Intelligent fault detection method for dual-mode communication unit

Abstract

The invention discloses an intelligent fault detection method for a dual-mode communication unit, relates to the technical field of communication fault detection, and aims to solve the problems that the installation environment of the dual-mode communication unit currently working in a complex environment is cramped, and workers cannot enter the dual-mode communication unit conveniently, so that time and labor are wasted in daily detection, and the detection efficiency is high. In order to solve the technical problems of low detection efficiency, low accuracy and incapability of finding faults in time in the prior art, the intelligent fault detection method for the dual-mode communication unit comprises the following steps: data acquisition: monitoring the working state of the dual-mode communication unit in real time through a sensor, acquiring parameter data including signal strength, communication frequency, temperature and voltage, and sending the parameter data to a server; the temperature and the voltage are directly read by the temperature sensor and the voltage sensor respectively, data are modeled and predicted, then real-time monitoring is performed to find faults, and the faults are diagnosed and processed, so that the daily detection efficiency of the dual-mode communication unit in a severe environment is greatly improved, and time and labor are saved.

Description

Technical Field

[0001] The present invention relates to the technical field of communication fault detection, and more specifically, to an intelligent fault detection method for a dual-mode communication unit. Background Art

[0002] At present, with the continuous development of communication technology, dual-mode communication units have been widely used in various fields. However, due to the complexity of dual-mode communication units and the diversity of working environments, the probability of their failure is relatively high.

[0003] At present, the fault detection of dual-mode communication units mainly relies on manual detection or a single detection method. The dual-mode communication units working in complex environments are inconvenient for staff to enter due to the cramped installation environment, which leads to time-consuming and laborious daily detection, low detection efficiency, low accuracy, and inability to detect faults in a timely manner.

[0004] In view of this, we propose an intelligent fault detection method for dual-mode communication units. Summary of the invention

[0005] The purpose of the present invention is to provide an intelligent fault detection method for a dual-mode communication unit to solve the technical problems that the dual-mode communication units currently working in complex environments are time-consuming and labor-intensive due to the cramped installation environment and inconvenient for staff to enter, resulting in low detection efficiency, low accuracy and inability to detect faults in a timely manner.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: an intelligent fault detection method for a dual-mode communication unit, comprising the following steps:

[0007] S1, data collection;

[0008] The working status of the dual-mode communication unit is monitored in real time through sensors, and parameter data including signal strength, communication frequency, temperature, and voltage are collected. The temperature and voltage are directly read by the temperature sensor and the voltage sensor respectively;

[0009] Signal strength can be expressed in dBm, and the calculation formula is:

[0010] P(dBm)=10×log(P(mW));

[0011] Wherein, P (mW) is the power value in milliwatts;

[0012] S2, single point detection;

[0013] Perform separate tests on each key node in the dual-mode communication unit to check whether its working status is normal, such as whether the working temperature and voltage of the chip are within the normal range, and whether the connection of the communication interface is stable;

[0014] S201, chip temperature;

[0015] For chip temperature, the normal temperature range can be set as:

[0016] T min ≤T≤T max ;

[0017] Wherein, T is the temperature value measured in real time;

[0018] T min and T max are the minimum and maximum allowable temperatures determined according to the chip specifications;

[0019] S202, chip voltage;

[0020] For chip voltage, the normal voltage range can be set as:

[0021] V min ≤V≤V max ;

[0022] Wherein, V is the temperature value measured in real time;

[0023] V min and V max are the minimum and maximum allowable temperatures determined according to the chip specifications;

[0024] S3, network detection;

[0025] Test the performance of the dual-mode communication unit in the networking state to check whether the network connection is stable, whether the communication between communication nodes is smooth, and whether the network throughput and delay indicators meet the requirements;

[0026] S301, network throughput;

[0027] Network throughput is expressed as the amount of data transmitted per unit time, such as the number of bits transmitted per second (bps). The calculation formula is:

[0028]

[0029] Among them, DatrVolume is the total amount of data transmitted in a period of time;

[0030] Time is the time interval;

[0031] S302, network delay;

[0032] Network latency is usually expressed in milliseconds (ms) and can be calculated by sending a test packet and measuring the round-trip time using the formula:

[0033]

[0034] Among them, RoundTripTime is the round trip time of the test data packet;

[0035] S4, communication quality detection;

[0036] Detect the communication quality of the dual-mode communication unit, including signal strength, signal-to-noise ratio, and bit error rate. By detecting the communication quality, problems in communication can be discovered in a timely manner;

[0037] S5, anti-interference test;

[0038] Conduct anti-interference tests on the dual-mode communication unit, simulate various interference sources, detect the communication performance of the dual-mode communication unit in an interference environment, use a signal generator to generate interference signals of different frequencies and intensities, and then measure the signal strength, signal-to-noise ratio, and bit error rate indicators of the dual-mode communication unit under interference. The anti-interference performance can be measured by the interference suppression ratio, and the calculation formula is:

[0039]

[0040] Among them, P signal with out inter ference is the signal power without interference;

[0041] P signal with inter ference is the signal power when there is interference;

[0042] S6. Data analysis;

[0043] The collected data is analyzed and processed, and machine learning algorithms are used to model and predict the data to determine whether the dual-mode communication unit has a fault;

[0044] S7, Fault analysis and location;

[0045] If the prediction result shows that the dual-mode communication unit is faulty, the fault is further analyzed and located;

[0046] S8, fault handling and verification;

[0047] Take appropriate troubleshooting measures according to the type and location of the fault. At the same time, re-test the dual-mode communication unit after the fault is repaired to verify whether the fault has been completely repaired. If the fault still exists, re-diagnose and handle the fault until the fault is completely repaired.

