Intelligent online monitoring equipment applied to cable joint

Abstract

The invention discloses intelligent online monitoring equipment applied to a cable joint, which belongs to the technical field of cable detection and comprises a transmission optical fiber, a fiber bragg grating sensor, an optical fiber demodulator, a data processing unit and an intelligent early warning unit. The transmission optical fiber is used for monitoring signal transmission; the sensor is installed at a key temperature measurement part of a cable joint and is connected in series through optical fiber fusion to form a sensing network. The optical fiber demodulator realizes multi-joint positioning monitoring and wavelength offset acquisition through a wavelength division multiplexing or time division multiplexing technology; the data processing unit converts the wavelength offset into temperature variation; and the intelligent early warning unit adopts a convolutional neural network, a bidirectional long-short-term memory network and a self-attention mechanism mixed model to realize fault grading early warning and temperature rise trend advanced pre-judgment. The cable joint fault early warning system does not need field power supply, is high in anti-electromagnetic interference capability, can realize all-weather online monitoring and accurate positioning, and effectively improves the accuracy and timeliness of cable joint fault early warning.

Description

Technical Field

[0001] This invention relates to the technical field of cable testing, and more specifically, to an intelligent online monitoring device for cable joints. Background Technology

[0002] As the core carrier of power transmission and energy distribution, cable lines are limited by cable manufacturing processes and actual engineering requirements, making continuous long-distance cable laying impossible without splicing. Cable joints have become an indispensable and critical component of cable lines. However, cable joints are necessary to connect the conductors of two cable sections, and the contact surface inevitably retains tiny physical gaps, oxide films, oil stains, or dust. These impurities directly increase the contact resistance of the joint, causing localized high-temperature points to form at the joint. When the joint temperature exceeds the temperature resistance threshold of the insulation layer, the insulation layer will melt or even burn, ultimately leading to short circuits, fires, and large-scale power outages.

[0003] Faced with the aforementioned severe safety risks, existing monitoring and protection methods are still insufficient, mainly in the following aspects: Relying on periodic manual inspections or handheld temperature measurement methods cannot achieve 24 / 7 online monitoring, resulting in significant time and spatial blind spots. This method struggles to detect early, slight heating phenomena and abnormal temperature rise rates at joints, often only discovering the problem when it has progressed to the middle or late stages and the temperature has significantly exceeded the limit, leading to severely delayed warnings and the loss of the optimal window for intervention. Existing technologies primarily focus on regional temperature monitoring or open flame identification after a fire, making it difficult to accurately pinpoint the specific joint causing the anomaly in a densely laid cable network. Once an alarm is triggered, maintenance personnel still need to spend considerable time checking each joint individually, failing to quickly determine the location after a fault occurs, greatly delaying repair time and reducing efficiency. This reduces operational efficiency; most monitoring systems can only provide basic temperature data and lack in-depth intelligent analysis of temperature change trends, rates, and correlation characteristics. They cannot effectively identify early hidden dangers caused by slight increases in contact resistance or slow seal failure, and they cannot achieve early warning based on predictive models. This keeps operation and maintenance work in a passive "post-event response" mode for a long time. For cable joints located in scattered, remote, and inaccessible locations such as high altitudes, high voltage, underwater, or narrow gaps, existing solutions usually install monitoring devices independently at each joint. These devices are often large and require additional on-site power supply (such as batteries or taps). Therefore, batteries must be replaced or charged manually every time. Maintenance is difficult in harsh environments and the life cycle cost is high. Their large size also limits their application in compact spaces. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent online monitoring device for cable joints, thereby solving the above-mentioned technical problems.

