Railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm

Abstract

The invention belongs to the technical field of railway bogie detection, and particularly relates to a railway bogie intelligent fault detection system based on multi-sensor fusion and a YOLOv8 algorithm, which comprises a multi-dimensional data acquisition module, a data preprocessing module, a fusion processing module, a fault identification module and an alarm module. According to the method, the collected multi-dimensional data are summarized and analyzed, so that the states of the railway bogies are monitored in real time and faults are quickly recognized, the railway bogies in the normal state can be quickly screened out, the detection time of the railway bogies is shortened, and the detection efficiency of the abnormal railway bogies is improved. Compared with the prior art, the method has the advantages that the railway bogies can be checked one by one correspondingly, and in addition, fault levels of the railway bogies with faults can be classified, so that the maintenance priorities of the railway bogies detected in the same batch can be determined, the maintenance work of the railway bogies can be carried out orderly, and the safety and efficiency of railway operation are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of railway bogie detection, and in particular relates to a railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm. Background Art

[0002] With the rapid development of railway transportation, railway bogies are key components of railway vehicles, and their safety and reliability are crucial to the entire railway transportation system. Traditional railway bogie fault detection methods mainly rely on manual inspection, which is not only inefficient but also easily affected by human factors, resulting in insufficient accuracy and timeliness of fault detection. Therefore, it is particularly urgent to develop an efficient and accurate railway bogie fault detection system.

[0003] In the prior art, the detection of railway bogies is mostly to identify faults in complex environments. There are various types of railway bogie faults, including but not limited to wheelset wear, loose bolts, foreign object jamming and overheating, which require the detection system to accurately identify and distinguish them. However, when analyzing the data under these fault types one by one from beginning to end, for normal railway bogies, it takes a long time, which is not conducive to improving the detection efficiency. Therefore, the railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm proposed in the present invention aims to solve the above problems. Summary of the invention

[0004] The purpose of the present invention is to provide a railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm, which can effectively improve the accuracy and efficiency of railway bogie fault detection.

[0005] The technical solution adopted by the present invention is as follows:

[0006] Railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm, including multi-dimensional data acquisition module, data preprocessing module, fusion processing module, fault identification module and alarm module;

[0007] The multi-dimensional data acquisition module is used to collect detection data of the railway bogie, and the detection data includes wheelset data, bolt data, foreign matter data and temperature data;

[0008] The data preprocessing module is used to preprocess the collected detection data and record the preprocessed detection data as reference feature data;

[0009] The fusion processing module is used to perform multi-dimensional fusion on the reference feature data to obtain fused feature data, and to perform vectorization processing on the fused feature data to obtain a fused feature vector;

[0010] The fault identification module is used to perform fault identification based on the fused feature vector, and output the detection state of the railway bogie according to the result of the fault identification, wherein the detection state includes a normal state and a fault state;

[0011] In the normal state, the detection state of the railway bogie is recorded in a system log;

[0012] In the fault state, the fault type of the railway bogie is output, and an alarm signal is sent through the alarm module according to the fault type.

[0013] In a preferred embodiment, the multidimensional data acquisition module includes a detection unit and a recording unit. The detection unit is used to collect wheelset data, bolt data, foreign object data and temperature data of the railway bogie. The recording unit is used to summarize and record the collected wheelset data, bolt data, foreign object data and temperature data as detection data, and store them in a database.

[0014] In a preferred embodiment, the data preprocessing module includes a data cleaning unit and a feature extraction unit. The data cleaning unit is used to remove noise and abnormal values ​​in the detection data, and the feature extraction unit is used to extract features related to fault identification from the cleaned detection data and record them as baseline feature data.

[0015] In a preferred solution, the fusion processing module includes a feature fusion unit and a vectorization processing unit. The feature fusion unit is used to perform multi-dimensional fusion of the baseline feature data to generate fused feature data. The vectorization processing unit is used to add feature descriptors to the fused feature data and summarize the feature descriptors to obtain a fused feature vector.

[0016] In a preferred solution, the reference feature data of the same railway bogie will be aggregated into the same feature subset, and will also be stored in different independent feature subsets according to the feature types, and each independent feature subset will only contain one type of reference feature data.