[0048] S9, optimize the communication transmission path;

[0049] According to the results of fault detection and processing, the communication transmission path of the dual-mode communication unit is optimized. If the signal strength of a communication node is weak, the communication path can be adjusted to avoid the node. If interference is found in a communication link, other communication links can be selected for communication.

[0050] The Dijkstra algorithm is used to optimize the communication transmission path, and the optimal communication path is determined by calculating the shortest path from each node to the target node:

[0051] Let the communication network be G = (V, E);

[0052] Among them, V is the node set;

[0053] E is the edge set;

[0054] For each node v∈V, let d(v) be the shortest distance from the source node to v, and p(v) be the predecessor node of v;

[0055] At initialization, d(s) = 0 (where s is the source node), for other nodes v≠s, d(v) = ∞;

[0056] For each node u whose shortest distance is not determined, select the node with the smallest d(u) and update the shortest distance of its adjacent nodes. Repeat the process until the shortest distances of all nodes are determined. Finally, by backtracking the predecessor nodes, the shortest path from the source node to the target node can be obtained.

[0057] The present invention uses a machine learning algorithm to model and predict data, and can quickly and accurately determine whether a dual-mode communication unit has a fault. In conjunction with real-time monitoring of the working status of the dual-mode communication unit through sensors, the efficiency and accuracy of fault detection are improved, and faults can be discovered in a timely manner, and the faults can be diagnosed and processed, which greatly improves the daily detection efficiency of the dual-mode communication unit in harsh environments, saving time and effort.

[0058] Preferably, the S4 communication quality detection further includes the following steps:

[0059] S401, signal-to-noise ratio;

[0060] The signal-to-noise ratio calculation formula is:

[0061]

[0062] Among them, P signal is the signal power;

[0063] P noise is the noise power;

[0064] S402, bit error rate;

[0065] The bit error rate calculation formula is:

[0066]

[0067] Among them, BER is the ratio of the number of error bits to the total number of transmitted bits;

[0068] Preferably, the S6 data analysis further comprises the following steps:

[0069] S601, data preprocessing;

[0070] Clean, denoise and normalize the collected data to improve the quality and usability of the data;

[0071] S602, feature extraction;

[0072] Extract features that can reflect the failure of the dual-mode communication unit from the preprocessed data, such as the rate of change of signal strength, deviation of communication frequency, and abnormal temperature change;

[0073] S603, model training;

[0074] Use historical data to train the machine learning model and build a fault detection model using a support vector machine, whose goal is to find a hyperplane that allows data points of different categories to be separated as correctly as possible while maximizing the interval;

[0075] Assume that the training data set is: {(X1,y1),(X2,y2),...,(X i ,y i )};

[0076] Among them, X i is the input feature vector;

[0077] y i ∈{-1,1} is the category label;

[0078] The optimization problem of support vector machine can be expressed as:

[0079] Where ω is the normal vector of the hyperplane;

[0080] b is the intercept;

[0081] C is the penalty parameter;

[0082] ξ i is the slack variable

[0083] The constraints are: i (ω T X i +b)≥1-ξ i ,ξ i ≥0;

[0084] The fault detection model is obtained by solving the above problems;

[0085] S604, fault prediction;

[0086] Input the real-time collected data into the fault detection model to predict whether there is a fault in the dual-mode communication unit;

[0087] Preferably, the data preprocessing in step S601 further includes the following steps:

[0088] S601a, data cleaning;

[0089] Outliers are removed through data cleaning. The Laida criterion is used. If the difference between the measured value and the mean value is greater than three times the standard deviation, the measured value is considered an outlier and removed:

[0090] Assume the data set is: {X1,X2...,X n};

[0091] The average value is:

[0092] The standard deviation is:

[0093] like Then X i is an outlier;

[0094] S601b, data denoising;

[0095] Data denoising uses methods such as mean filtering, and the new data value is the average of several surrounding data. Suppose the current data point is X i , the surrounding K data points are:

[0096] {X i-j ,...,X i-j ,X i+j ,...,X i+j};

[0097] in (k is an even number) or (k is an odd number);

[0098] The denoised data value is:

[0099]

[0100] S601c, data normalization;

[0101] Data normalization can use the maximum and minimum normalization method, and the calculation formula is:

[0102]

[0103] Where X is the original data value;

[0104] X min and X max is the minimum and maximum value of the data;

[0105] Preferably, the feature extraction in S602 further includes the following steps:

[0106] S602a, signal strength change rate;

[0107] The signal strength change rate calculation formula is:

[0108]

[0109] Among them, P1 is the signal strength at the previous moment;

[0110] P0 is the signal strength at the previous moment;

[0111] S602b, communication frequency deviation;

[0112] The communication frequency deviation calculation formula is:

[0113]

[0114] Wherein, F is the current communication frequency;

[0115] F0 is the standard communication frequency;

[0116] S602c, abnormal temperature change;

[0117] Abnormal temperature changes can be measured by calculating the temperature change rate. Assuming the current temperature is T1, the temperature at the previous moment is T0, and the time interval is Δt, the temperature change rate is:

[0118]

[0119] If the temperature change rate exceeds the threshold R, the temperature is considered abnormal;

[0120] Preferably, the S7 fault analysis and location also includes the following steps:

[0121] S701, fault type determination;

[0122] Determine the type of fault based on its characteristics and manifestations, whether it is a hardware fault, software fault, or communication fault;

[0123] S702, fault location positioning;

[0124] By analyzing the working principle and structure of the dual-mode communication unit and combining the characteristics of the fault, the fault location is located;

[0125] Preferably, the S701 fault type determination further includes the following steps:

[0126] S701a, hardware failure determination;

[0127] Assume that the judgment standard of abnormal temperature rise is that the temperature exceeds the threshold T thershold , that is, T>T thershold It is judged that there may be a hardware failure;

[0128] The voltage instability criterion is that the percentage of voltage deviation from the normal range exceeds the threshold V percentage ,Right now It is judged that there may be a hardware failure;

[0129] S701b, software fault determination;

[0130] The judgment mark of program crash is set as the appearance of a specific error code or exception information;

[0131] S701c, communication failure determination;

[0132] Assume that the criterion for judging the increase of the bit error rate is that the bit error rate exceeds the threshold BER threshold , that is, BER>BER threshold It is judged that there may be a communication failure;

[0133] Preferably, the fault location locating step S702 further includes the following steps:

[0134] S702a, hardware fault location: If it is a hardware fault, the fault location can be determined by detecting the on-off state of the circuit, the working state of the chip, etc.;

[0135] S702b, software fault location: If it is a software fault, the fault location can be determined by analyzing the program's running log, error code, etc.;

[0136] An intelligent fault detection system for a dual-mode communication unit, comprising a working state real-time monitoring module for monitoring the working state of the dual-mode communication unit, a key node real-time detection module for detecting key nodes on the dual-mode communication unit, a networking state performance detection module for real-time detection of the networking state performance of the dual-mode communication unit, a communication quality real-time monitoring module for detecting the communication quality of the dual-mode communication unit, and an anti-interference real-time testing module for performing an anti-interference test on the dual-mode communication unit;

[0137] The working status real-time monitoring module, the key node real-time detection module, the networking status performance detection module, the communication quality real-time monitoring module and the anti-interference real-time testing module are electrically connected to an intelligent fault real-time monitoring unit for real-time fault detection of the dual-mode communication unit in combination with the collected data;

[0138] The intelligent fault real-time monitoring unit includes a data analysis and processing module for analyzing and processing data, and the data analysis and processing module is electrically connected to a machine learning algorithm modeling module for modeling based on a machine learning algorithm and a data preprocessing module for preprocessing the analyzed data;

[0139] The machine learning data modeling module generates a fault detection model, and the fault detection model is electrically connected to a machine learning algorithm prediction module for fault prediction and a machine learning model training module for training the fault detection model, and the machine learning algorithm prediction module and the data preprocessing module are both electrically connected to the machine learning training module;

[0140] The machine learning model training module is electrically connected to a fault analysis and positioning unit for fault analysis and positioning, and the fault analysis and positioning unit is electrically connected to a fault handling and verification module for providing fault handling and verification suggestions, and the fault handling and verification module is electrically connected to a communication transmission path optimization module for optimizing the communication transmission path.

[0141] Preferably, the fault analysis and locating unit comprises a fault type determination module for determining the fault type and a fault location location unit for locating the fault location, and the fault type determination module is electrically connected to the fault location location unit;

[0142] The fault location locating unit includes a hardware fault locating module for locating a hardware fault and a software fault locating module for locating a software fault.

[0143] Compared with the prior art, the present invention has the following beneficial effects:

[0144] 1. The present invention uses a machine learning algorithm to model and predict data, and can quickly and accurately determine whether the dual-mode communication unit has a fault. In conjunction with real-time monitoring of the working status of the dual-mode communication unit through sensors, the efficiency and accuracy of fault detection are improved, and faults can be discovered in a timely manner, and the faults can be diagnosed and processed, which greatly improves the daily detection efficiency of the dual-mode communication unit in harsh environments, saving time and effort;

[0145] 2. The present invention collects the operation data of the dual-mode communication unit in real time through the working status real-time monitoring module, the key node real-time detection module, the networking status performance detection module, the communication quality real-time monitoring module and the anti-interference real-time testing module, and cooperates with the intelligent fault real-time monitoring unit to combine the collected data for fault analysis, without relying on manual detection, reducing the manpower cost and time cost, thereby reducing the cost of fault detection;

