation stage. The dynamic power migration stage corresponds to the transition period when the backup heating cable starts up and increases power while the main heating cable synchronously decreases power. During this stage, the main and backup power continuously changes, and the tank temperature and SF6 pressure are in a transient adjustment process. The data contains a large amount of operating condition fluctuation interference. The steady-state parallel operation stage corresponds to the period when the main and backup power are both adjusted to the target value and remain constant, and the thermal state of the tank gradually tends to stabilize. The data in this stage can truly reflect the inherent heating and heat transfer characteristics differences between the main and backup heating cables.

[0051] The system then adds a running stage identifier to each sample in the initial master-slave mapping learning sample set. The running stage identifier is a classification label that is bound to the sample timestamp and is used to mark the power migration stage to which the sample belongs. The label determination is based on the deviation between the power set value of the master and slave heat tracing cables and the steady-state target value at the current sampling time: when the power of the slave heat tracing cable does not reach the target value or the power of the master heat tracing cable does not drop to the target exit value, the corresponding sample is labeled as a dynamic migration stage; when the power of both is stable at the target value and continues for more than 3 sampling periods, the corresponding sample is labeled as a steady-state running stage.

[0052] After labeling is completed, the system performs targeted filtering on the initial sample set based on the operation stage identifier, directly removing all transient samples marked as dynamic migration stage, and retaining only samples marked as steady-state operation stage. Finally, an effective master-slave mapping learning sample set is obtained. The effective master-slave mapping learning sample set is a high-quality training sample set that removes the interference of operating condition fluctuations and contains only steady-state operating condition pairing data. It can accurately map the characteristic differences of the master and slave heating cables under the same heating load, and avoid the thermal lag and power fluctuation interference of transient processes on the accuracy of mapping fitting.

[0053] Furthermore, in the method provided in this application embodiment, the multi-parameter compensation mapping relationship is used to map the associated backup tracing cable compensation detection data to the data representation space where the main tracing cable fault compensation detection data is located; the mapping error is calculated based on the difference between the mapping output of the multi-parameter compensation mapping relationship and the main tracing cable fault compensation detection data; when the mapping error is less than the preset error threshold, the tracing cable mapping detection data that has a continuous data representation relationship with the main tracing cable fault sampling data is obtained using the multi-parameter compensation mapping model.

[0054] Specifically, the operation is executed by the PLC local controller unit deployed in the double-layer insulated local control box. The core function is to achieve benchmark alignment of primary and backup data through a unified transformation of the data representation space. The multi-parameter compensation mapping relationship is essentially a set of iteratively trained multi-dimensional nonlinear mapping functions. Its core function is to project the associated backup heating cable compensation detection data, which has completed corrections for voltage fluctuations, temperature measurement positions, and pressure hysteresis, onto the data representation space where the primary heating cable fault compensation detection data resides. The data representation space refers to a standardized data benchmark system composed of the unified numerical range, parameter response sensitivity, thermal conduction hysteresis characteristics, and historical trend baseline corresponding to the primary heating cable's operating data. The upper-level high-voltage circuit breaker fault detection processor performs equipment status analysis based on this fixed benchmark space. If the input data deviates from this benchmark space, it will directly... This leads to deviations in the judgment logic, triggering false alarms or missed alarms; after each round of mapping coefficient iteration, the full amount of associated backup heat tracing cable compensation detection data is substituted into the current mapping function for transformation to obtain the mapping output value corresponding to each sampling time. Then, the mapping output value at each time is compared with the main heat tracing cable fault compensation detection data under the same sampling timestamp dimension by dimension to calculate the mapping error. Since there are significant differences in the dimensions, ranges and numerical magnitudes of the three types of parameters, namely current, tank surface temperature and SF6 gas pressure, the range extreme value normalization processing of each dimension parameter needs to be performed before calculating the mapping error. Different physical quantities are uniformly mapped to the dimensionless interval of [0,1] of the high-voltage electrical equipment heat tracing control and operation data processing. After eliminating the dimensional differences, weighted synthesis is performed. The normalization formula is: ; In the formula, x is the actual detected value of the corresponding parameter after compensation and correction; These are the upper and lower limits of the measurement range for each parameter; where: the working current range of the heating cable is taken from 0 to the rated working current; the tank surface temperature range is taken from the local extreme minimum ambient temperature to the upper limit of the target temperature control of the gas chamber; and the SF6 gas pressure range is taken from the low gas pressure lock-up pressure of the equipment to the rated working pressure. Based on the safety risk level of the high-voltage circuit breaker operation, the expert weighting method is used to assign error weights to the three types of parameters, and the sum of the weights is 1, with priority tilted towards the core safety parameters.