[0017] In a preferred embodiment, the fault identification module includes a fault classification unit and a state output unit. The fault classification unit is used to classify the fault of the railway bogie according to the fused feature vector, and the state output unit is used to output the detection state of the railway bogie, including a normal state and a fault state.

[0018] In a preferred solution, when the state output unit is executed, the fusion feature vector of the railway bogie is collected and compared with a preset standard feature vector, and the specific steps are as follows:

[0019] Inputting the fused feature vector and the standard feature vector together into a preset precondition function, and recording the output result of the precondition function as an evaluation score;

[0020] Comparing the evaluation score with a preset standard threshold;

[0021] When the evaluation score is higher than or equal to the standard threshold, it indicates that the detection data corresponding to the railway bogie is normal, and the detection state of the railway bogie is directly output as a normal state;

[0022] When the evaluation score is lower than the standard threshold, it indicates that there is an abnormality in the detection data corresponding to the railway bogie, and the wheelset data, bolt data, foreign object data and temperature data are compared one by one to determine the abnormal data, and the fault type of the railway bogie is determined based on the abnormal data.

[0023] In a preferred embodiment, the fault identification module further comprises a fault level classification unit, and the fault level classification unit is used to classify the faults of the railway bogie according to the fault category;

[0024] A plurality of fault levels are set in the fault level classification unit, and the fault levels are positively correlated with the number of railway bogie fault categories;

[0025] Among them, the higher the fault level of the railway bogie, the higher the maintenance priority of the corresponding railway bogie.

[0026] The present invention also provides a railway bogie intelligent fault detection method based on multi-sensor fusion and YOLOv8 algorithm, which is applied to the above-mentioned railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm, and includes the following steps:

[0027] Collecting detection data of railway bogies, wherein the detection data includes wheelset data, bolt data, foreign matter data and temperature data;

[0028] Preprocessing the collected detection data, and recording the preprocessed detection data as reference feature data;

[0029] Perform multi-dimensional fusion on the benchmark feature data to obtain fused feature data, and perform vectorization processing on the fused feature data to obtain a fused feature vector;

[0030] Perform fault identification based on the fused feature vector, and output the detection state of the railway bogie according to the result of the fault identification, wherein the detection state includes a normal state and a fault state;

[0031] In the normal state, the detection state of the railway bogie is recorded in a system log;

[0032] In the fault state, the fault type of the railway bogie is output, and an alarm signal is sent through the alarm module according to the fault type.

[0033] And, an electronic device, the electronic device comprising:

[0034] at least one processor;

[0035] and a memory communicatively coupled to the at least one processor;

[0036] In which, the memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the above-mentioned railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm.

[0037] The technical effects achieved by the present invention are:

[0038] The present invention realizes real-time monitoring of the status of railway bogies and rapid identification of faults by summarizing and analyzing the collected multi-dimensional data, so that railway bogies in normal status can be quickly screened out, thereby shortening the inspection time of railway bogies. For abnormal railway bogies, corresponding one-by-one verification work will be carried out. In addition, for faulty railway bogies, their fault levels will be classified, so that the maintenance priority of railway bogies inspected in the same batch can be determined, ensuring that the maintenance work of railway bogies is carried out in an orderly manner, and improving the safety and efficiency of railway operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of the system module of the present invention;

[0040] Figure 2 It is a schematic flow chart of the method of the present invention;

[0041] Figure 3 It is a schematic diagram of the structure of an electronic device of the present invention. DETAILED DESCRIPTION

[0042] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0043] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0044] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure or characteristic that may be included in at least one implementation of the present invention. The term "in a preferred embodiment" that appears in different places in this specification does not refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0045] See also Figure 1 As shown, the present invention provides a railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm, including a multi-dimensional data acquisition module, a data preprocessing module, a fusion processing module, a fault identification module and an alarm module;

[0046] The multi-dimensional data acquisition module is used to collect the inspection data of railway bogies, including wheelset data, bolt data, foreign matter data and temperature data;

[0047] The data preprocessing module is used to preprocess the collected detection data and record the preprocessed detection data as reference feature data;

[0048] The fusion processing module is used to perform multi-dimensional fusion on the reference feature data to obtain fused feature data, and to perform vectorization processing on the fused feature data to obtain a fused feature vector;

[0049] The fault identification module is used to perform fault identification based on the fused feature vector, and output the detection state of the railway bogie according to the result of the fault identification, wherein the detection state includes a normal state and a fault state;

[0050] Under normal conditions, the detection status of the railway bogie is recorded in the system log;

[0051] In the fault state, the fault type of the railway bogie is output, and an alarm signal is sent through the alarm module according to the fault type.

