A method and system for detecting communication cable faults

Abstract

The present invention provides a communication cable fault detection method and system, which relate to the field of data processing technology. The method comprises: monitoring voltage signals at various locations of a communication cable through a voltage sensor; judging whether there is a disturbance according to the voltage signal, and if so, proceeding to the next step; calculating a cross-correlation parameter according to the voltage signal; judging whether the cross-correlation parameter is greater than a preset cross-correlation value, and if so, determining that the communication cable is initially diagnosed to have a fault and proceeding to the next step; predicting the voltage signal through a Kalman filter to obtain a voltage estimation signal; calculating a half-cycle standard deviation according to a residual signal; generating a state diagram with the cross-correlation parameter at each moment as the horizontal coordinate and the half-cycle standard deviation at each moment as the vertical coordinate; performing a fine diagnosis through a convolutional neural network according to the state diagram to determine the type of fault; and when a fault is detected in the communication cable, using the position of the voltage sensor as the fault location for positioning and early warning.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a communication cable fault detection method and system. Background Art

[0002] Communication cables connect people and devices by transmitting signals, supporting various communication methods such as telephone communication, Internet, TV signals, wireless communication, etc., enabling people to communicate remotely, share information and obtain entertainment, playing an increasingly important role in people's lives. Cable faults can cause communication interruption or instability, affecting the normal communication services of users. By timely detecting and repairing cable faults, the stability and continuity of the communication system can be ensured.

[0003] During the use of communication cables, initial faults are often difficult to capture by traditional relay protection equipment due to their short duration. However, if the initial faults are not handled in time, the frequency of initial faults is likely to increase over time, eventually destroying the insulation layer of the communication cable and turning into permanent faults, causing losses. Summary of the invention

[0004] In order to solve the technical problem in the prior art that the current relay protection equipment is difficult to capture initial faults, and as time goes by, the frequency of initial faults increases, eventually damaging the insulation layer of the communication cable, and then transforming into permanent faults and causing losses, the present invention provides a communication cable fault detection method and system.

[0005] The technical solution provided by the embodiment of the present invention is as follows:

[0006] First aspect:

[0007] A communication cable fault detection method provided by an embodiment of the present invention includes:

[0008] S1: Monitor the voltage signals at various locations of the communication cable through a voltage sensor;

[0009] S2: judging whether there is disturbance according to the voltage signal, if so, entering S3, otherwise, returning to S1 to continue monitoring;

[0010] S3: Calculating a cross-correlation parameter according to the voltage signal;

[0011] S4: Determine whether the cross-correlation parameter is greater than a preset cross-correlation value. If so, determine that the communication cable is initially diagnosed to have a fault and proceed to S5. Otherwise, return to S1 to continue monitoring.

[0012] S5: predicting the voltage signal through a Kalman filter to obtain a voltage estimation signal;

[0013] S6: Calculate a residual signal between the voltage signal and the voltage estimation signal;

[0014] S7: Calculate a half-cycle standard deviation according to the residual signal;

[0015] S8: Generate a state diagram with the cross-correlation parameter at each moment as the abscissa and the half-cycle standard deviation at each moment as the ordinate;

[0016] S9: According to the state diagram, a convolutional neural network is used to perform a detailed diagnosis to determine the fault type;

[0017] S10: Record the number, location, preliminary diagnosis result and detailed diagnosis result of the voltage sensor. When a fault is detected in the communication cable, use the location of the voltage sensor as the fault location for positioning and warning.

[0018] Second aspect:

[0019] An embodiment of the present invention provides a communication cable fault detection system, comprising:

[0020] processor;

[0021] A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the communication cable fault detection method as described in the first aspect is implemented.

[0022] The third aspect:

[0023] An embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the communication cable fault detection method as described in the first aspect is implemented.

[0024] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0025] In the present invention, a preliminary diagnosis is first performed based on the cross-correlation parameters, and then a detailed diagnosis is performed through the Kalman filter and the convolutional neural network, which can capture the initial fault in a timely and accurate manner, avoid the initial fault from turning into a permanent fault, and reduce the communication cable fault loss. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0027] Figure 1A schematic diagram of a flow chart of a communication cable fault detection method provided by an embodiment of the present invention;

[0028] Figure 2 A schematic structural diagram of a communication cable fault detection system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0030] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0031] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same. "of", "corresponding, relevant" and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same.

