Cable external damage detection method and system based on optical fiber expansion multi-sensor fusion

Abstract

The invention relates to the technical field of high-voltage cable protection, and provides a cable external damage detection method and system based on optical fiber expansion multi-sensor fusion, and the method comprises the steps: obtaining a real-time vibration characteristic and a real-time partial discharge characteristic; then based on the trained external damage prediction model; inputting the real-time partial discharge characteristics at the moment T after the moment corresponding to the real-time vibration characteristics and the preset time T into the external damage prediction model, and calculating to obtain an external damage occurrence probability; and inputting real-time vibration characteristics at (-t, t) moments within a time period of preset time t before and after the moment corresponding to the real-time partial discharge characteristics into the external damage prediction model, and calculating to obtain an external damage occurrence probability. Through training of the neural network prediction model, relevance among vibration features, partial discharge features and external damage results is learned, and the problem that prediction accuracy of a single parameter is not high is solved; meanwhile, the vibration characteristics and the partial discharge characteristics are divided into different situations, signals at corresponding moments after time shifting are obtained, and the prediction accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of high-voltage cable protection, and in particular to a cable external damage detection method and system based on optical fiber expansion and multi-sensor fusion. Background Art

[0002] In order to avoid the potential safety hazards caused by direct exposure of cables to the outside world and to ensure the aesthetics of the city, many cables are laid underground. Power pipeline corridors, also known as cable tunnels and power tunnels, are important municipal infrastructure in cities. Cable tunnels are deep underground and extend to every corner of the city.

[0003] However, with the gradual development and advancement of urban infrastructure, many infrastructure construction projects require excavation of foundations, which poses a serious safety crisis for underground cables. Once the cables are accidentally dug up, it will cause regional power outages and circuit breakers, causing serious impacts on residents' lives and corporate production.

[0004] Traditionally, the protection of underground cables is just to arrange special personnel to patrol. This method has low automation and intelligence, and the actual protection effect is not ideal. At present, the monitoring of underground cables mainly adopts distributed optical fiber sensing technology. However, the signal recognition of the current distributed optical fiber sensor is easily affected by environmental factors, such as weather, animals, industrial activities, etc., resulting in a high false alarm rate of optical fiber sensors. And because aliasing signals are common in actual situations, it is not conducive to the detection of external damage events by distributed optical fiber sensors.

[0005] For cable damage detection, there are two main forms of distributed optical fiber sensing technology:

[0006] The first is to perform partial discharge detection on the cable. The causes of power cable failure are complex and varied. Gaps and impurities mixed in during the cable manufacturing process, scratches and pollution caused during the laying and installation process, and electrical trees generated during the aging process will cause partial discharges to varying degrees. Partial discharge is also an important factor causing cable insulation degradation. Therefore, partial discharge detection covers the vast majority of causes of cable failures. The level of cable partial discharge (generally referred to as partial discharge) generally reflects the safety of the cable. However, generally speaking, lighter levels of partial discharge do not necessarily mean that external damage has occurred. Higher levels of partial discharge usually mean that external damage has already occurred. Therefore, although partial discharge detection can detect the occurrence of external damage, it is not accurate and cannot provide early warning;

[0007] The other is the detection of cable vibration. Generally speaking, cable damage caused by external force, impact by external objects, earthquakes, etc. is usually accompanied by vibration. Therefore, the level of vibration is correlated with cable damage. However, there is no direct correlation between cable vibration and damage. Slight vibration may cause damage when the cable is severely aged, and stronger vibration does not necessarily mean damage when the cable installation structure is sufficiently optimized. Therefore, the accuracy of predicting cable damage through vibration detection needs to be improved. Summary of the invention

[0008] The purpose of the present invention is to provide a cable external damage detection method and system based on optical fiber expansion multi-sensor fusion to solve the above-mentioned technical problems.

[0009] The present invention provides a cable external damage detection method based on optical fiber expansion multi-sensor fusion, comprising the steps of:

[0010] S1, training a preset neural network prediction model according to an external failure training sample to obtain an external failure prediction model; wherein each data sample in the external failure training sample includes input data and an external failure result, the input data includes vibration characteristics and partial discharge characteristics of the cable, the external failure result includes whether an external failure occurs and whether an external failure does not occur, and the external failure training sample includes at least one of a simulated data sample and a real data sample; the vibration characteristics include vibration amplitude and continuous vibration duration, and the partial discharge characteristics include current characteristics and / or voltage characteristics;

[0011] S2, obtaining a vibration signal detected in real time by a fiber optic interferometer sensor arranged along the cable, and obtaining a partial discharge signal detected in real time by a partial discharge sensor arranged along the cable;