[0052] In the present invention, with the rapid development of the railway transportation industry, the maintenance and inspection requirements for railway bogies are getting higher and higher. In order to improve the inspection efficiency and accuracy of railway bogies, in this embodiment, the fault detection system of the railway bogie includes a multi-dimensional data acquisition module, a data preprocessing module, a fusion processing module, a fault identification module and an alarm module. The main function of the multi-dimensional data acquisition module is to collect various types of inspection data of the railway bogie in real time. The inspection data covers multiple dimensions such as wheelset data, bolt data, foreign matter data and temperature data. Through the inspection data, the system can fully understand the operating status of the bogie and provide a rich information source for subsequent fault detection. The data preprocessing module is responsible for the preliminary processing of the collected inspection data, including data cleaning, format conversion, noise filtering and other steps to ensure the quality and consistency of the inspection data. The preprocessed inspection data will be recorded as benchmark feature data to provide an accurate basis for subsequent analysis and processing. The core task of the fusion processing module is to perform multi-dimensional fusion processing on the benchmark feature data. Through feature fusion, the system can integrate data of different dimensions into a unified feature space, so as to obtain a more comprehensive and accurate feature description, so as to facilitate unified fault identification. The fusion processing module will vectorize the fused feature data to generate a fused feature vector to provide necessary input for the fault identification module. The fault identification module uses the fused feature vector to identify the fault. Through the YOLOv8 algorithm, it can quickly and accurately identify whether the bogie has a fault and output the corresponding detection status. In this embodiment, the detection status is divided into a normal state and a fault state. In the normal state, the system will record the detection status of the railway bogie in the system log. The log data can be used for subsequent analysis and maintenance decisions to help the railway department better manage railway facilities. In the fault state, the system will not only output the fault category of the railway bogie, but also send a corresponding alarm signal through the alarm module. The alarm signal can be a visual alarm, a sound alarm or other forms of alarms to ensure that relevant personnel can understand and handle the fault in a timely manner.

[0053] In a preferred embodiment, the multidimensional data acquisition module includes a detection unit and a recording unit. The detection unit is used to collect wheelset data, bolt data, foreign object data and temperature data of the railway bogie. The recording unit is used to summarize and record the collected wheelset data, bolt data, foreign object data and temperature data as detection data, and store them in a database.

[0054] In this embodiment, the multi-dimensional data acquisition module is mainly composed of a detection unit and a recording unit. The main responsibility of the detection unit is to collect and monitor key data information of the railway bogie. Specifically, it will collect wheelset data, including the size, position and wear of the wheelset, bolt data, covering the tightening state, torque value and whether there is looseness of the bolts, foreign matter data, that is, detecting whether there are foreign objects attached or stuck on the bogie, these foreign objects may affect the safe operation of the railway vehicle, temperature data, monitoring the temperature changes of the bogie and its related components to prevent failures caused by overheating, so that the detection unit can fully grasp the operating status of the railway bogie, and the recording unit is responsible for summarizing and collating the detection data collected by the detection unit, that is, integrating the wheelset data, bolt data, foreign matter data and temperature data into a complete detection data report. The recording unit will also store the summarized detection data in the database to ensure the security and traceability of the detection data. The establishment of the database provides convenience for the long-term preservation and ready access of data, and also facilitates the management and analysis of data, providing strong data support for the maintenance and overhaul of railway bogies.

[0055] In a preferred embodiment, the data preprocessing module includes a data cleaning unit and a feature extraction unit. The data cleaning unit is used to remove noise and abnormal values ​​in the detection data. The feature extraction unit is used to extract features related to fault identification from the cleaned detection data and record them as baseline feature data.