[0032] In the embodiments of the present invention, sometimes a subscript such as W1 may be mistakenly written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are consistent.

[0033] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0034] Reference Manual Attached Figure 1 , shows a flow chart of a communication cable fault detection method provided by an embodiment of the present invention.

[0035] The embodiment of the present invention provides a communication cable fault detection method, which can be implemented by a communication cable fault detection device, which can be a terminal or a server. The processing flow of the communication cable fault detection method can include the following steps:

[0036] S1: Monitor the voltage signals at various locations of the communication cable through voltage sensors.

[0037] In actual application, the voltage sensor can be installed at a suitable position according to the topological structure of the communication cable.

[0038] S2: Determine whether there is a disturbance based on the voltage signal. If so, go to S3; otherwise, return to S1 and continue monitoring.

[0039] Specifically, the voltage signal is a sine wave under an ideal state, and whether there is a disturbance can be determined by judging whether the current voltage signal still maintains a sine wave.

[0040] In a possible implementation manner, judging whether there is a disturbance according to the voltage signal in S2 specifically includes:

[0041] Determine whether the voltage signal meets the following conditions. If so, determine that there is no disturbance; otherwise, determine that there is a disturbance:

[0042] V1<|y t | <V2

[0043] Among them, V1 represents the lower limit of disturbance, V2 represents the upper limit of disturbance, and y t Represents the voltage signal at time t.

[0044] The disturbance lower limit value V1 is specifically:

[0045] V1=α1|y t-T / 2 |

[0046] Among them, α1 represents the lower limit coefficient of disturbance, y t-T / 2 It represents the voltage signal corresponding to the voltage signal at time t in the previous half cycle, and T represents the cycle.

[0047] Among them, those skilled in the art can set the size of the disturbance lower limit coefficient α1 according to actual conditions, and the present invention does not limit it.

[0048] The disturbance upper limit value V2 is specifically:

[0049] V2=α2|y t-T / 2 |

[0050] Among them, α2 represents the upper limit coefficient of disturbance.

[0051] Among them, those skilled in the art can set the size of the disturbance upper limit coefficient α2 according to actual conditions, and the present invention does not limit it.

[0052] In the present invention, by judging the upper and lower limits of the voltage signal, it is possible to more accurately determine whether a disturbance exists, thereby providing an important basis for subsequent fault detection and processing, and helping to improve the stability and reliability of the communication cable.

[0053] S3: Calculate the cross-correlation parameters based on the voltage signal.

[0054] In a possible implementation manner, the cross-correlation parameter is specifically calculated as follows:

[0055]

[0056] Among them, C i represents the cross-correlation parameter at time t, T represents the period, A1(t) represents the voltage signal of the half period before time t, A2(t) represents the average value of the voltage signal in the half cycle before time t, and A2(t) represents the voltage signal in the half cycle after time t. Represents the average value of the voltage signal in half a cycle after time t.

[0057] S4: Determine whether the cross-correlation parameter is greater than the preset cross-correlation value. If so, determine that the communication cable is initially diagnosed to have a fault and enter S5. Otherwise, return to S1 to continue monitoring.

[0058] Among them, those skilled in the art can set the size of the preset mutual correlation value according to actual conditions, and the present invention does not limit it.

[0059] Specifically, when an initial fault occurs, the voltage signal tends to deviate from the sinusoidal waveform. The cross-correlation parameters can be used to measure the degree to which the voltage signal deviates from the sinusoidal waveform, and the waveform characteristics of the voltage signal can be objectively evaluated without being affected by subjective factors.

[0060] In the present invention, the cross-correlation parameters are compared with preset cross-correlation values, and a preliminary fault judgment is made based on the comparison results, which can achieve automated, standardized, and timely response fault detection, thereby improving the efficiency and accuracy of fault diagnosis. At the same time, since refined detection is time-consuming, a preliminary diagnosis is performed at the front end of refined detection. Only when the preliminary diagnosis determines that a fault exists, refined detection is performed, which can save computing resources, reduce unnecessary refined detection, and thus reduce monitoring costs.