[0055] Among them, the surface temperature of the tank directly reflects the thermal state of the gas chamber and is directly related to the risk of SF6 liquefaction in the heat tracing control and operation data processing of high-voltage electrical equipment, with a weight of 0.45; SF6 gas pressure is the core indicator for judging the insulation and breaking performance of the equipment, with a weight of 0.35; and the working current of the heat tracing cable is used to characterize the electrical operating status of the circuit, with a weight of 0.20.

[0056] The mapping error uses the weighted average root mean square error as a unified quantitative indicator, comprehensively considering the overall mapping deviation of multiple dimensions of parameters. The calculation formula is as follows: ; In the formula: N is the total number of valid samples, and M is the total number of parameter dimensions. In this scheme... ; The error weights corresponding to the j-th dimension parameters; The normalized mapping output value of the j-th dimension parameter of the i-th sample; This is the true value of the j-th dimension parameter of the i-th sample after normalization; this formula, by taking the root mean square after weighted square error for each sample, can comprehensively reflect the overall fitting accuracy of the mapping model across all samples and dimensions.

[0057] The system pre-sets a corresponding preset error threshold as the accuracy judgment benchmark. This threshold is a dimensionless value, corresponding to the upper limit of the normalized comprehensive deviation. The preset threshold of this scheme is 0.02 for high-voltage electrical equipment heat tracing control and operation data processing, which is equivalent to 2% of the high-voltage electrical equipment heat tracing control and operation data processing deviation relative to the rated range.

[0058] In the model training and accuracy verification stages, the time-series hold-out method is used to divide the effective steady-state sample set, strictly distinguishing the computational scope of the training set and the validation set to avoid data leakage: according to the sampling time sequence, the steady-state samples of the first 70% of the high-voltage electrical equipment heat tracing control and operation data processing are taken as the training set, which is only used to iteratively solve the coefficients of the multi-parameter compensation mapping function to drive the model iterative convergence; the remaining steady-state samples of the last 30% of the high-voltage electrical equipment heat tracing control and operation data processing are used as an independent validation set, which does not participate in the coefficient iterative solution, but is only used to evaluate the generalization performance of the model.

[0059] The RMSE (Recovery Mean Squared Error) of the integrated high-voltage electrical equipment heat tracing control and operation data processing in the training set serves only as an intermediate criterion for judging model convergence during iteration. The RMSE of the integrated high-voltage electrical equipment heat tracing control and operation data processing in the independent validation set serves as the formal criterion for judging the final accuracy of the model. The system strictly avoids directly using training error as a conclusion for generalization accuracy. Only when the RMSE of the integrated high-voltage electrical equipment heat tracing control and operation data processing in the independent validation set is less than or equal to a preset error threshold can it be determined that the accuracy of the mapping model meets the requirements for engineering applications. If the training set error has converged but the validation set error exceeds the threshold, it indicates that the model has the risk of overfitting. It is necessary to extend the overlapping observation window duration, add new steady-state samples, re-divide the dataset, and retrain until the validation set accuracy meets the standard.