[0012] S3, performing normalization processing and feature extraction on the vibration signal to obtain real-time vibration features; performing normalization processing and feature extraction on the partial discharge signal to obtain real-time partial discharge features;

[0013] S41, when the real-time vibration feature is greater than a first preset vibration threshold and less than a second preset vibration threshold, obtaining a real-time partial discharge feature at a time T after a preset time T after a time corresponding to the real-time vibration feature; wherein the second preset vibration threshold is greater than the first preset vibration threshold;

[0014] S51, inputting the real-time vibration characteristics and the real-time partial discharge characteristics at time T obtained in S41 into the external failure prediction model to calculate the probability of external failure;

[0015] S42, when the real-time partial discharge characteristic is less than a first preset partial discharge threshold, obtaining a real-time vibration characteristic at a time (-t, t) within a preset time period t before and after the time corresponding to the real-time partial discharge characteristic;

[0016] S52, inputting the real-time partial discharge characteristics and the real-time vibration characteristics at the time (-t, t) obtained in S42 into the external failure prediction model to calculate the probability of external failure.

[0017] Furthermore, the method further comprises the steps of:

[0018] S43, when the real-time vibration characteristic is greater than the second preset vibration threshold and less than the third preset vibration threshold, determining whether the real-time partial discharge characteristic at the current moment is greater than the first partial discharge safety threshold; wherein the third preset vibration threshold is greater than the second preset vibration threshold;

[0019] If so, an external break alarm signal is directly issued;

[0020] If not, then obtain the real-time partial discharge feature at the time T after the time corresponding to the real-time vibration feature plus the preset time T, and go to step S41;

[0021] Wherein, the second preset vibration threshold is greater than the first preset vibration threshold.

[0022] Furthermore, when the real-time vibration characteristic is greater than a third preset vibration threshold, an external damage alarm signal is directly issued.

[0023] Furthermore, the method further comprises the steps of:

[0024] S43, when the real-time partial discharge characteristic is lower than the second preset partial discharge threshold and greater than the first preset partial discharge threshold, determining whether the real-time vibration characteristic at the current moment is greater than the first vibration safety threshold; the first vibration safety threshold is less than the third preset vibration threshold;

[0025] If so, a new external break alarm signal is directly issued;

[0026] If not, then obtain the real-time vibration characteristics at the time (-t, t) within a preset time period t before and after the time corresponding to the real-time partial discharge characteristics, and enter step S42.

[0027] Furthermore, the method further comprises the steps of:

[0028] The real-time vibration characteristics and the real-time partial discharge characteristics are input into the external failure prediction model to calculate the probability of external failure occurrence.

[0029] Furthermore, the time T is 1 to 60 seconds.

[0030] Furthermore, the t time period is (-1 to -60 seconds, 1 to 60 seconds).

[0031] The present invention also provides a cable external damage detection system based on optical fiber extension multi-sensor fusion, including an optical fiber interference sensor arranged with the cable for real-time detection of the cable's vibration signal, a partial discharge sensor arranged with the cable for real-time detection of the cable's partial discharge signal, and a monitoring background server; the monitoring background server includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the steps of the cable external damage detection method based on optical fiber extension multi-sensor fusion as described in any one of the above items are implemented.

[0032] The present invention provides a cable external damage detection method and system based on optical fiber expansion multi-sensor fusion, which obtains the vibration signal detected in real time by the optical fiber interference sensor arranged along the cable, and obtains the partial discharge signal detected in real time by the partial discharge sensor arranged along the cable; performs normalization processing and feature extraction on the vibration signal to obtain the real-time vibration feature; performs normalization processing and feature extraction on the partial discharge signal to obtain the real-time partial discharge feature; then based on the trained external damage prediction model; when the real-time vibration feature is greater than a first preset vibration threshold, obtain the real-time partial discharge feature at the moment T after the moment corresponding to the real-time vibration feature plus a preset time T, input the real-time vibration feature and the real-time partial discharge feature at the moment T into the external damage prediction model, and calculate the probability of external damage; when the real-time partial discharge feature is less than or equal to the first preset partial discharge threshold, obtain the real-time vibration feature at the moment (-t, t) within a preset time period t before and after the moment corresponding to the real-time partial discharge feature, input the real-time partial discharge feature and the real-time vibration feature at the moment (-t, t) into the external damage prediction model, and calculate the probability of external damage. Firstly, the vibration characteristics and partial discharge characteristics are combined, and the correlation between the vibration characteristics, partial discharge characteristics and external failure results is learned through the training of the neural network prediction model, thereby improving the problem of low prediction accuracy of a single parameter. At the same time, by dividing the vibration characteristics and partial discharge characteristics into different situations, the signals at the corresponding moments after time shift are obtained to adapt to the causal relationship between vibration and partial discharge, thereby improving the accuracy of the prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a flow chart of a cable damage detection method based on optical fiber expansion and multi-sensor fusion in one embodiment of the present invention.