[0061] S5: Predict the voltage signal through the Kalman filter to obtain a voltage estimation signal.

[0062] Among them, the Kalman filter is a mathematical tool for estimating the state of a system, and is usually used to deal with dynamic systems with noise. The Kalman filter is a recursive filter that can reduce the estimation error by continuously updating the estimated value of the system state. When processing voltage signals, the Kalman filter can be used to remove noise, smooth signals, or decompose signals into different components. By properly designing the state space model and the observation equation, the Kalman filter can be applied to the decomposition of the voltage signal to obtain the basic components of the voltage signal.

[0063] In a possible implementation, S5 specifically includes sub-steps S501 to S505:

[0064] S501: The voltage estimation signal under the ideal state is represented by a cosine signal:

[0065]

[0066] Among them, S t represents the voltage estimation signal at time t, a represents the amplitude, ω0 represents the angular frequency, t represents the time, Indicates phase.

[0067] S502: Based on the basic properties of the cosine signal, construct the voltage dynamic equation:

[0068] S t+1 +S t-1 =2cos(ω0)S t +ψ t

[0069] Among them, S t+1 represents the voltage estimation signal at time t+1, S t-1 represents the voltage estimation signal at time t-1, ψ t represents the error at time t.

[0070] S503: According to the voltage dynamic equation, the state space of the Kalman filter is constructed:

[0071] X t+1 =MX t +bψ t

[0072] y t =h T X t +v t

[0073] Among them, y t represents the voltage signal at time t in the actual state, X represents the state vector, h represents the measurement matrix, h = [1 0], () T Represents the transpose operation of a matrix.

[0074] It should be noted that by constructing the voltage dynamic equation and the state space model of the Kalman filter, the voltage signal can be decomposed into basic components and measurement errors under ideal conditions. This can more clearly understand the composition structure of the voltage signal and facilitate further analysis and processing.

[0075] S504: predicting the voltage signal in the actual state measured by the voltage sensor through a Kalman filter to obtain a voltage estimation signal.

[0076] It should be noted that the Kalman filter is a very effective prediction and estimation tool and is widely used in noisy dynamic systems to estimate the state of the system. It is a recursive filter that can estimate the state of a dynamic system from a series of incomplete and noisy measurements. The Kalman filter first predicts the state based on the estimate of the previous state and the physical model of the system. This prediction includes an estimate of the state at the next moment, without considering new measurement data. When a new voltage measurement is provided by the sensor, the Kalman filter compares this new measurement with the predicted state and adjusts the state estimate to more accurately reflect the actual measurement. This step involves the statistical characteristics of the uncertainty of the measurement (noise) and the uncertainty of the prediction. Based on the error in the prediction and the error in the actual measurement, the Kalman filter adjusts the prediction by calculating a coefficient called the "Kalman gain". This gain determines the weight of the measured data and the predicted data to minimize the estimation error. The process is repeated, and with each new measurement update, the Kalman filter continuously adjusts its estimate, gradually approaching the true state of the system.

[0077] Furthermore, in the case of communication cables, the Kalman filter can effectively predict voltage signals, mainly because voltage signals are often affected by multiple factors, such as cable aging, environmental interference, etc. By building a dynamic model that includes these factors, the Kalman filter can accurately predict voltage changes in actual measurements, thereby helping maintenance teams to identify potential voltage problems and faults in a timely manner and perform more effective maintenance and management. This method plays a key role in voltage monitoring and fault prediction through its excellent noise suppression capabilities and prediction accuracy.

[0078] In a possible implementation manner, S504 specifically includes:

[0079] Initialize the state vector and error covariance.

[0080] Calculate the instantaneous Kalman gain of the Kalman filter:

[0081] K t =P t -h T (hP t h T +R t ) -1

[0082] Among them, K t represents the Kalman gain at time t, P t represents the error covariance at time t, h represents the measurement matrix, R t represents the measurement noise at time t.