[0146] 3. Through single-point detection, networking detection and communication quality detection, the performance of the dual-mode communication unit can be comprehensively detected, and various potential faults can be discovered in a timely manner. Combined with fault analysis and location as well as fault handling and verification, the type and location of the fault can be quickly and accurately determined, ensuring the effectiveness of fault handling, avoiding the recurrence of faults, and providing strong support for fault handling. BRIEF DESCRIPTION OF THE DRAWINGS

[0147] Figure 1 is a schematic block diagram of an intelligent fault detection system in the present invention;

[0148] Figure 2 It is a schematic block diagram of an intelligent real-time fault monitoring unit in the present invention;

[0149] Figure 3 The figure is a schematic block diagram of the fault analysis and location unit of the present invention. DETAILED DESCRIPTION

[0150] Embodiment 1:

[0151] The present invention relates to an intelligent fault detection method for a dual-mode communication unit, comprising the following steps:

[0152] S1, data collection;

[0153] The working status of the dual-mode communication unit is monitored in real time through sensors, and parameter data including signal strength, communication frequency, temperature, and voltage are collected. The temperature and voltage are directly read by the temperature sensor and the voltage sensor respectively;

[0154] Signal strength can be expressed in dBm, and the calculation formula is:

[0155] P(dBm)=10×log(P(mW));

[0156] Wherein, P (mW) is the power value in milliwatts;

[0157] S2, single point detection;

[0158] Perform separate tests on each key node in the dual-mode communication unit to check whether its working status is normal, such as whether the working temperature and voltage of the chip are within the normal range, and whether the connection of the communication interface is stable;

[0159] S201, chip temperature;

[0160] For chip temperature, the normal temperature range can be set as:

[0161] T min ≤T≤T max ;

[0162] Wherein, T is the temperature value measured in real time;

[0163] T min and T max are the minimum and maximum allowable temperatures determined according to the chip specifications;

[0164] S202, chip voltage;

[0165] For chip voltage, the normal voltage range can be set as:

[0166] V min ≤V≤V max ;

[0167] Wherein, V is the temperature value measured in real time;

[0168] V min and V max are the minimum and maximum allowable temperatures determined according to the chip specifications;

[0169] S3, network detection;

[0170] Test the performance of the dual-mode communication unit in the networking state to check whether the network connection is stable, whether the communication between communication nodes is smooth, and whether the network throughput and delay indicators meet the requirements;

[0171] S301, network throughput;

[0172] Network throughput is expressed as the amount of data transmitted per unit time, such as the number of bits transmitted per second (bps). The calculation formula is:

[0173]

[0174] Among them, DatrVolume is the total amount of data transmitted in a period of time;

[0175] Time is the time interval;

[0176] S302, network delay;

[0177] Network latency is usually expressed in milliseconds (ms) and can be calculated by sending a test packet and measuring the round-trip time using the formula:

[0178]

[0179] Among them, RoundTripTime is the round trip time of the test data packet;

[0180] S4, communication quality detection;

[0181] Detect the communication quality of the dual-mode communication unit, including signal strength, signal-to-noise ratio, and bit error rate. By detecting the communication quality, problems in communication can be discovered in a timely manner;

[0182] Through single-point detection, networking detection and communication quality detection, the performance of the dual-mode communication unit can be comprehensively detected, and various potential faults can be discovered in a timely manner. Combined with fault analysis and location as well as fault handling and verification, the type and location of the fault can be quickly and accurately determined, ensuring the effectiveness of fault handling, avoiding the recurrence of faults, and providing strong support for fault handling.

[0183] S5, anti-interference test;

[0184] Conduct anti-interference tests on the dual-mode communication unit, simulate various interference sources, detect the communication performance of the dual-mode communication unit in an interference environment, use a signal generator to generate interference signals of different frequencies and intensities, and then measure the signal strength, signal-to-noise ratio, and bit error rate indicators of the dual-mode communication unit under interference. The anti-interference performance can be measured by the interference suppression ratio, and the calculation formula is:

[0185]

[0186] Among them, P signal without inter ference is the signal power without interference;

[0187] P signal with inter ference is the signal power when there is interference;

[0188] S6. Data analysis;

[0189] The collected data is analyzed and processed, and machine learning algorithms are used to model and predict the data to determine whether there is a fault in the dual-mode communication unit;

[0190] S7, Fault analysis and location;

[0191] If the prediction result shows that the dual-mode communication unit is faulty, the fault is further analyzed and located;

[0192] S8, fault handling and verification;

[0193] Take appropriate troubleshooting measures according to the type and location of the fault. At the same time, re-test the dual-mode communication unit after the fault is repaired to verify whether the fault has been completely repaired. If the fault still exists, re-diagnose and handle the fault until the fault is completely repaired.

[0194] S9, optimize the communication transmission path;

[0195] According to the results of fault detection and processing, the communication transmission path of the dual-mode communication unit is optimized. If the signal strength of a communication node is weak, the communication path can be adjusted to avoid the node. If interference is found in a communication link, other communication links can be selected for communication.