[0005] To achieve the above objectives, the present invention provides the following technical solution: An intelligent online monitoring device for cable joints includes a transmission optical fiber, a fiber Bragg grating sensor (FBG), an optical fiber demodulator, a data processing unit, and an intelligent early warning unit. The transmission optical fiber can be directly embedded inside the outer sheath and under the armor layer of the cable during the cable production process, realizing the structural integration of the sensing unit and the cable. For on-site construction with already laid cables, the transmission optical fiber can be tightly fixed to the outer wall of the cable using high-temperature resistant straps, clamps, or special adhesives, achieving physical contact coupling with the cable. The FBG sensor, as a sensing unit, is directly installed at the key temperature measurement point of the cable joint to sense real-time temperature changes on the joint surface. Each FBG sensor is connected to the transmission optical cable via a fiber optic fusion splicer for low-loss, high-reliability fusion splicing, thereby connecting multiple sensors in series to the same optical fiber link to form a quasi-distributed sensing network. The sensing signal is then transmitted long-distance to the fiber optic demodulator at the monitoring center via the transmission optical cable. The FBG is sheathed with a polymer tube, and the optical fiber is in a relaxed, bent state inside the tube, with both ends fixed with adhesive. When the external structure deforms, the force is not transmitted to the optical fiber, but heat can be conducted through the tube wall, ensuring the FBG wavelength shift is controlled. The effects are primarily limited to temperature. The fiber optic demodulator comprises a laser source, a modulation unit, and a signal receiving unit. The laser output from the laser source is modulated and amplified by the modulation unit to ensure that the output wavelength range of the light source covers the initial Bragg wavelengths of all FBG sensors and that the intensity of the output light can support the required transmission distance. The signal receiving unit receives the reflection peaks at all FBGs along the entire fiber in real time. Its spectral detection range must cover the initial Bragg wavelengths of all FBG sensors and their maximum expected offset within the measurement range. The data processing unit processes the reflection peaks in real time, obtaining the wavelength of each reflection peak and the time it takes for the reflection peak to reach the fiber optic demodulator. It then locates the wavelength of each FBG reflection peak and subtracts the Bragg wavelength of each FBG pre-stored in the data processing unit. To obtain wavelength offset Meanwhile, the data processing unit calculates the wavelength shift based on the FBG's sensing principle. Converted to temperature change The intelligent early warning unit issues warnings based on real-time temperature change gradients and absolute temperatures. The absolute temperature is determined by combining real-time calculated wavelength data with ambient temperature and a coupling compensation coefficient. The core function of this coupling compensation coefficient is to correct measurement errors introduced by incomplete mechanical coupling between the FBG sensor and the cable connector, as well as by the thermal insulation of the sealing layer.

[0006] Furthermore, the FBG positioning method can employ wavelength division multiplexing (WDM) or time division multiplexing (TDM). When using WDM, a different Bragg wavelength is written into the FBG at each cable joint. The fiber optic demodulator scans the entire wavelength range and distinguishes different FBGs by identifying the wavelength of the reflection peak. When the demodulator detects a certain wavelength, it corresponds to a specific connector position. When the FBG positioning adopts the time-division multiplexing method, a pulsed laser source is used or a continuous laser is modulated into an optical pulse through a modulation unit. The optical pulse propagates in the optical fiber, and each FBG will reflect a part of the light. By measuring the time difference of the reflected pulse return and combining it with the speed of light, the distance between the FBG and the demodulator can be calculated to realize the positioning of the cable connector.

[0007] Furthermore, the principle of temperature and strain monitoring using FBG is that when laser light passes through FBG, it is reflected at the FBG wavelength. The narrow band of light centered on the reflection peak remains unaffected, allowing other wavelengths of light to continue propagating. From grating period and the effective refractive index of the fiber core The decision is made if the following formula is satisfied:

[0008] When the ambient temperature of the FBG changes or it is subjected to external forces, it will change and This leads to the FBG itself When a shift occurs, the demodulator detects the center wavelength of the reflection peak. movement amount To sense changes acting on the FBG, wavelength shift With changing temperature and strain The relationship satisfies the following formula:

[0009] in For temperature coefficient, The strain coefficient and the strain are both known quantities before monitoring and are related to the physical parameters of the fiber optic material itself. Therefore, the temperature and strain acting on the FBG can be obtained through the above formula. To reduce the influence of the external environment on the temperature demodulation accuracy, a polymer sleeve is fitted over the FBG. The fiber is in a relaxed, bent state inside the sleeve, and both ends are fixed with adhesive. When the external structure deforms, the force will not be transmitted to the fiber. At this time, the wavelength shift is... It is basically determined by temperature:

[0010] Furthermore, the fault diagnosis model built into the intelligent early warning unit receives inputs including real-time temperature changes obtained from the data processing unit. and absolute temperature The model uses a CNN convolutional neural network to extract local mutation patterns and obtain the temperature rise rate, accurately capturing the short-term steep rise characteristics of abnormal heating in cable joints. It then combines a Bi-LSTM bidirectional long short-term memory network to learn the long-term temporal dependencies of temperature data and explore the long-term evolution law of joint operation status. At the same time, a self-attention layer is added to adaptively allocate weights, allowing the model to focus on key time nodes with drastic temperature rise and high failure risk, weakening the interference of normal temperature rise caused by environmental and compliance loads. The model is set with dual output heads. One is a classification head, which is used to output the failure probability of three levels: normal, fault warning, and emergency alarm, to achieve accurate fault classification and early warning. The other is a regression head, which is used to predict the short-term temperature value in the future, to complete the advanced prediction of temperature rise trend and avoid the risk of thermal collapse and fire in advance. Furthermore, the fault discrimination model built into the intelligent early warning unit uses a large amount of historical temperature time-series datasets of tagged cable joints during the training phase. This dataset includes data on various operating conditions such as normal operation, overload temperature rise, poor contact temperature rise, sealing failure and moisture-induced temperature rise, and structural loosening and deformation temperature rise. After data augmentation and time-series standardization preprocessing, the data is input into a hybrid network for iterative training to optimize model weights and improve the accuracy of fault classification and temperature rise trend prediction. During the inference phase, the model receives all FBG temperature data transmitted by the data processing unit in batches in real time. When the temperature rise rate is consistently higher than the historical baseline or the environmental temperature rise model, the temperature curve shows a stable monotonous upward trend, and the model predicts the specific time when the short-term temperature value will approach the safety limit of the sealing material, the model issues a fault warning and outputs the predicted temperature curve for the next 10-30 minutes. Combined with the fault cable joint location information output by the data processing unit, the model provides maintenance personnel with a basis for handling the situation. When the temperature rises rapidly in a very short time and the estimated temperature reaches or exceeds the ignition point, an emergency alarm is output, indicating that there may be a high-temperature fire inside the cable joint or a fire in the external environment. The power supply needs to be cut off and the location and repairs need to be carried out.

[0011] In summary, compared with the prior art, the beneficial effects of the present invention are: This invention provides an intelligent online monitoring device for cable joints. Its monitoring is highly targeted; by installing FBG sensors at key locations on each cable joint and connecting multiple FBGs in series to the same transmission optical fiber, it achieves centralized, real-time, and high-precision location monitoring of multiple cable joints. Monitoring data is uploaded in real-time via the transmission optical fiber to a fiber optic demodulator for unified processing, eliminating the need for separate power supplies for each monitoring point on-site and avoiding the cumbersome maintenance of traditional distributed monitoring equipment that requires regular battery replacements. This invention uses optical signal sensing, with no electrical signal transmission throughout, resulting in outstanding anti-interference capabilities. It avoids strong electromagnetic interference from high-voltage cable joints, eliminating the risk of leakage and short circuits. The monitoring data is accurate and stable, adaptable to various high-voltage, strong electromagnetic, and humid environments. The thin and flexible optical fiber facilitates installation in confined spaces. Sensors are deployed only at preset points on the joint, eliminating redundant sensing areas, which not only accurately focuses on the core state of the joint but also reduces system complexity and data processing burden. The intelligent early warning unit employs a hybrid model combining CNN, Bi-LSTM, and Self-Attention, effectively integrating local mutation feature extraction, long-term time-series dependency learning, and key information focusing capabilities. Compared to traditional fixed-threshold alarm methods, this significantly reduces false alarm and false negative rates. This intelligent early warning unit can intelligently predict short-term temperature rise trends, providing early warnings of potential thermal collapse and fire risks. It offers maintenance personnel effective decision-making support and a window for handling issues, achieving the intelligent maintenance goal of early fault detection and early response. Attached Figure Description