[0083] According to the instantaneous Kalman gain, the state vector is updated:

[0084]

[0085] in, represents the estimated value of the state vector at time t, represents the current value of the state vector at time t, y t Represents the voltage signal at time t under actual conditions.

[0086] It should be noted that the Kalman filter utilizes the relationship between the state estimation and the measured value, and updates the state vector through the instantaneous Kalman gain, so that the estimated value is closer to the true value, which can improve the estimation accuracy of the voltage estimation signal.

[0087] According to the updated state vector, the voltage estimation signal at each moment is determined.

[0088] In the present invention, the voltage signal under the actual state measured by the voltage sensor is predicted by a Kalman filter to solve the voltage estimation signal, which can improve the accuracy, real-time, adaptability and anti-interference ability of the decomposition result, while saving resources, and helping to improve the effect and efficiency of communication cable fault detection and diagnosis.

[0089] S6: Calculate a residual signal between the voltage signal and the voltage estimation signal.

[0090] In a possible implementation manner, the residual signal is calculated as follows:

[0091] VSR t =y t -S t

[0092] Among them, VSR t represents the residual signal at time t, S t represents the voltage estimation signal at time t, y t Represents the voltage signal at time t under the actual state measured by the voltage sensor.

[0093] It should be noted that the residual signal is the difference between the actual measured voltage signal and the voltage estimation signal predicted by the model (such as the Kalman filter). This difference reflects the degree of deviation between the actual signal and the expected behavior. When the communication cable is working properly, the residual signal should be small because the actual signal will be very close to the output of the prediction model. If the residual is large, this may indicate that the cable is abnormal or faulty. At the same time, the size and pattern of the residual can help determine the nature and location of the fault. For example, if the voltage signal of a particular section is always lower than the predicted value, this may indicate that there is a problem with broken wires or poor contact in the section. Statistical analysis of the residual signal (such as calculating the standard deviation of the half cycle) further provides information about the stability and severity of the fault. By analyzing the time series of the residual signal, small changes that may be overlooked by basic voltage sensor measurements can be detected. Combining advanced analysis techniques (such as convolutional neural networks) to analyze the residual signal can accurately diagnose the type of fault, making maintenance and repair work more targeted and effective.

[0094] S7: Calculate the half-cycle standard deviation based on the residual signal.

[0095] In a possible implementation manner, the half-period standard deviation is calculated as follows:

[0096]

[0097] Among them, Sd t represents the half-period standard deviation at time t, T represents the period, Represents the mean value of the residual signal.

[0098] It should be noted that the half-cycle standard deviation is a statistic of the residual signal, which is used to measure the degree of change of the residual signal. When the half-cycle standard deviation is large, it indicates that the residual signal fluctuates greatly, and there may be anomalies or faults. Therefore, by calculating the half-cycle standard deviation, it can help detect abnormal situations and discover potential problems in time.

[0099] S8: Generate a state diagram with the cross-correlation parameter at each moment as the horizontal coordinate and the half-cycle standard deviation at each moment as the vertical coordinate.

[0100] It should be noted that the convolutional neural network has a better processing effect on image data, so the state diagram is constructed based on the cross-correlation parameters and the half-cycle standard deviation.

[0101] S9: Based on the state diagram, a convolutional neural network is used to perform detailed diagnosis and determine the fault type.

[0102] Among them, Convolutional Neural Network (CNN) is a deep learning model specially designed for processing neural networks with grid-structured data.

[0103] In a possible implementation, the convolutional neural network includes an input layer, a convolution layer, a pooling layer, a fully connected layer, a classification layer, and an output layer, and S9 specifically includes:

[0104] S901: In the input layer, obtain the state diagram.

[0105] S902: In the convolution layer, extract the features of the state diagram:

[0106]

[0107] in, represents the output of the current convolutional layer, represents the output of the previous convolutional layer, Represents the convolution kernel weight of the current convolution layer, Represents the bias term of the current convolutional layer, M j represents the selected input feature map and σ() represents the activation function.

[0108] The convolution layer extracts local features of the input data by applying a series of learnable filters (also called convolution kernels). Each filter performs a sliding convolution operation on the input data to generate a series of feature maps, where each feature map corresponds to a local feature of the input data. The convolution operation can effectively capture the spatial structure information of the input data, and through the back-propagation algorithm, the network can automatically learn the optimal filter parameters.