[0060] Subsequently, the mapping error is the comprehensive root mean square deviation calculated by allocating coefficients based on the safety impact weights of different parameters such as current, tank surface temperature, and SF6 gas pressure. Among them, tank temperature and SF6 gas pressure are assigned higher weights because they are directly related to the insulation and arc extinguishing performance of the circuit breaker, followed by current. The system has a corresponding preset error threshold, which is set according to the minimum accuracy requirement of the high-voltage circuit breaker fault detection processor for data continuity. It is usually 2% of the rated parameter value. If the value exceeds this threshold, the processor will identify it as data abnormality.

[0061] When the calculated mapping error is greater than the preset error threshold, the system determines that the model has not converged and continues to accumulate new steady-state samples to iteratively optimize the mapping coefficients. When the mapping error is less than or equal to the preset error threshold, the system determines that the fitting accuracy of the mapping model meets the requirements. The system immediately solidifies the coefficients of the mapping function to form a stable multi-parameter compensation mapping model. Subsequently, the real-time collected compensation detection data of the associated backup heat tracing cable is continuously input into the solidified model. After conversion and calculation, the heat tracing cable mapping detection data is output. This data has a continuous data representation relationship with the main heat tracing cable fault sampling data. That is, the output mapping data is completely connected with the main heat tracing cable operation data before the fault in terms of numerical magnitude, slope of change trend, fluctuation amplitude, and response characteristics to changes in ambient temperature. There are no numerical jumps or trend breaks. It is equivalent to the detection data generated by the continuous normal operation of the main heat tracing cable.

[0062] Furthermore, in the method provided in this application embodiment, a heat tracing operation data sequence is obtained by continuously reconstructing the real-time detection data of the main heat tracing cable and the mapping detection data of the heat tracing cable. The method includes: if the current time node is the time node corresponding to the overlapping observation endpoint, disconnecting the circuit corresponding to the main heat tracing cable so that the circuit corresponding to the associated backup heat tracing cable operates independently and obtaining independent detection data of the associated backup heat tracing cable; and continuously reconstructing the real-time detection data of the main heat tracing cable, the mapping detection data of the heat tracing cable, and the independent detection data of the associated backup heat tracing cable according to the time nodes to obtain the heat tracing operation data sequence.

[0063] Specifically, the process is executed by a PLC local controller deployed in a double-layer insulated local control box. After the overlapping observation window dynamically converges and the mapping accuracy meets the standard, the hardware loop switching and data sequence reconstruction are completed synchronously. During implementation, the system uses the internal millisecond-level unified system time base as a benchmark to compare the current timestamp with the locked overlapping observation endpoint in real time. The overlapping observation endpoint is the time node when the mapping error of the multi-parameter compensation mapping relationship first converges to the preset error threshold, which is also the termination time of the main and backup parallel heating phase. When the two completely coincide, the system outputs a cutoff control signal to the solid-state relay corresponding to the main heating cable circuit, and reliably disconnects the power supply to the main heating cable by relying on the contactless disconnection of the semiconductor device. To prevent secondary risks such as burnout and leakage caused by the main heat tracing cable being energized for a long time due to a fault, the associated backup heat tracing cable officially enters an independent operating state at this time, independently undertaking the entire heating load of the corresponding circuit breaker gas chamber. Its operating parameters are continuously collected by the matching Acrel BA05AI / I type current transmitter, tank surface patch PT100 temperature sensor, and SF6 gas pressure transmitter. After voltage fluctuation correction, temperature measurement position deviation correction, and pressure hysteresis compensation processing consistent with the pre-compensation link, independent detection data of the associated backup heat tracing cable is formed. This data is also input in real time into the solidified converged multi-parameter compensation mapping model and converted into equivalent characterization data consistent with the main heat tracing cable data benchmark.