[0012] Figure 1 This is a schematic diagram of an intelligent online monitoring device for cable joints. Figure 2 Installation diagram of fiber Bragg grating sensor Figure 1 ; Figure 3 Installation diagram of fiber Bragg grating sensor Figure 2 ; Wherein: 1-Fiber Bragg grating sensor, 2-Cable connector, 3-Transmission fiber. Detailed Implementation

[0013] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments given herein are for illustration and explanation only and are not intended to limit the present invention. Specific Implementation Example 1

[0015] like Figure 1 and 2 As shown in the first embodiment of the present invention, an intelligent online monitoring device for cable joints is provided, including a transmission optical fiber, an FBG sensor, an optical fiber demodulator, a data processing unit, and an intelligent early warning unit.

[0016] The transmission optical fiber is directly embedded within the outer sheath and under the armor layer of the cable during the cable manufacturing process, achieving structural integration of the sensing unit and the cable. At the cable joint, two cables are connected, and an FBG sensor is fused between the two transmission optical fibers, maintaining a tight connection between the FBG sensor and the cable joint. This allows for real-time sensing of temperature changes on the joint surface. Each FBG sensor is connected to the transmission optical cable via a fiber optic fusion splicer for low-loss, high-reliability bonding, thus connecting multiple sensors in series in the same optical fiber link to form a quasi-distributed sensing network. The sensing signals are then transmitted over long distances via the transmission optical cable to the fiber optic demodulator at the monitoring center. The FBG sensors are pre-customized, and each FBG sensor has a different initial reflection wavelength. The fiber optic demodulator includes a laser source, a modulation unit, and a signal receiving unit. The laser source and the modulation unit are connected by an optical connection. The output laser is modulated and amplified by the modulation unit to ensure that the output wavelength range of the light source covers the initial Bragg wavelength of all FBG sensors and its maximum expected offset within the measurement range, and to ensure that the intensity of the output light can support the required transmission distance. The modulation unit is connected to the transmission optical fiber fixed on the cable through a circulator and an optical fiber interface. The FBG reflection peak signal emitted by the laser after passing through the FBG sensor passes through the circulator and is received by the receiving unit. The monitoring signal received by the demodulator is connected to the data processing unit through a circuit. The data processing unit transmits the data to the fault discrimination model of the intelligent early warning unit.

[0017] Specifically, the fault diagnosis model built into the intelligent early warning unit receives inputs including real-time temperature changes obtained from the data processing unit. and absolute temperature The model uses CNN to extract local mutation patterns from the fused temperature time-series data to obtain the temperature rise rate, accurately capturing the short-term steep rise characteristics of abnormal heating in cable joints. It then combines a Bi-LSTM bidirectional long short-term memory network to learn the long-term temporal dependencies of temperature data and explore the long-term evolution pattern of the joint's operating state. At the same time, a Self-Attention layer is added to adaptively allocate weights, allowing the model to focus on key time nodes with drastic temperature rise and high failure risk, weakening the interference of normal temperature rise caused by environmental and compliance loads. The model is set with dual output heads. One is a classification head, which is used to output the failure probability of three levels: normal, fault warning, and emergency alarm, to achieve accurate fault classification and early warning. The other is a regression head, which is used to predict the short-term temperature value in the future, to complete the advanced prediction of temperature rise trend and avoid the risk of thermal collapse and fire in advance.

[0018] Furthermore, the fault discrimination model built into the intelligent early warning unit uses a large amount of historical temperature time-series datasets of tagged cable joints during the training phase. This dataset includes data on various operating conditions such as normal operation, overload temperature rise, poor contact temperature rise, sealing failure and moisture-induced temperature rise, and structural loosening and deformation temperature rise. After data augmentation and time-series standardization preprocessing, the data is input into a hybrid network for iterative training to optimize model weights and improve the accuracy of fault classification and temperature rise trend prediction. During the inference phase, the model receives all FBG temperature data transmitted by the data processing unit in batches in real time. When the temperature rise rate is consistently higher than the historical baseline or the environmental temperature rise model, the temperature curve shows a stable monotonous upward trend, and the model predicts the specific time when the short-term temperature value will approach the safety limit of the sealing material, the model issues a fault warning and outputs the predicted temperature curve for the next 10-30 minutes. Combined with the fault cable joint location information output by the data processing unit, the model provides maintenance personnel with a basis for handling the situation. When the temperature rises rapidly in a very short time and the estimated temperature reaches or exceeds the ignition point, an emergency alarm is output, indicating that there may be a high-temperature fire inside the cable joint or a fire in the external environment. The power supply needs to be cut off and the location and repairs need to be carried out.