[0109] S903: In the pooling layer, the extracted features are reduced in dimension:

[0110]

[0111] Among them, down represents the downsampling function, represents the output of the current pooling layer, Represents the output of the previous pooling layer.

[0112] Among them, the pooling layer is used to reduce the spatial dimension of the feature map, thereby reducing the number of network parameters and the amount of computation, and improving the robustness of the network. Commonly used pooling operations include maximum pooling and average pooling, which take the maximum value or average value in the feature map as output.

[0113] It should be noted that the convolutional neural network can gradually extract and combine abstract features in the state diagram through multiple convolutional layers and pooling layers, so as to conduct a more comprehensive analysis and diagnosis of the fault. This helps to identify multiple aspects and characteristics of the fault and improve the accuracy and robustness of the diagnosis.

[0114] S904: In the fully connected layer, the features are summarized to obtain the hidden state.

[0115] Among them, the fully connected layer is used to summarize and integrate the features extracted by the convolution layer and the pooling layer, and map them to the space of the output category. The fully connected layer usually contains multiple neurons, each of which is connected to all neurons in the previous layer.

[0116] It should be noted that convolutional neural networks can automatically learn and extract features from state graphs without manually designing feature extractors. This enables the system to better adapt to state graphs of different types and complexities, and can capture potential patterns and rules that are difficult to identify manually.

[0117] S905: In the classification layer, the probability value of each fault type is calculated:

[0118] P t =Softmax(W δh h t +b δ )

[0119] P t =[P 1t ,…,P it ,…,P nt ]

[0120] Among them, P t represents the probability value of each fault type at time t, Softmax() represents the Softmax activation function, and h t represents the hidden state at time t, W δh represents the weight between the fully connected layer and the classification layer, b δ represents the bias between the fully connected layer and the classification layer, P it It represents the probability value of the i-th fault type at time t, i = 1, 2, …, n, and n represents the total number of fault types.

[0121] S906: In the output layer, the fault type with the largest probability value is taken as the detailed diagnosis result and outputted.

[0122] In the present invention, fine diagnosis through convolutional neural network can automatically extract features, comprehensively analyze state diagrams, adapt to complex situations, achieve real-time and efficiency, improve accuracy and consistency, so as to better realize the diagnosis and processing of communication cable faults, and can capture early faults in a timely and accurate manner, avoid early faults from turning into permanent faults, and reduce communication cable fault losses.

[0123] In a possible implementation, the convolutional neural network is trained as follows:

[0124] Taking the minimum mean square error loss function as the goal, the optimal network parameters of the convolutional neural network are determined by using an improved particle swarm optimization algorithm.

[0125] Among them, the mean square error loss function is specifically:

[0126]

[0127] Among them, L() represents the loss function, θ represents the model parameters, and z i represents the fine diagnosis result of the i-th training sample, Represents the classification label of the i-th training sample, i = 1, 2, ..., N, and N represents the total number of training samples in the sample data set.

[0128] Specifically, the inverse of the mean square error loss function is used as the fitness function of the particle swarm optimization algorithm.

[0129] Initialize a particle swarm, which contains multiple particles. Each particle represents a feasible set of convolutional neural network parameters. The convolutional neural network parameters include convolution kernel size, number of convolution kernels, convolution step, pooling window size, pooling step, number of neurons, learning rate, and drop rate.

[0130] Calculate the fitness of each particle;

[0131] Determine individual optimal positions and global optimal positions;

[0132] Each particle updates its speed and position based on its individual optimal position and the global optimal position:

[0133]

[0134] in, represents the velocity of the i-th particle at the t+1th iteration, represents the velocity of the i-th particle at the t-th iteration, ω represents the inertia weight factor, c1 represents the individual optimal acceleration factor, c2 represents the global optimal acceleration factor, r1 and r2 both represent random numbers between 0 and 1, represents the individual optimal position of the i-th particle at the t-th iteration, represents the global optimal position at the tth iteration, represents the position of the i-th particle at the t-th iteration, h represents the flight coefficient, represents the position of the i-th particle at the t+1th iteration.