[0064] The system then performs continuous reconstruction processing. This continuous reconstruction refers to the process of stitching together and integrating operational data from different sources and stages in chronological order, using a unified time axis as a benchmark. Weighted smoothing is then used to eliminate numerical jumps at the junctions, ultimately forming a continuous time sequence without breaks or benchmark jumps. To ensure the security of the monitoring logic and the traceability of faults, the smoothed sequence generated by the reconstruction must not replace the original operational records and fault event information. The system adopts a dual-channel parallel output mechanism, strictly distinguishing between continuous data on the thermal state of the gas chamber and fault events in the heat tracing circuit. The smoothed sequence obtained through continuous reconstruction serves only as a unified benchmark data source for trend analysis of the thermal state of the gas chamber and prediction of SF6 liquefaction, and is sent to the status analysis channel of the high-voltage circuit breaker fault detection processor. The original measurement data, the fault flag of the main heat tracing cable, the real-time mapping error value of the current mapping model, and the mapping validity status identifier are simultaneously sent to the event alarm channel of the fault detection processor through an independent parallel channel. The two types of data are independent and do not overlap, preventing the smoothed equivalent normal sequence from masking the actual faults that have occurred in the main heat tracing cable, and ensuring that circuit fault events can be independently identified and alarm triggered.

[0065] For real-time linear point filling and smoothing at the reconstructed connection points, the system adopts a fixed-length historical cache window mechanism. The point filling and smoothing operations only rely on the historical valid sampling points before the current moment and do not call any sample data from future moments. The cache window retains by default the measured data and mapping data of 5 historical sampling cycles before and after the connection node. The numerical transition at the connection point is completed through a time-weighted smoothing algorithm. The entire process is based solely on the historical data that has been collected, ensuring that the reconstructed logic meets the requirements for real-time online operation and has no dependence on future data. It can be directly adapted to the real-time scanning operation mode of the PLC local controller for high-voltage electrical equipment heat tracing control and operation data processing.

[0066] During reconstruction, the system splices three data segments in chronological order: The first segment is the full real-time detection data of the main heat tracing cable before the start of the overlapping observation, i.e., the measured dataset of the main circuit operating normally before the fault occurred, serving as the preceding baseline segment of the sequence, completely preserving the parameter trends and historical characteristics of the equipment's normal operation; the second segment is the heat tracing cable mapping detection data generated within the overlapping observation window, i.e., the equivalent main circuit data obtained by mapping and transforming the backup circuit data during the parallel phase, serving as a transition segment to fill the time interval of the main circuit gradually exiting operation after the fault; the third segment is the independent detection data of the associated backup heat tracing cable after the end of the overlapping observation and after transformation by the mapping model, serving as the subsequent continuation segment of the sequence, continuously outputting the changing trend of the transition segment; at the connection node between adjacent segments, the system takes data from 5 sampling periods before and after and performs time-weighted smoothing processing, so that the weight of the preceding data gradually decreases over time and the weight of the subsequent data gradually increases synchronously, completely eliminating the small numerical jumps at the connection point, and finally forming a heat tracing operation data sequence with continuous time axis, unified data benchmark, and smooth changing trend, which is synchronously sent to the high-voltage circuit breaker fault detection processor.

[0067] Furthermore, in the method provided in this application embodiment, the multi-parameter detection sample data includes at least the gas chamber surface temperature, SF6 gas pressure, main heat tracing cable operating current, and associated backup heat tracing cable operating current.

[0068] Specifically, the multi-parameter detection sample data is the smallest complete data unit that constitutes the master-slave mapping learning sample set. At each sampling moment within the overlapping observation window, the data is synchronously triggered by the PLC local controller deployed in the double-layer insulated local control box based on the internal millisecond-level unified system time base, covering at least four types of parameters: gas chamber surface temperature, SF6 gas pressure, main heat tracing cable working current, and associated backup heat tracing cable working current.

[0069] The surface temperature of the air chamber is collected by a patch-type PT100 temperature sensor attached to the installation side of the heating cable on the outer wall of the high-voltage circuit breaker air chamber. The patch-type PT100 is a thin-film platinum resistance temperature measuring element with a nominal resistance of 100Ω at 0℃. It can be closely attached to the surface of the metal tank to achieve high-precision contact temperature measurement. Its output resistance signal is connected to the XD-E6PT-P resistance temperature expansion module matched with the PLC. After analog-to-digital conversion, it outputs the real-time temperature value, which can directly reflect the actual effect of the heating cable conducting heat to the metal structure of the tank. It is the core parameter characterizing the overall thermal state of the air chamber.