[0034] Figure 2 It is a schematic diagram of the hardware structure of a cable external break detection system based on optical fiber expansion multi-sensor fusion in one embodiment of the present invention.

[0035] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings in conjunction with the embodiments. DETAILED DESCRIPTION

[0036] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0037] In the subsequent description, the suffixes such as "module", "component" or "unit" used to represent elements are only used to facilitate the description of the present invention, and have no specific meanings. Therefore, "module", "component" or "unit" can be used in a mixed manner.

[0038] The present invention will now be described in further detail with reference to the accompanying drawings.

[0039] Please combine Figure 1 and Figure 2 , Figure 1 FIG. 1 is a flow chart of a cable external damage detection method based on optical fiber expansion multi-sensor fusion in one embodiment of the present invention. Figure 2 This is a schematic diagram of the framework structure of the cable external damage detection system based on optical fiber expansion and multi-sensor fusion.

[0040] The cable external damage detection method based on optical fiber expansion multi-sensor fusion includes the following steps:

[0041] S1. Training a preset neural network prediction model according to an external failure training sample to obtain an external failure prediction model; wherein each data sample in the external failure training sample includes input data and an external failure result, the input data includes vibration characteristics and partial discharge characteristics of the cable, the external failure result includes whether external failure occurred and whether external failure occurred, and the external failure training sample includes at least one of a simulated data sample and a real data sample; the vibration characteristics include vibration amplitude and continuous vibration duration, and the partial discharge characteristics include current characteristics and / or voltage characteristics.

[0042] Specifically, in order to improve the diversity of training samples, the data volume can be increased by simulating the safe operation data of cables; at the same time, more real data can be obtained to expand the sample credibility and improve the accuracy of model training.

[0043] S2, obtaining a vibration signal detected in real time by a fiber optic interferometer sensor arranged along the cable, and obtaining a partial discharge signal detected in real time by a partial discharge sensor arranged along the cable.

[0044] S3, performing normalization processing and feature extraction on the vibration signal to obtain real-time vibration features; performing normalization processing and feature extraction on the partial discharge signal to obtain real-time partial discharge features.

[0045] S41, when the real-time vibration feature is greater than the first preset vibration threshold and less than the second preset vibration threshold, obtain the real-time partial discharge feature at the moment T after the moment corresponding to the real-time vibration feature plus the preset time T; wherein the second preset vibration threshold is greater than the first preset vibration threshold; generally speaking, the partial discharge phenomenon caused by vibration is often not generated immediately, and the detection of the partial discharge sensor also has a delay, so the real-time vibration feature is time-shifted to increase the accuracy of the prediction.

[0046] S51, inputting the real-time vibration characteristics and the real-time partial discharge characteristics at time T obtained in S41 into the external failure prediction model to calculate the probability of external failure.

[0047] S42, when the real-time partial discharge feature is less than or lower than a first preset partial discharge threshold, obtain the real-time vibration feature at the moment (-t, t) within a preset time period t before and after the moment corresponding to the real-time partial discharge feature; the real-time partial discharge feature also has a time difference with the real-time vibration feature. By widening the collection time period of the corresponding real-time vibration feature when the real-time partial discharge feature reaches a dangerous value, the accuracy of the prediction can be effectively increased.

[0048] S52, inputting the real-time partial discharge characteristics and the real-time vibration characteristics at the time (-t, t) obtained in S42 into the external failure prediction model to calculate the probability of external failure.

[0049] Furthermore, in a preferred embodiment, the method further comprises the steps of:

[0050] S43, when the real-time vibration characteristic is greater than the second preset vibration threshold and less than the third preset vibration threshold, determining whether the real-time partial discharge characteristic at the current moment is greater than the first partial discharge safety threshold; wherein the third preset vibration threshold is greater than the second preset vibration threshold;

[0051] If so, an external break alarm signal is directly issued;

[0052] If not, then obtain the real-time partial discharge feature at the time T after the time corresponding to the real-time vibration feature plus the preset time T, and go to step S41;

[0053] Wherein, the second preset vibration threshold is greater than the first preset vibration threshold.