[0056] In this embodiment, the data preprocessing module covers a data cleaning unit and a feature extraction unit. Specifically, the main responsibility of the data cleaning unit is to remove noise and outliers in the detection data to ensure the quality and accuracy of the data, usually including identifying and correcting errors in the data, filling missing values, removing duplicate records, and processing outliers, etc., so as to provide a reliable data basis for subsequent data analysis and feature extraction. The feature extraction unit further extracts key features that are helpful for fault identification from the cleaned detection data based on data cleaning. This process involves in-depth analysis of the data to identify patterns and trends related to the fault. The feature extraction unit will extract representative features from the original data and record them as baseline feature data. The baseline feature data will serve as an important input for subsequent fault diagnosis and prediction models to help the system more accurately identify and predict potential fault problems.

[0057] In a preferred embodiment, the fusion processing module includes a feature fusion unit and a vectorization processing unit. The feature fusion unit is used to perform multi-dimensional fusion of the baseline feature data to generate fused feature data. The vectorization processing unit is used to add feature descriptors to the fused feature data and summarize the feature descriptors to obtain a fused feature vector.

[0058] As mentioned above, the fusion processing module is mainly composed of a feature fusion unit and a vector processing unit. Its function is to perform multi-dimensional fusion processing on the baseline feature data to generate richer and more comprehensive fusion feature data. The vector processing unit further works on the basis of the feature fusion unit. Its main task is to add feature descriptors to the fusion feature data. The feature descriptors can describe in detail various aspects and attributes of the fusion feature data, such as the wheelset size, position, wear condition of the railway bogie, the tightening state, torque value, looseness of the bolts, the type and position of foreign matter, and detailed changes in temperature data. Through the feature descriptors, the vector processing unit can convert the fusion feature data into a structured fusion feature vector, which is convenient for the fault identification module to perform efficient and accurate fault analysis and identification, and provide corresponding data support for subsequent data analysis and application.

[0059] In a preferred embodiment, the reference feature data of the same railway bogie will be aggregated into the same feature subset and will also be stored in different independent feature subsets according to the feature type, and each independent feature subset only contains one type of reference feature data.

[0060] In this embodiment, after the benchmark feature data is output, all benchmark feature data under the same railway bogie will be aggregated into a specific feature subset. At the same time, the benchmark feature data will be stored in different independent feature subsets according to their respective feature types. Each independent feature subset only contains one type of benchmark feature data, ensuring that the data classification is clear and easy to manage. This storage method facilitates centralized processing of data, as well as separate analysis and management of different types of feature data, and facilitates subsequent summary analysis and independent analysis of the benchmark feature data.

[0061] In a preferred embodiment, the fault identification module includes a fault classification unit and a state output unit. The fault classification unit is used to classify the fault of the railway bogie according to the fused feature vector, and the state output unit is used to output the detection state of the railway bogie, including a normal state and a fault state.

[0062] As mentioned above, the fault identification module is composed of a fault classification unit and a state output unit. The fault classification unit uses a fusion feature vector to accurately classify and identify various types of faults that may occur in the railway bogie, and comprehensively analyzes various detection data to determine whether the bogie has corresponding fault problems. The state output unit is responsible for presenting the detection results in a clear manner to ensure that relevant personnel can quickly understand the current state of the railway bogie. Specifically, the state output unit will output two main states: normal state and fault state. When the railway bogie operates normally and no abnormality is detected, the state output unit will display a normal state to indicate that the bogie is in good condition. On the contrary, if any fault or abnormality is detected, the state output unit will display a fault state and provide corresponding fault information for timely maintenance and processing, effectively ensuring the reliability and safety of the railway bogie.

[0063] In a preferred embodiment, when the state output unit is executed, the fusion feature vector of the railway bogie is collected and compared with the preset standard feature vector. The specific steps are as follows:

[0064] The fused feature vector and the standard feature vector are input into a preset precondition function, and the output result of the precondition function is recorded as the evaluation score;

[0065] Compare the assessment score with a preset standard threshold;

[0066] When the evaluation score is higher than or equal to the standard threshold, it indicates that the detection data corresponding to the railway bogie is normal, and the detection state of the railway bogie is directly output as a normal state;

[0067] When the evaluation score is lower than the standard threshold, it indicates that there is an abnormality in the detection data corresponding to the railway bogie, and the wheelset data, bolt data, foreign object data and temperature data are compared one by one to determine the abnormal data, and the fault type of the railway bogie is determined based on the abnormal data.