[0196] The Dijkstra algorithm is used to optimize the communication transmission path, and the optimal communication path is determined by calculating the shortest path from each node to the target node:

[0197] Let the communication network be G = (V, E);

[0198] Among them, V is the node set;

[0199] E is the edge set;

[0200] For each node v∈V, let d(v) be the shortest distance from the source node to v, and p(v) be the predecessor node of v;

[0201] At initialization, d(s) = 0 (where s is the source node), for other nodes v≠s, d(v) = ∞;

[0202] For each node u whose shortest distance is not determined, select the node with the smallest d(u) and update the shortest distance of its adjacent nodes. Repeat the process until the shortest distances of all nodes are determined. Finally, by backtracking the predecessor nodes, the shortest path from the source node to the target node can be obtained.

[0203] The present invention uses a machine learning algorithm to model and predict data, and can quickly and accurately determine whether a dual-mode communication unit has a fault. In conjunction with real-time monitoring of the working status of the dual-mode communication unit through sensors, the efficiency and accuracy of fault detection are improved, and faults can be discovered in a timely manner, and the faults can be diagnosed and processed, which greatly improves the daily detection efficiency of the dual-mode communication unit in harsh environments, saving time and effort.

[0204] In an embodiment of the present invention, the S4 communication quality detection further includes the following steps:

[0205] S401, signal-to-noise ratio;

[0206] The signal-to-noise ratio calculation formula is:

[0207]

[0208] Among them, P signal is the signal power;

[0209] P noise is the noise power;

[0210] S402, bit error rate;

[0211] The bit error rate calculation formula is:

[0212]

[0213] Among them, BER is the ratio of the number of error bits to the total number of transmitted bits;

[0214] In an embodiment of the present invention, S6 data analysis further includes the following steps:

[0215] S601, data preprocessing;

[0216] Clean, denoise and normalize the collected data to improve the quality and usability of the data;

[0217] S602, feature extraction;

[0218] Extract features that can reflect the failure of the dual-mode communication unit from the preprocessed data, such as the rate of change of signal strength, deviation of communication frequency, and abnormal temperature change;

[0219] S603, model training;

[0220] Use historical data to train the machine learning model and build a fault detection model using a support vector machine, whose goal is to find a hyperplane that allows data points of different categories to be separated as correctly as possible while maximizing the interval;

[0221] Assume that the training data set is: {(X1,y1),(X2,y2),...,(X i ,y i )};

[0222] Among them, X i is the input feature vector;

[0223] y i∈{-1,1} is the category label;

[0224] The optimization problem of support vector machine can be expressed as:

[0225] Where ω is the normal vector of the hyperplane;

[0226] b is the intercept;

[0227] C is the penalty parameter;

[0228] ξ i is the slack variable

[0229] The constraints are: i (ω T X i +b)≥1-ξ i ,ξ i ≥0;

[0230] The fault detection model is obtained by solving the above problems;

[0231] S604, fault prediction;

[0232] Input the real-time collected data into the fault detection model to predict whether there is a fault in the dual-mode communication unit;

[0233] In an embodiment of the present invention, S601 data preprocessing further includes the following steps:

[0234] S601a, data cleaning;

[0235] Outliers are removed through data cleaning. The Laida criterion is used. If the difference between the measured value and the mean value is greater than three times the standard deviation, the measured value is considered an outlier and removed:

[0236] Assume the data set is: {X1,X2...,X n};

[0237] The average value is:

[0238] The standard deviation is:

[0239] like Then X i is an outlier;

[0240] S601b, data denoising;

[0241] Data denoising uses methods such as mean filtering, and the new data value is the average of several surrounding data. Suppose the current data point is X i , the surrounding K data points are:

[0242] {Xi-j ,...,X i-j ,X i+j ,...,X i+j};

[0243] in (k is an even number) or (k is an odd number);

[0244] The denoised data value is:

[0245]

[0246] S601c, data normalization;

[0247] Data normalization can use the maximum and minimum normalization method, and the calculation formula is:

[0248]

[0249] Where X is the original data value;

[0250] X min and X max is the minimum and maximum value of the data;

[0251] In an embodiment of the present invention, S602 feature extraction further includes the following steps:

[0252] S602a, signal strength change rate;

[0253] The signal strength change rate calculation formula is:

[0254]

[0255] Among them, P1 is the signal strength at the previous moment;

[0256] P0 is the signal strength at the previous moment;

[0257] S602b, communication frequency deviation;

[0258] The communication frequency deviation calculation formula is:

[0259]

[0260] Wherein, F is the current communication frequency;

[0261] F0 is the standard communication frequency;

[0262] S602c, abnormal temperature change;

[0263] Abnormal temperature changes can be measured by calculating the temperature change rate. Assuming the current temperature is T1, the temperature at the previous moment is T0, and the time interval is Δt, the temperature change rate is:

[0264]

[0265] If the temperature change rate exceeds the threshold R, the temperature is considered abnormal;