[0135] Among them, those skilled in the art can set the size of the inertia weight factor ω according to actual conditions, and the present invention does not limit this.

[0136] Optionally, the inertia weight factor ω is specifically:

[0137]

[0138] Among them, ω t represents the inertia weight factor at the tth iteration, t max represents the maximum number of iterations, t represents the current number of iterations, λ represents the decrease coefficient, and exp represents the exponential function with the natural constant e as the base.

[0139] It should be noted that as the number of iterations t increases, the value of the inertia weight factor ω gradually decreases, which helps to improve the exploratory nature of particles in the early stages of the algorithm and increase the algorithm's ability to search extensively in the solution space. In the later stages of the iteration, the inertia weight factor ω decreases, which helps particles concentrate on searching for local optimal solutions and improves the convergence of the algorithm.

[0140] Update the fitness of each particle.

[0141] Calculate the neighbor particles of each particle:

[0142]

[0143] Among them, N i represents the neighbors of the i-th particle, f() represents the fitness function, represents the fitness of the i-th particle at the t-th iteration, represents the fitness of the global optimal position of the ith particle at the tth iteration, exp[] represents an exponential function with the natural constant e as the base, R represents a 1×d random matrix consisting of d random numbers in the range [-1,1], d represents the dimension of the particle, and ⊙ represents an element-by-element multiplication operation.

[0144] It should be noted that when calculating neighbor particles, the comparison of fitness function values ​​is used, that is, the position of neighbor particles is determined by comparing the fitness of the current particle with the fitness of the global optimal particle. Dynamic and independent information sharing is carried out by continuously updating the current position and the global optimal position, and then their local information is fully developed to dynamically adjust the position of neighbor particles, making it more likely to move in the direction of higher fitness, thereby accelerating the convergence speed of the algorithm.

[0145] Calculate the fitness of each neighbor particle.

[0146] If the neighbor particle has a higher fitness, the current particle is replaced by the neighbor particle, otherwise, the current particle remains in its original state.

[0147] It should be noted that by calculating neighbor particles, the neighborhood information between particles can be introduced, making the search process more localized, which is conducive to a more refined search near the local optimal solution. When the fitness of the neighbor particle is higher, the current particle can be directly replaced by the neighbor particle, thereby accelerating the convergence speed of the search process, reducing the number of unnecessary iterations, and improving the efficiency of the algorithm.

[0148] Update individual optimality and global optimality.

[0149] Determine whether the current number of iterations has reached the maximum number of iterations; if so, output the network parameter set represented by the particle with the highest fitness as the optimal network parameter set; otherwise, return to continue iterating.

[0150] In the present invention, the optimal network parameters of the convolutional neural network are determined by an improved particle swarm optimization algorithm, which can improve the accuracy of fault detection.

[0151] S10: Record the number, location, preliminary diagnosis result and detailed diagnosis result of the voltage sensor. When a fault is detected in the communication cable, use the location of the voltage sensor as the fault location for positioning and warning.

[0152] In the present invention, by recording the position of the sensor, the fault position can be located quickly and accurately, so that maintenance personnel can quickly arrive at the site for repair, thus shortening the fault handling time.

[0153] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0154] In the present invention, a preliminary diagnosis is first performed based on the cross-correlation parameters, and then a detailed diagnosis is performed through the Kalman filter and the convolutional neural network, which can capture the initial fault in a timely and accurate manner, avoid the initial fault from turning into a permanent fault, and reduce the communication cable fault loss.

[0155] Reference Manual Attached Figure 2 , showing a structural schematic diagram of a communication cable fault detection system provided by the present invention.

[0156] The present invention further provides a communication cable fault detection system 20, which is applied to the above-mentioned communication cable fault detection method, comprising:

[0157] Processor 201;

[0158] The memory 202 stores computer-readable instructions, and when the computer-readable instructions are executed by the processor 201, the communication cable fault detection method described in the method embodiment is implemented.

[0159] The communication cable fault detection system 20 provided by the present invention can execute the above-mentioned communication cable fault detection method and achieve the same or similar technical effects. To avoid repetition, the present invention will not go into details.