[0070] The SF6 gas pressure is collected by the pressure transmitter matching the gas chamber body. The pressure transmitter is a sensing device that linearly converts the gas pressure in the closed gas chamber into a 4-20mA standard analog signal, and its output signal is connected to the analog input channel of the PLC. This parameter can indirectly reflect the liquefaction degree of the SF6 gas in the gas chamber, is a core safety indicator for judging whether the insulation performance and breaking capacity of the circuit breaker meet the standards, and can also be used to verify the accuracy of the tank temperature conduction calculation and correct the temperature measurement deviation caused by thermal hysteresis.

[0071] The working current of the main heating tape and the working current of the associated standby heating tape are respectively collected by the Acrel BA05AI / I current transmitters connected in series at the power supply ends of the two loops. The current transmitters can linearly convert the power frequency large current in the loops into a standard weak current analog signal suitable for PLC collection, which is sent to the controller through the XD-E8AD-A analog expansion module, and can accurately capture the milliamp-level tiny fluctuations of the loop current. It can not only be directly converted to obtain the actual operating power of the heating tape, but also serve as a basis for identifying fault states such as wire breakage, virtual connection, and power deviation.

[0072] The collection actions of the four types of parameters are synchronously triggered by the PLC under the same sampling clock, and each set of sample data corresponds to a unique timestamp, ensuring that the electrical parameters and thermal parameters under the same sampling node correspond strictly, and the working conditions of the main and standby loop data are completely consistent, which provides basic data matched with working conditions for subsequent multi-parameter compensation correction and mapping relationship learning.

[0073] In summary, the intelligent heating tape control method based on multi-parameter compensation and correction provided by the embodiments of the present application has the following technical effects: By constructing the main and standby heating association model to clarify the loop mapping relationship, setting the dynamically converging overlapping observation window to provide transition time sequence, executing complementary power transfer control to maintain constant heating power, carrying out multi-parameter compensation mapping learning to realize data reference alignment, and completing continuity reconstruction to output a unified monitoring sequence, the undisturbed smooth switching of the main and standby heating tapes is finally realized. On the heating level, it ensures the stability of the total heating power of the gas chamber during the whole switching process, completely eliminates the risk of temperature fluctuation in the switching gap, reliably maintains the SF6 gas pressure, and builds a solid safety line for equipment operation. On the data level, it generates a continuous heating operation data sequence without breakpoints and reference jumps, so that the upper fault detection processor can carry out state judgment based on a unified data reference, effectively avoiding false alarms and missed alarms caused by data mutations. At the same time, the complete and continuous operation data can retain the equipment trend characteristics, provide reliable data support for SF6 liquefaction prediction, heating tape full-life assessment and energy consumption statistics, and improve the operation reliability and monitoring accuracy of the heating system.

[0074] The second embodiment is based on the same inventive concept as the intelligent heating tape control method based on multi-parameter compensation and correction in the previous embodiments, such as Figure 2As shown in the embodiment of this application, a smart control system for heat tracing cables based on multi-parameter compensation and correction is provided. The system includes: The configuration module 11 is used to establish a primary and backup heat tracing association model based on the configuration relationship of the primary and backup heat tracing devices corresponding to the high-voltage circuit breaker gas chamber. The primary and backup heat tracing devices include at least one primary heat tracing device and one backup heat tracing device. The acquisition module 12 is used to acquire real-time detection data of the primary heat tracing device and generate an overlapping observation window of the primary and backup heat tracing devices when the primary heat tracing device triggers a fault by analyzing the real-time detection data of the primary heat tracing device. The migration module 13 is used to simultaneously start the associated backup heat tracing device to perform complementary power migration control in the overlapping observation window of the primary and backup heat tracing devices according to the primary and backup heat tracing association model, and acquire synchronous detection data of the primary and backup heat tracing devices. The compensation module 14 is used to learn the multi-parameter compensation mapping relationship of the synchronous detection data of the primary and backup heat tracing devices and acquire heat tracing mapping detection data. The reconstruction module 15 is used to perform continuous reconstruction based on the real-time detection data of the primary heat tracing device and the heat tracing mapping detection data to obtain the heat tracing operation data sequence and send it to the high-voltage circuit breaker fault detection processor.