[0266] In an embodiment of the present invention, S7 fault analysis and location further includes the following steps:

[0267] S701, fault type determination;

[0268] Determine the type of fault based on its characteristics and manifestations, whether it is a hardware fault, software fault, or communication fault;

[0269] S702, fault location positioning;

[0270] By analyzing the working principle and structure of the dual-mode communication unit and combining the characteristics of the fault, the fault location is located;

[0271] In an embodiment of the present invention, S701 fault type determination further includes the following steps:

[0272] S701a, hardware failure determination;

[0273] Assume that the judgment standard of abnormal temperature rise is that the temperature exceeds the threshold T thershold , that is, T>T thershold It is judged that there may be a hardware failure;

[0274] The voltage instability criterion is that the percentage of voltage deviation from the normal range exceeds the threshold V percentage ,Right now It is judged that there may be a hardware failure;

[0275] S701b, software fault determination;

[0276] The judgment mark of program crash is set as the appearance of a specific error code or exception information;

[0277] S701c, communication failure determination;

[0278] Assume that the criterion for judging the increase of the bit error rate is that the bit error rate exceeds the threshold BER threshold , that is, BER>BER threshold It is judged that there may be a communication failure;

[0279] In an embodiment of the present invention, S702 fault location positioning further includes the following steps:

[0280] S702a, hardware fault location: If it is a hardware fault, the fault location can be determined by detecting the on-off state of the circuit, the working state of the chip, etc.;

[0281] S702b, software fault location: If it is a software fault, the fault location can be determined by analyzing the program's running log, error code, etc.;

[0282] Embodiment 2:

[0283] like Figures 1 to 3 As shown, an intelligent fault detection system for a dual-mode communication unit includes a working status real-time monitoring module for monitoring the working status of the dual-mode communication unit, a key node real-time detection module for detecting key nodes on the dual-mode communication unit, a networking status performance detection module for real-time detection of the networking status performance of the dual-mode communication unit, a communication quality real-time monitoring module for detecting the communication quality of the dual-mode communication unit, and an anti-interference real-time testing module for performing anti-interference testing on the dual-mode communication unit;

[0284] The working status real-time monitoring module, the key node real-time detection module, the networking status performance detection module, the communication quality real-time monitoring module and the anti-interference real-time testing module are electrically connected together with an intelligent fault real-time monitoring unit for real-time fault detection of the dual-mode communication unit in combination with the collected data;

[0285] The intelligent fault real-time monitoring unit includes a data analysis and processing module for analyzing and processing data, and the data analysis and processing module is electrically connected to a machine learning algorithm modeling module for modeling based on a machine learning algorithm and a data preprocessing module for preprocessing the analyzed data;

[0286] The machine learning data modeling module generates a fault detection model, and the fault detection model is electrically connected to a machine learning algorithm prediction module for fault prediction and a machine learning model training module for training the fault detection model, and the machine learning algorithm prediction module and the data preprocessing module are both electrically connected to the machine learning training module;

[0287] The machine learning model training module is electrically connected to a fault analysis and positioning unit for fault analysis and positioning, and the fault analysis and positioning unit is electrically connected to a fault handling and verification module for providing fault handling and verification suggestions, and the fault handling and verification module is electrically connected to a communication transmission path optimization module for optimizing the communication transmission path. The present invention collects the operating data of the dual-mode communication unit in real time through a working status real-time monitoring module, a key node real-time detection module, a networking status performance detection module, a communication quality real-time monitoring module, and an anti-interference real-time testing module, and cooperates with the intelligent fault real-time monitoring unit to perform fault analysis in combination with the collected data, without relying on manual detection, reducing labor costs and time costs, thereby reducing the cost of fault detection.

[0288] As another embodiment of the present invention, the fault analysis and location unit includes a fault type determination module for determining the fault type and a fault location location unit for locating the fault location, and the fault type determination module is electrically connected to the fault location location unit;

[0289] The fault location locating unit includes a hardware fault locating module for locating a hardware fault and a software fault locating module for locating a software fault.

[0290] The embodiments of the present invention disclose preferred embodiments, but are not limited thereto. A person skilled in the art can easily understand the spirit of the present invention based on the above embodiments and make different extensions and changes. However, as long as they do not deviate from the spirit of the present invention, they are all within the protection scope of the present invention.