[0160] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0161] In the present invention, a preliminary diagnosis is first performed based on the cross-correlation parameters, and then a detailed diagnosis is performed through the Kalman filter and the convolutional neural network, which can capture the initial fault in a timely and accurate manner, avoid the initial fault from turning into a permanent fault, and reduce the communication cable fault loss.

[0162] It should be understood that the processor in the embodiment of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0163] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0164] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0165] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0166] In the present invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0167] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0168] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0169] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0170] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0171] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0172] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0173] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0174] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the communication cable fault detection method as described in the method embodiment.

[0175] A computer-readable storage medium provided by the present invention can implement the steps and effects of the communication cable fault detection method of the above method embodiment. To avoid repetition, the present invention will not go into details.

[0176] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0177] In the present invention, a preliminary diagnosis is first performed based on the cross-correlation parameters, and then a detailed diagnosis is performed through the Kalman filter and the convolutional neural network, which can capture the initial fault in a timely and accurate manner, avoid the initial fault from turning into a permanent fault, and reduce the communication cable fault loss.

[0178] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0179] There are a few points to note:

[0180] (1) The drawings of the embodiments of the present invention only relate to the structures related to the embodiments of the present invention, and other structures may refer to the general design.

[0181] (2) For the sake of clarity, in the drawings used to describe the embodiments of the present invention, the thickness of the layers or regions is exaggerated or reduced, that is, these drawings are not drawn according to the actual scale. It is understood that when an element such as a layer, film, region or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or there may be intermediate elements.

[0182] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to obtain new embodiments.

[0183] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A communication cable fault detection method, characterized in that: include: S1: Monitor the voltage signals at various locations of the communication cable through a voltage sensor; S2: judging whether there is disturbance according to the voltage signal, if so, entering S3, otherwise, returning to S1 to continue monitoring; S3: Calculating a cross-correlation parameter according to the voltage signal; S4: Determine whether the cross-correlation parameter is greater than a preset cross-correlation value. If so, determine that the communication cable is initially diagnosed to have a fault and proceed to S5. Otherwise, return to S1 to continue monitoring. S5: predicting the voltage signal through a Kalman filter to obtain a voltage estimation signal; S6: Calculate a residual signal between the voltage signal and the voltage estimation signal; S7: Calculate a half-cycle standard deviation according to the residual signal; S8: Generate a state diagram with the cross-correlation parameter at each moment as the abscissa and the half-cycle standard deviation at each moment as the ordinate; S9: According to the state diagram, a convolutional neural network is used to perform a detailed diagnosis to determine the fault type; S10: Record the number, location, preliminary diagnosis result and detailed diagnosis result of the voltage sensor. When a fault is detected in the communication cable, use the location of the voltage sensor as the fault location for positioning and warning.

2. The communication cable fault detection method according to claim 1, characterized in that: The step of determining whether there is a disturbance according to the voltage signal in S2 specifically includes: Determine whether the voltage signal satisfies the following conditions. If so, determine that there is no disturbance; otherwise, determine that there is a disturbance: V1<|y t | <V2 Among them, V1 represents the lower limit of disturbance, V2 represents the upper limit of disturbance, and y t Represents the voltage signal at time t; The disturbance lower limit value V1 is specifically: V1=α1|y t-T / 2 | Among them, α1 represents the lower limit coefficient of disturbance, y t-T / 2 represents the voltage signal corresponding to the voltage signal at time t in the previous half cycle, where T represents the cycle; The disturbance upper limit value V2 is specifically: <h2 style=";text-align:left;direction:ltr">V2 = α2|y<h2 style=";text-align:left;direction:ltr"> t-T / 2 <h2 style=";text-align:left;direction:ltr"> | Among them, α2 represents the upper limit coefficient of disturbance.

3. The communication cable fault detection method according to claim 1, characterized in that: The cross-correlation parameter is calculated as follows: Among them, C i represents the cross-correlation parameter at time t, T represents the period, A1(t) represents the voltage signal of the half period before time t, A2(t) represents the average value of the voltage signal in the half cycle before time t, and A2(t) represents the voltage signal in the half cycle after time t. Represents the average value of the voltage signal in half a cycle after time t.