[0019] During cable joint monitoring, six FBG sensors are connected to a single optical fiber. The initial Bragg wavelength of the FBG sensor corresponding to cable joint J1 is 1535nm, the initial Bragg wavelength of the FBG sensor at cable joint J2 is 1540nm, and the initial Bragg wavelengths of the FBG sensors at cable joints J3, J4, J5, and J6 are 1545nm, 1550nm, 1555nm, and 1560nm, respectively. The laser source output of the fiber optic demodulator outputs a scanning laser covering the 1530nm-1565nm range. After being amplified by the modulation unit, the laser is injected into the transmission fiber. Each FBG reflects a specific wavelength of light signal, and the signal receiving unit receives the reflection peaks at all FBGs along the entire fiber online. The data processing unit processes the reflection peaks in real time to obtain the wavelength of each reflection peak. Based on the cable joint corresponding to the reflection peak wavelength of each FBG, the location monitoring of the cable joint is achieved, and the pre-stored Bragg wavelength of each FBG in the data processing unit is subtracted. To obtain wavelength offset Meanwhile, the data processing unit calculates the wavelength shift based on the FBG's sensing principle. Through formula Converted to temperature change Combining the ambient temperature of 20℃ and a coupling compensation coefficient of 0.8, the absolute temperature T is determined. The core function of this coupling compensation coefficient is to correct the measurement errors introduced by incomplete mechanical coupling between the FBG sensor and the cable joint and the thermal insulation of the sealing layer. The CNN-Bi-LSTM-Self-Attention model of the intelligent early warning unit processes the temperature time-series data in real time. When the temperature of a cable joint rises from 45℃ to 65℃ at a rate of 5℃ / min, the model classification head outputs a fault warning probability of 85%, and the regression head predicts that the temperature will reach 85℃ in 30 minutes (close to the safe upper limit of 90℃ for the sealing material). The intelligent early warning unit pushes warning information to maintenance personnel, including the joint number, location information, and predicted temperature curve. Maintenance personnel rush to the site based on the location information, check the condition of the cable joint, and replace it in time to prevent cable fire.

[0020] As a preferred technical solution in this embodiment, the FBG is encased in a polymer sleeve, and the optical fiber is in a relaxed, bent state inside the sleeve, with both ends fixed with adhesive. When the external structure deforms, the force will not be transmitted to the optical fiber, but heat can be conducted through the sleeve wall, ensuring the FBG wavelength shift. It is only affected by temperature.

[0021] Example 2

[0022] like Figure 1 and Figure 3 As shown, in on-site construction where cables have already been laid, the optical fiber can be tightly fixed to the outer wall of the cable using high-temperature resistant straps, clamps, or special adhesives to achieve physical contact coupling with the cable. Time-division multiplexing is used for locating cable joints. The fiber optic demodulator is equipped with a pulsed laser source (10ns pulse width, 1kHz repetition frequency), or the continuous laser signal is modulated into a pulse signal through a modulation unit. The time difference of the reflected pulse return is measured. Calculate the distance to the FBG sensor:

[0023] Where c is the speed of light in the optical fiber. The refractive index of the fiber optic cable is used. The distance to each FBG sensor is obtained to locate the cable joint. When the temperature of a cable joint reaches 65℃, the model classification head outputs a fault warning probability of 85%, and the regression head predicts that the temperature will reach 85℃ in 30 minutes (approaching the safety limit of 90℃ for the sealing material). The intelligent early warning unit pushes a warning message to maintenance personnel, including the joint number, location information, and predicted temperature curve. Maintenance personnel then rush to the site based on the location information, inspect the cable joint, and replace it promptly to prevent cable fires.