Claims

1. An intelligent fault detection method for a dual-mode communication unit, characterized in that: The following steps are involved: S1, data collection; The working status of the dual-mode communication unit is monitored in real time through sensors, and parameter data including signal strength, communication frequency, temperature, and voltage are collected. The temperature and voltage are directly read by the temperature sensor and the voltage sensor respectively; Signal strength can be expressed in dBm, and the calculation formula is: P(dBm)=10×log(P(mW)); Wherein, P (mW) is the power value in milliwatts; S2, single point detection; Perform separate tests on each key node in the dual-mode communication unit to check whether its working status is normal, such as whether the working temperature and voltage of the chip are within the normal range, and whether the connection of the communication interface is stable; S201, chip temperature; For chip temperature, the normal temperature range can be set as: T min ≤T≤T max ; Wherein, T is the temperature value measured in real time; T min and T max are the minimum and maximum allowable temperatures determined according to the chip specifications; S202, chip voltage; For chip voltage, the normal voltage range can be set as: In min ≤V≤V max ; Wherein, V is the temperature value measured in real time; V min and V max are the minimum and maximum allowable temperatures determined according to the chip specifications; S3, network detection; Test the performance of the dual-mode communication unit in the networking state to check whether the network connection is stable, whether the communication between communication nodes is smooth, and whether the network throughput and delay indicators meet the requirements; S301, network throughput; Network throughput is expressed as the amount of data transmitted per unit time, such as the number of bits transmitted per second (bps). The calculation formula is: Among them, DatrVolume is the total amount of data transmitted in a period of time; Time is the time interval; S302, network delay; Network latency is usually expressed in milliseconds (ms) and can be calculated by sending a test packet and measuring the round-trip time using the formula: Among them, RoundTripTime is the round trip time of the test data packet; S4, communication quality detection; Detect the communication quality of the dual-mode communication unit, including signal strength, signal-to-noise ratio, and bit error rate. By detecting the communication quality, problems in communication can be discovered in a timely manner; S5, anti-interference test; Conduct anti-interference tests on the dual-mode communication unit, simulate various interference sources, detect the communication performance of the dual-mode communication unit in an interference environment, use a signal generator to generate interference signals of different frequencies and intensities, and then measure the signal strength, signal-to-noise ratio, and bit error rate indicators of the dual-mode communication unit under interference. The anti-interference performance can be measured by the interference suppression ratio, and the calculation formula is: Among them, P signalwith out interference is the signal power without interference; P signalwithinterference is the signal power when there is interference; S6. Data analysis; The collected data is analyzed and processed, and machine learning algorithms are used to model and predict the data to determine whether there is a fault in the dual-mode communication unit; S7, Fault analysis and location; If the prediction result shows that the dual-mode communication unit is faulty, the fault is further analyzed and located; S8, fault handling and verification; Take appropriate troubleshooting measures according to the type and location of the fault. At the same time, re-test the dual-mode communication unit after the fault is repaired to verify whether the fault has been completely repaired. If the fault still exists, re-diagnose and handle the fault until the fault is completely repaired. S9, optimize the communication transmission path; According to the results of fault detection and processing, the communication transmission path of the dual-mode communication unit is optimized. If the signal strength of a communication node is weak, the communication path can be adjusted to avoid the node. If interference is found in a communication link, other communication links can be selected for communication. The Dijkstra algorithm is used to optimize the communication transmission path, and the optimal communication path is determined by calculating the shortest path from each node to the target node: Let the communication network be G = (V, E); Among them, V is the node set; E is the edge set; For each node v∈V, let d(v) be the shortest distance from the source node to v, and p(v) be the predecessor node of v; At initialization, d(s) = 0 (where s is the source node), for other nodes v≠s, d(v) = ∞; For each node u whose shortest distance is not determined, select the node with the smallest d(u) and update the shortest distance of its adjacent nodes. Repeat the process until the shortest distances of all nodes are determined. Finally, by backtracking the predecessor nodes, the shortest path from the source node to the target node can be obtained.

2. The intelligent fault detection method of a dual-mode communication unit according to claim 1, characterized in that: The S4 communication quality detection also includes the following steps: S401, signal-to-noise ratio; The signal-to-noise ratio calculation formula is: Among them, P signal is the signal power; P noise is the noise power; S402, bit error rate; The bit error rate calculation formula is: Among them, BER is the ratio of the number of error bits to the total number of transmitted bits.

3. The intelligent fault detection method of a dual-mode communication unit according to claim 1, characterized in that: The S6 data analysis also includes the following steps: S601, data preprocessing; Clean, denoise and normalize the collected data to improve the quality and usability of the data; S602, feature extraction; Extract features that can reflect the failure of the dual-mode communication unit from the preprocessed data, such as the rate of change of signal strength, deviation of communication frequency, and abnormal temperature change; S603, model training; Use historical data to train the machine learning model and build a fault detection model using a support vector machine, whose goal is to find a hyperplane that allows data points of different categories to be separated as correctly as possible while maximizing the interval; Assume that the training data set is: {(X1,y1),(X2,y2),...,(X i ,y i )}; Among them, X i is the input feature vector; y i ∈{-1,1} is the category label; The optimization problem of support vector machine can be expressed as: Where ω is the normal vector of the hyperplane; b is the intercept; C is the penalty parameter; ξ i is the slack variable The constraints are: i (ω T X i +b)≥1-ξ i ,ξ i ≥0; The fault detection model is obtained by solving the above problems; S604, fault prediction; The real-time collected data is input into the fault detection model to predict whether the dual-mode communication unit has a fault.