[0075] Furthermore, the acquisition module 12 is also used to perform the following steps: when the main and backup tropical tropics trigger a fault, a main and backup tropical tropics overlapping observation window is generated. The main and backup tropical tropics overlapping observation window includes an overlapping observation start point and an overlapping observation end point. The overlapping observation start point is the time node when the main and backup tropical tropics trigger a fault. The overlapping observation end point is the time node that meets the preset end point condition. The preset end point condition is the time node that corresponds to when the mapping error of the multi-parameter compensation mapping relationship is less than or equal to a preset error threshold.

[0076] Furthermore, the acquisition module 12 is also used to perform the following steps: initialize the overlapping observation endpoint based on the overlapping observation starting point, and generate an initial primary and backup accompanying heat wave overlapping observation window; within the initial primary and backup accompanying heat wave overlapping observation window, sample the primary accompanying heat wave and the associated backup accompanying heat wave at a first sampling frequency to obtain primary accompanying heat wave fault sampling data and associated backup accompanying heat wave sampling data; perform multi-parameter compensation mapping relationship learning based on the primary accompanying heat wave fault sampling data and associated backup accompanying heat wave sampling data, and calculate the mapping error until convergence; when the mapping error is greater than the preset error threshold, optimize and adjust the time node of the initial overlapping observation endpoint until the mapping error is less than or equal to the preset error threshold.

[0077] Furthermore, the acquisition module 12 is also used to perform the following steps: when the mapping error is greater than the preset error threshold, the first sampling frequency is optimized first under the sampling frequency limit; until the optimized sampling frequency is the sampling frequency limit, the time node of the initial overlapping observation endpoint is optimized and adjusted.

[0078] Furthermore, the compensation module 14 is also used to perform the following steps: time-series alignment of the main and backup tracing cable fault detection data and the associated backup tracing cable detection data in the synchronous detection data of the main and backup tracing cables to generate a main-backup mapping learning sample set; obtaining multi-parameter detection sample data corresponding to the same sampling time node; compensating the main-backup mapping learning sample set according to the multi-parameter detection sample data to obtain main tracing cable fault compensation detection data and associated backup tracing cable compensation detection data; using the associated backup tracing cable compensation detection data as the mapping input and the main tracing cable fault compensation detection data at the same sampling time node as the mapping target to learn the multi-parameter compensation mapping relationship and obtain tracing cable mapping detection data.

[0079] Furthermore, the compensation module 14 is also used to perform the following steps: adding a running stage identifier to the primary-backup mapping learning sample set according to the power migration stage corresponding to the complementary power migration control; filtering out the primary-backup mapping learning sample set according to the running stage identifier to obtain a valid primary-backup mapping learning sample set.

[0080] Furthermore, the compensation module 14 is also used to perform the following steps: the multi-parameter compensation mapping relationship is used to map the associated backup heat tracing cable compensation detection data to the data representation space where the main heat tracing cable fault compensation detection data is located; the mapping error is calculated based on the difference between the mapping output of the multi-parameter compensation mapping relationship and the main heat tracing cable fault compensation detection data; when the mapping error is less than the preset error threshold, the heat tracing cable mapping detection data that has a continuous data representation relationship with the main heat tracing cable fault sampling data is obtained using the multi-parameter compensation mapping model.