[0024] It should be understood that the above embodiments are one or more embodiments of the present invention. There are many other embodiments and variations based on the present invention. Any variations and modifications made by those skilled in the art without making pioneering innovations are within the protection scope of the present invention.

Claims

1. An intelligent online monitoring device for cable joints, characterized in that, It includes transmission optical fiber, fiber Bragg grating sensor, fiber demodulator, data processing unit and intelligent early warning unit; The transmission optical fiber is used to transmit optical signals. During cable production, the transmission optical fiber is embedded inside the cable's outer sheath, under the armor layer, or fixed to the outer wall of the laid cable. The fiber Bragg grating sensor, as a sensing unit, is installed at a key temperature measurement point on the cable joint to sense real-time changes in the joint's surface temperature. Each fiber Bragg grating sensor is low-loss fused to the transmission optical fiber using a fiber optic fusion splicer. Multiple fiber Bragg grating sensors are connected in series to the same optical fiber link to form a quasi-distributed sensing network. The fiber demodulator includes a laser source, a modulation unit, and a signal receiving unit. The laser source transmits optical signals, the modulation unit modulates and amplifies the output laser wavelength range, and the signal receiving unit receives the reflection peak signals from the fiber Bragg grating sensors. The data processing unit processes the reflection peak signals in real-time to obtain the wavelength and arrival time of each reflection peak and calculates the wavelength offset. The temperature change ΔT is converted into an amount of temperature change. The absolute temperature T is determined by combining the ambient temperature and the coupling compensation coefficient. The fiber Bragg grating sensor is located by wavelength division multiplexing or time division multiplexing. The intelligent early warning unit has a built-in fault discrimination model. It inputs fused temperature time series data and outputs fault classification early warning results and temperature prediction values ​​for the next 10-30 minutes.

2. The intelligent online monitoring device for cable joints according to claim 1, characterized in that, When the fiber Bragg grating sensor is positioned using wavelength division multiplexing, a different Bragg wavelength is written to the fiber Bragg grating sensor at each cable joint. The fiber optic demodulator scans the entire wavelength range and distinguishes different fiber Bragg grating sensors by identifying the wavelength of the reflection peak, thus achieving connector positioning.

3. The intelligent online monitoring device for cable joints according to claim 1, characterized in that, When the fiber Bragg grating sensor is positioned using a time-division multiplexing method, a pulsed laser source is used or a continuous laser is modulated into an optical pulse through a modulation unit. The distance between the fiber Bragg grating sensor and the demodulator is calculated by measuring the time difference of the reflected pulse return and combining it with the speed of light, thereby realizing the positioning of the cable joint.

4. The intelligent online monitoring device for cable joints according to claim 1, characterized in that, The fiber Bragg grating sensor is sheathed with a polymer tube, and the optical fiber is relaxed and bent inside the tube, with both ends fixed with adhesive to achieve decoupling of temperature and strain; the wavelength offset in the data processing unit... The conversion relationship with the temperature change ΔT is as follows: ,in This is the temperature coefficient.

5. The intelligent online monitoring device for cable joints according to claim 1, characterized in that, The fault discrimination model of the intelligent early warning unit includes a convolutional neural network layer, a bidirectional long short-term memory network layer, a self-attention mechanism layer, and a dual output head. The dual output head includes a classification head and a regression head. The classification head outputs the fault probability of three levels: normal, fault warning, and emergency alarm. The regression head predicts the temperature value for the next 10-30 minutes.

6. The intelligent online monitoring device for cable joints according to claim 5, characterized in that, The fault discrimination model is trained using historical temperature time series datasets that include multiple operating conditions such as normal operation, overload temperature rise, poor contact temperature rise, sealing failure and moisture-induced temperature rise, and structural loosening and deformation temperature rise during the training phase. The datasets are then trained after data augmentation and time series standardization preprocessing.

7. The intelligent online monitoring device for cable joints according to claim 1, characterized in that, The modulation unit of the fiber optic demodulator modulates the wavelength range of the laser to cover the initial Bragg wavelength of all fiber Bragg grating sensors, and the spectral detection range of the signal receiving unit must cover the initial Bragg wavelength of all FBG sensors and their maximum expected offset within the measurement range.

Images