4. The intelligent fault detection method of a dual-mode communication unit according to claim 3, characterized in that: The data preprocessing step S601 also includes the following steps: S601a, data cleaning; Outliers are removed through data cleaning. The Laida criterion is used. If the difference between the measured value and the mean value is greater than three times the standard deviation, the measured value is considered an outlier and removed: Assume the data set is: {X1,X2...,X n }; The average value is: The standard deviation is: like Then X i is an outlier; S601b, data denoising; Data denoising uses methods such as mean filtering, and the new data value is the average of several surrounding data. Suppose the current data point is X i , the surrounding K data points are: {X i-j ,...,X i-j ,X i+j ,...,X i+j }; in (k is an even number) or (k is an odd number); The denoised data value is: S601c, data normalization; Data normalization can use the maximum and minimum normalization method, and the calculation formula is: Where X is the original data value; X min and X max are the minimum and maximum values ​​of the data.

5. The intelligent fault detection method of a dual-mode communication unit according to claim 4, characterized in that: The feature extraction step S602 also includes the following steps: S602a, signal strength change rate; The signal strength change rate calculation formula is: Among them, P1 is the signal strength at the previous moment; P0 is the signal strength at the previous moment; S602b, communication frequency deviation; The communication frequency deviation calculation formula is: Wherein, F is the current communication frequency; F0 is the standard communication frequency; S602c, abnormal temperature change; Abnormal temperature changes can be measured by calculating the temperature change rate. Assuming the current temperature is T1, the temperature at the previous moment is T0, and the time interval is Δt, the temperature change rate is: If the temperature change rate exceeds the threshold R, the temperature is considered abnormal.

6. The intelligent fault detection method of a dual-mode communication unit according to claim 1, characterized in that: The S7 fault analysis and location also includes the following steps: S701, fault type determination; Determine the type of fault based on its characteristics and manifestations, whether it is a hardware fault, software fault, or communication fault; S702, fault location positioning; By analyzing the working principle and structure of the dual-mode communication unit and combining the characteristics of the fault, the fault location can be located.

7. The intelligent fault detection method of a dual-mode communication unit according to claim 6, characterized in that: The S701 fault type determination also includes the following steps: S701a, hardware failure determination; Assume that the judgment standard of abnormal temperature rise is that the temperature exceeds the threshold value T thershold , that is, T>T thershold It is judged that there may be a hardware failure; The voltage instability criterion is that the percentage of voltage deviation from the normal range exceeds the threshold V percentage ,Right now It is judged that there may be a hardware failure; S701b, software fault determination; The judgment mark of program crash is set as the appearance of a specific error code or exception information; S701c, communication failure determination; Assume that the criterion for judging the increase of the bit error rate is that the bit error rate exceeds the threshold BER threshold , that is, BER>BER threshold It is judged that there may be a communication failure.

8. The intelligent fault detection method of a dual-mode communication unit according to claim 7, characterized in that: The fault location positioning S702 also includes the following steps: S702a, hardware fault location: If it is a hardware fault, the fault location can be determined by detecting the on-off state of the circuit, the working state of the chip, etc.; S702b, software fault location: If it is a software fault, the fault location can be determined by analyzing the program's running log, error code, etc.

9. An intelligent fault detection system for a dual-mode communication unit, according to the intelligent fault detection method for a dual-mode communication unit according to claims 1 to 8, characterized in that: It includes a working status real-time monitoring module for monitoring the working status of the dual-mode communication unit, a key node real-time detection module for detecting key nodes on the dual-mode communication unit, a networking status performance detection module for real-time detection of the networking status performance of the dual-mode communication unit, a communication quality real-time monitoring module for detecting the communication quality of the dual-mode communication unit, and an anti-interference real-time testing module for performing anti-interference testing on the dual-mode communication unit; The working status real-time monitoring module, the key node real-time detection module, the networking status performance detection module, the communication quality real-time monitoring module and the anti-interference real-time testing module are electrically connected to an intelligent fault real-time monitoring unit for real-time fault detection of the dual-mode communication unit in combination with the collected data; The intelligent fault real-time monitoring unit includes a data analysis and processing module for analyzing and processing data, and the data analysis and processing module is electrically connected to a machine learning algorithm modeling module for modeling based on a machine learning algorithm and a data preprocessing module for preprocessing the analyzed data; The machine learning data modeling module generates a fault detection model, and the fault detection model is electrically connected to a machine learning algorithm prediction module for fault prediction and a machine learning model training module for training the fault detection model, and the machine learning algorithm prediction module and the data preprocessing module are both electrically connected to the machine learning training module; The machine learning model training module is electrically connected to a fault analysis and positioning unit for fault analysis and positioning, and the fault analysis and positioning unit is electrically connected to a fault handling and verification module for providing fault handling and verification suggestions, and the fault handling and verification module is electrically connected to a communication transmission path optimization module for optimizing the communication transmission path.

10. The intelligent fault detection system of a dual-mode communication unit according to claim 9, characterized in that: The fault analysis and positioning unit includes a fault type judgment module for judging the fault type and a fault position positioning unit for positioning the fault position, and the fault type judgment module is electrically connected to the fault position positioning unit; The fault location locating unit includes a hardware fault locating module for locating a hardware fault and a software fault locating module for locating a software fault.

Images