[0081] Furthermore, the reconstruction module 15 is also used to perform the following steps: if the current time node is the time node corresponding to the overlapping observation endpoint, disconnect the circuit corresponding to the main heat tracing cable so that the circuit corresponding to the associated backup heat tracing cable can operate independently and obtain independent detection data of the associated backup heat tracing cable; continuously reconstruct the real-time detection data of the main heat tracing cable, the mapping detection data of the heat tracing cable, and the independent detection data of the associated backup heat tracing cable according to the time node to obtain the heat tracing operation data sequence.

[0082] Furthermore, the compensation module 14 is also used to perform the following steps: the multi-parameter detection sample data includes at least the gas chamber surface temperature, SF6 gas pressure, main heat tracing cable operating current and associated backup heat tracing cable operating current.

[0083] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A smart control method for tracing cables based on multi-parameter compensation and correction, characterized in that, The method includes: Based on the configuration relationship of the main and backup heat tracing devices corresponding to the gas chamber of the high-voltage circuit breaker, a main and backup heat tracing association model is established. The main and backup heat tracing devices include at least one main heat tracing device and one backup heat tracing device. Acquire real-time detection data of the main and backup tropics, and generate an overlapping observation window of the main and backup tropics when the main and backup tropics trigger a fault by analyzing the real-time detection data of the main and backup tropics. According to the primary and backup heat tracing correlation model, complementary power migration control is simultaneously activated in the primary and backup heat tracing overlap observation window to obtain synchronous detection data of the primary and backup heat tracing. Multi-parameter compensation mapping relationship learning is performed on the synchronous detection data of the main and backup tracing cables to obtain tracing cable mapping detection data; Based on the real-time detection data of the main heat tracing cable and the mapping detection data of the heat tracing cable, a continuous reconstruction is performed to obtain the heat tracing operation data sequence, which is then sent to the high-voltage circuit breaker fault detection processor.

2. The intelligent control method for tracing cables based on multi-parameter compensation and correction as described in claim 1, characterized in that, When the main and backup tropical tropics trigger a fault, an overlapping observation window for the main and backup tropical tropics is generated. The overlapping observation window for the main and backup tropical tropics includes an overlapping observation start point and an overlapping observation end point. The overlapping observation start point is the time node when the main and backup tropical tropics trigger a fault. The overlapping observation endpoint is the time node that meets the preset endpoint condition, which is the time node when the mapping error of the multi-parameter compensation mapping relationship is less than or equal to a preset error threshold.

3. The intelligent control method for tracing cables based on multi-parameter compensation and correction as described in claim 2, characterized in that, The overlapping observation endpoint is the time node that meets the preset endpoint conditions. The method includes: Based on the aforementioned overlapping observation starting point, the overlapping observation ending point is initialized, and the initialization window for the primary and backup tropical overlapping observations is generated. Within the initial primary and backup heat line overlap observation window, data are sampled for the primary heat line and the associated backup heat line at a first sampling frequency to obtain primary heat line fault sampling data and associated backup heat line sampling data. Based on the main cable malfunction sampling data and the associated backup cable malfunction sampling data, a multi-parameter compensation mapping relationship is learned, and the mapping error is calculated until convergence. When the mapping error is greater than the preset error threshold, the time node of the initial overlapping observation endpoint is optimized and adjusted until the mapping error is less than or equal to the preset error threshold.

4. The intelligent control method for tracing cables based on multi-parameter compensation and correction as described in claim 3, characterized in that, When the mapping error is greater than the preset error threshold, the first sampling frequency is optimized first under the sampling frequency limit. The time node of the initial overlapping observation endpoint is optimized and adjusted until the optimized sampling frequency reaches the sampling frequency limit.