[0054] In this embodiment, if the real-time characteristic is large, the risk of external rupture may increase; at this time, if the real-time partial discharge characteristic has reached a certain level, an external rupture alarm signal is directly issued as an early warning; if the real-time partial discharge characteristic has not reached the critical value (the first partial discharge safety threshold), further prediction and observation can be performed.

[0055] Further, when the real-time vibration characteristic is greater than the third preset vibration threshold, an external damage alarm signal is directly issued. In this embodiment, by setting the third preset vibration threshold (i.e., the maximum safety threshold), a direct warning is issued for large vibrations.

[0056] Furthermore, in a preferred embodiment, the method further comprises the steps of:

[0057] S43, when the real-time partial discharge characteristic is lower than the second preset partial discharge threshold and greater than the first preset partial discharge threshold, determining whether the real-time vibration characteristic at the current moment is greater than the first vibration safety threshold; the first vibration safety threshold is less than the third preset vibration threshold;

[0058] If so, a new external break alarm signal is directly issued;

[0059] If not, then obtain the real-time vibration characteristics at the time (-t, t) within a preset time period t before and after the time corresponding to the real-time partial discharge characteristics, and enter step S42.

[0060] In this embodiment, if the real-time partial discharge characteristic is large, the risk of external damage may increase as long as vibration occurs. At this time, if the real-time vibration characteristic has reached a certain level, an external damage alarm signal is directly issued as an early warning. If the real-time vibration characteristic has not reached the critical value (the first vibration safety threshold), further prediction and observation can be carried out.

[0061] Furthermore, the method further comprises the steps of:

[0062] When the real-time partial discharge characteristic is greater than a third preset partial discharge threshold, an external break alarm signal is directly issued.

[0063] In this embodiment, a third preset partial discharge threshold (ie, the maximum safety threshold) is set to directly warn of large vibrations.

[0064] Furthermore, the method further comprises the steps of:

[0065] The real-time vibration characteristics and the real-time partial discharge characteristics are input into the external failure prediction model to calculate the probability of external failure. It is understandable that prediction can also be performed using the real-time vibration characteristics and the real-time partial discharge characteristics.

[0066] Furthermore, the time T is 1 to 60 seconds.

[0067] Furthermore, the t time period is (-1 to -60 seconds, 1 to 60 seconds).

[0068] The present invention also provides a cable external break detection system based on optical fiber extension multi-sensor fusion, comprising an optical fiber interference sensor 10 arranged with the cable for real-time detection of the cable's vibration signal, a partial discharge sensor 20 arranged with the cable for real-time detection of the cable's partial discharge signal, and a monitoring background server 60; the monitoring background server comprises a memory 61, a processor 62, and a computer program 63 stored in the memory 61 and executable on the processor 62, and when the processor 62 executes the computer program 63, the steps of the cable external break detection method based on optical fiber extension multi-sensor fusion as described in any one of the above items are implemented.

[0069] The present invention provides a cable external damage detection method and system based on optical fiber expansion multi-sensor fusion, which obtains the vibration signal detected in real time by the optical fiber interference sensor arranged along the cable, and obtains the partial discharge signal detected in real time by the partial discharge sensor arranged along the cable; performs normalization processing and feature extraction on the vibration signal to obtain the real-time vibration feature; performs normalization processing and feature extraction on the partial discharge signal to obtain the real-time partial discharge feature; then based on the trained external damage prediction model; when the real-time vibration feature is greater than a first preset vibration threshold, obtain the real-time partial discharge feature at the moment T after the moment corresponding to the real-time vibration feature plus a preset time T, input the real-time vibration feature and the real-time partial discharge feature at the moment T into the external damage prediction model, and calculate the probability of external damage; when the real-time partial discharge feature is less than or equal to the first preset partial discharge threshold, obtain the real-time vibration feature at the moment (-t, t) within a preset time period t before and after the moment corresponding to the real-time partial discharge feature, input the real-time partial discharge feature and the real-time vibration feature at the moment (-t, t) into the external damage prediction model, and calculate the probability of external damage. Firstly, the vibration characteristics and partial discharge characteristics are combined, and the correlation between the vibration characteristics, partial discharge characteristics and external failure results is learned through the training of the neural network prediction model, thereby improving the problem of low prediction accuracy of a single parameter. At the same time, by dividing the vibration characteristics and partial discharge characteristics into different situations, the signals at the corresponding moments after time shift are obtained to adapt to the causal relationship between vibration and partial discharge, thereby improving the accuracy of the prediction.