4. The communication cable fault detection method according to claim 1, characterized in that: The S5 specifically includes: S501: The voltage estimation signal under the ideal state is represented by a cosine signal: Among them, S t represents the voltage estimation signal at time t, a represents the amplitude, ω0 represents the angular frequency, t represents the time, Indicates phase; S502: Based on the basic properties of the cosine signal, construct the voltage dynamic equation: S t+1 +S t-1 =2cos(ω0)S t +ψ t Among them, S t+1 represents the voltage estimation signal at time t+1, S t-1 represents the voltage estimation signal at time t-1, ψ t represents the error at time t; S503: Constructing the state space of the Kalman filter according to the voltage dynamic equation: X t+1 =MX t +bψ t y t =h T X t +v t Among them, y t represents the voltage signal at time t in the actual state, X represents the state vector, and X t+1 = h represents the measurement matrix, h = [1 0], () T Represents the transpose operation of a matrix; S504: predicting the voltage signal in the actual state measured by the voltage sensor through a Kalman filter to obtain a voltage estimation signal.

5. The communication cable fault detection method according to claim 4, characterized in that: The S504 specifically includes: Initialize the state vector and error covariance; Calculate the instantaneous Kalman gain of the Kalman filter: K t =P t -h T (hP t h T +R t ) -1 Among them, K t represents the Kalman gain at time t, P t represents the error covariance at time t, h represents the measurement matrix, R t represents the measurement noise at time t; According to the instantaneous Kalman gain, the state vector is updated: in, represents the estimated value of the state vector at time t, represents the current value of the state vector at time t, y t Represents the voltage signal at time t in the actual state; According to the updated state vector, the voltage estimation signal at each moment is determined.

6. The communication cable fault detection method according to claim 1, characterized in that: The residual signal is calculated as follows: VSR t =y t -S t Among them, VSR t represents the residual signal at time t, S t represents the voltage estimation signal at time t, y t Represents the voltage signal at time t under the actual state measured by the voltage sensor.

7. The communication cable fault detection method according to claim 1, characterized in that: The calculation method of the half-cycle standard deviation is specifically as follows: Among them, Sd t represents the half-period standard deviation at time t, T represents the period, Represents the mean value of the residual signal.

8. The communication cable fault detection method according to claim 1, characterized in that: The convolutional neural network includes an input layer, a convolution layer, a pooling layer, a fully connected layer, a classification layer and an output layer. The S9 specifically includes: S901: In the input layer, obtaining the state diagram; S902: In the convolution layer, extract the features of the state diagram: in, represents the output of the current convolutional layer, represents the output of the previous convolutional layer, Represents the convolution kernel weight of the current convolution layer, Represents the bias term of the current convolutional layer, M j represents the selected input feature map, σ() represents the activation function; S903: In the pooling layer, the extracted features are reduced in dimension: Among them, down represents the downsampling function, represents the output of the current pooling layer, Represents the output of the previous pooling layer; S904: In the fully connected layer, the features are summarized to obtain a hidden state; S905: In the classification layer, the probability value of each fault type is calculated: P t =Softmax(W δh h t +b δ ) P t =[P 1t ,…,P it ,…,P nt ] Among them, P t represents the probability value of each fault type at time t, Softmax() represents the Softmax activation function, and h t represents the hidden state at time t, W δh represents the weight between the fully connected layer and the classification layer, b δ represents the bias between the fully connected layer and the classification layer, P it represents the probability value of the i-th fault type at time t, i = 1, 2, ..., n, and n represents the total number of fault types; S906: In the output layer, the fault type with the largest probability value is taken as the detailed diagnosis result and outputted.

9. The communication cable fault detection method according to claim 8, characterized in that: The training method of the convolutional neural network is specifically as follows: Taking the minimum mean square error loss function as the goal, the optimal network parameters of the convolutional neural network are determined by using an improved particle swarm optimization algorithm.

10. A communication cable fault detection system, characterized in that: include: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the communication cable fault detection method according to any one of claims 1 to 9 is implemented.

Images