5. The intelligent control method for tracing cables based on multi-parameter compensation and correction as described in claim 3, characterized in that, The method involves learning a multi-parameter compensation mapping relationship from the synchronous detection data of the primary and backup tracing cables to obtain tracing cable mapping detection data. The fault detection data of the main heat trough and the associated backup heat trough detection data in the synchronous detection data of the main and backup heat troughs are time-series aligned to generate a main and backup mapping learning sample set. Acquire multi-parameter detection sample data corresponding to the same sampling time point; Based on the multi-parameter detection sample data, the primary and backup mapping learning sample sets are compensated respectively to obtain primary tropical tractor fault compensation detection data and associated backup tropical tractor compensation detection data. Using the associated backup tracing cable compensation detection data as the mapping input, and the main tracing cable fault compensation detection data at the same sampling time node as the mapping target, multi-parameter compensation mapping relationship learning is performed to obtain tracing cable mapping detection data.

6. The intelligent control method for tracing cables based on multi-parameter compensation and correction as described in claim 5, characterized in that, After generating the master-slave mapping learning sample set, the method also includes: Based on the power migration stage corresponding to the complementary power migration control, add a running stage identifier to the primary / backup mapping learning sample set; The primary-backup mapping learning sample set is filtered out according to the operation stage identifier to obtain a valid primary-backup mapping learning sample set.

7. The intelligent control method for tracing cables based on multi-parameter compensation and correction as described in claim 5, characterized in that, The multi-parameter compensation mapping relationship is used to map the associated backup heat tracing cable compensation detection data to the data representation space where the main heat tracing cable fault compensation detection data is located. The mapping error is calculated based on the difference between the mapping output of the multi-parameter compensation mapping relationship and the fault compensation detection data of the main and accompanying tropical lines. When the mapping error is less than or equal to the preset error threshold, the mapping detection data of the tracing tropics that has a continuous data representation relationship with the fault sampling data of the main tracing tropics is obtained by using the multi-parameter compensation mapping model.

8. The intelligent control method for tracing cables based on multi-parameter compensation and correction as described in claim 2, characterized in that, Based on the real-time detection data of the main heat tracing cable and the mapping detection data of the heat tracing cable, a continuous reconstruction is performed to obtain the heat tracing operation data sequence. The method includes: If the current time node is the time node corresponding to the overlapping observation endpoint, the circuit corresponding to the main heat tracing cable will be disconnected so that the circuit corresponding to the associated backup heat tracing cable can operate independently and obtain independent detection data of the associated backup heat tracing cable. The real-time detection data of the main heat tracing cable, the mapping detection data of the heat tracing cable, and the independent detection data of the associated backup heat tracing cable are continuously reconstructed according to the time nodes to obtain the heat tracing operation data sequence.

9. The intelligent control method for tracing cables based on multi-parameter compensation and correction as described in claim 5, characterized in that, The multi-parameter detection sample data includes at least the gas chamber surface temperature, SF6 gas pressure, main heat tracing cable operating current, and associated backup heat tracing cable operating current.

10. A smart control system for heat tracing cables based on multi-parameter compensation and correction, characterized in that, The system is used to implement the intelligent control method for tracing cables based on multi-parameter compensation correction as described in any one of claims 1 to 9, wherein the system comprises: The configuration module is used to establish a main and backup heat tracing association model based on the configuration relationship of the main and backup heat tracing devices corresponding to the high-voltage circuit breaker gas chamber. The main and backup heat tracing devices include at least one main heat tracing device and one backup heat tracing device. The acquisition module is used to acquire real-time detection data of the main and backup tropics. By analyzing the real-time detection data of the main and backup tropics, an overlapping observation window of the main and backup tropics is generated when the main and backup tropics trigger a fault. The migration module is used to simultaneously start the associated backup heat tracing cable to perform complementary power migration control in the overlapping observation window of the main and backup heat tracing cables according to the main and backup heat tracing cable association model, and obtain synchronous detection data of the main and backup heat tracing cables. The compensation module is used to learn the multi-parameter compensation mapping relationship of the synchronous detection data of the main and backup tracing cables to obtain the tracing cable mapping detection data. The reconstruction module is used to continuously reconstruct the heat tracing operation data sequence based on the real-time detection data of the main heat tracing cable and the mapping detection data of the heat tracing cable, and then send it to the high-voltage circuit breaker fault detection processor.