[0070] In the description of this specification, the description with reference to the terms "an embodiment", "another embodiment", "other embodiments", or "first embodiment to Xth embodiment" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, method steps or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0071] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0072] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A cable external damage detection method based on optical fiber expansion multi-sensor fusion, characterized in that: Includes steps: S1, training a preset neural network prediction model according to an external failure training sample to obtain an external failure prediction model; wherein each data sample in the external failure training sample includes input data and an external failure result, the input data includes vibration characteristics and partial discharge characteristics of the cable, the external failure result includes whether an external failure occurs and whether an external failure does not occur, and the external failure training sample includes at least one of a simulated data sample and a real data sample; S2, obtaining a vibration signal detected in real time by a fiber optic interferometer sensor arranged along the cable, and obtaining a partial discharge signal detected in real time by a partial discharge sensor arranged along the cable; S3, performing normalization processing and feature extraction on the vibration signal to obtain real-time vibration features; performing normalization processing and feature extraction on the partial discharge signal to obtain real-time partial discharge features; S41, when the real-time vibration feature is greater than a first preset vibration threshold, obtaining a real-time partial discharge feature at a time T after a preset time T after a time corresponding to the real-time vibration feature; S51, inputting the real-time vibration characteristics and the real-time partial discharge characteristics at time T obtained in S41 into the external failure prediction model to calculate the probability of external failure; S42, when the real-time partial discharge characteristic is less than a first preset partial discharge threshold, obtaining a real-time vibration characteristic at a time (-t, t) within a preset time period t before and after the time corresponding to the real-time partial discharge characteristic; S52, inputting the real-time partial discharge characteristics and the real-time vibration characteristics at the time (-t, t) obtained in S42 into the external failure prediction model to calculate the probability of external failure.

2. The cable external damage detection method based on optical fiber expansion multi-sensor fusion according to claim 1 is characterized in that: Also includes the steps: S43, when the real-time vibration characteristic is greater than the second preset vibration threshold and less than the third preset vibration threshold, determining whether the real-time partial discharge characteristic at the current moment is greater than the first partial discharge safety threshold; wherein the third preset vibration threshold is greater than the second preset vibration threshold; If so, an external break alarm signal is directly issued; If not, then obtain the real-time partial discharge feature at the time T after the time corresponding to the real-time vibration feature plus the preset time T, and go to step S41; Wherein, the second preset vibration threshold is greater than the first preset vibration threshold.

3. The cable external damage detection method based on optical fiber expansion multi-sensor fusion according to claim 2 is characterized in that: When the real-time vibration characteristic is greater than a third preset vibration threshold, an external damage alarm signal is directly issued.

4. The cable external damage detection method based on optical fiber expansion multi-sensor fusion according to claim 3 is characterized in that: Also includes the steps: S44, when the real-time partial discharge characteristic is lower than the second preset partial discharge threshold and greater than the first preset partial discharge threshold, determining whether the real-time vibration characteristic at the current moment is greater than the first vibration safety threshold; the first vibration safety threshold is less than the third preset vibration threshold; If so, a new external break alarm signal is directly issued; If not, then obtain the real-time vibration characteristics at the time (-t, t) within a preset time period t before and after the time corresponding to the real-time partial discharge characteristics, and enter step S42.

5. The cable external damage detection method based on optical fiber extension multi-sensor fusion according to claim 4 is characterized in that: Also includes the steps: When the real-time partial discharge characteristic is greater than a third preset partial discharge threshold, an external break alarm signal is directly issued.

6. The cable external damage detection method based on optical fiber extension multi-sensor fusion according to claim 1 is characterized in that: Also includes the steps: The real-time vibration characteristics and the real-time partial discharge characteristics are input into the external failure prediction model to calculate the probability of external failure occurrence.

7. The cable external damage detection method based on optical fiber extension multi-sensor fusion according to any one of claims 1 to 6 is characterized in that: The time T is 1 to 60 seconds.

8. The cable external damage detection method based on optical fiber extension multi-sensor fusion according to any one of claims 1 to 6 is characterized in that: The t time period is (-1 to -60 seconds, 1 to 60 seconds).

9. A cable external damage detection system based on optical fiber expansion multi-sensor fusion, characterized in that: It includes a fiber optic interferometer sensor arranged with the cable for real-time detection of the vibration signal of the cable, a partial discharge sensor arranged with the cable for real-time detection of the partial discharge signal of the cable, and a monitoring background server; the monitoring background server includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, it implements the steps of the cable external damage detection method based on fiber optic extension multi-sensor fusion as described in any one of claims 1-8.

Images