[0068] In the above, during the execution of the state output unit, the fused feature vector of the railway bogie is first collected to comprehensively reflect the operating state of the railway bogie. Subsequently, the fused feature vector is compared with the preset standard feature vector. Specifically, the fused feature vector and the standard feature vector are input into the preset precondition function together, wherein the expression of the precondition function is: In the formula, R represents the evaluation score, n represents the number of feature points in the fusion feature vector and the standard feature vector, and a i and b iRepresent the fusion feature vector and the standard feature vector respectively. The output result of the precondition function will be recorded as an evaluation score. The evaluation score reflects the degree of difference between the current railway bogie state and the standard state. The evaluation score will then be compared with the preset standard threshold. The standard threshold is a preset value used to determine whether the detection data of the railway bogie is within the normal range. It is based on a large amount of historical data and experience. When the evaluation score is higher than or equal to the standard threshold, it indicates that the detection data corresponding to the railway bogie is in a normal state. At this time, the state output unit will directly output the detection state of the railway bogie as a normal state, which means that the railway bogie is in good condition and no further inspection or maintenance is required. However, when the evaluation score is lower than the standard threshold, it indicates that the railway bogie is in a good condition and no further inspection or maintenance is required. If there is an abnormality in the detection data corresponding to the railway bogie, the status output unit will start a further fault diagnosis process. First, the wheelset data, bolt data, foreign object data and temperature data will be compared one by one. This process involves a detailed analysis of each data item to determine whether there is a fault. By comparing one by one, it can be determined which specific item or items of data are abnormal. Once the abnormal data is determined, the status output unit will determine the fault type of the railway bogie based on these abnormal data. The fault category may include but is not limited to wheelset wear, loose bolts, foreign object interference or abnormal temperature. After determining the fault category, the status output unit will output the corresponding fault information so that maintenance personnel can take targeted maintenance measures to ensure that the railway bogie can return to normal operating status.

[0069] In a preferred embodiment, the fault identification module further includes a fault level classification unit, which is used to classify the faults of the railway bogie according to the fault category;

[0070] Multiple fault levels are set in the fault level classification unit, and the fault level is positively correlated with the number of railway bogie fault categories;

[0071] Among them, the higher the fault level of the railway bogie, the higher the maintenance priority of the corresponding railway bogie.

[0072] In this embodiment, the fault identification module further includes a fault level classification unit. The main function of the fault level classification unit is to perform detailed level classification of the faults occurring in the railway bogie according to different fault categories. A plurality of different fault levels are configured inside the fault level classification unit. There is a positive correlation between the number of fault levels and the number of fault categories that may occur in the railway bogie. Specifically, as the fault level of the railway bogie increases, the corresponding maintenance priority will also increase accordingly. For faults with higher fault levels, the railway maintenance department will give priority to arranging resources and manpower to handle them, so as to ensure the normal operation of the railway bogie and the safety of railway transportation.

[0073] like Figure 2 As shown, the railway bogie intelligent fault detection method based on multi-sensor fusion and YOLOv8 algorithm is applied to the above-mentioned railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm, including the following steps:

[0074] Collect inspection data of railway bogies, including wheelset data, bolt data, foreign matter data and temperature data;

[0075] Preprocessing the collected detection data, and recording the preprocessed detection data as reference feature data;

[0076] Perform multi-dimensional fusion on the benchmark feature data to obtain fused feature data, and perform vectorization processing on the fused feature data to obtain a fused feature vector;

[0077] Fault identification is performed based on the fused feature vector, and the detection state of the railway bogie is output according to the result of the fault identification, wherein the detection state includes a normal state and a fault state;

[0078] Under normal conditions, the detection status of the railway bogie is recorded in the system log;

[0079] In the fault state, the fault type of the railway bogie is output, and an alarm signal is sent through the alarm module according to the fault type.

[0080] In the above steps, various types of inspection data of railway bogies will be collected first. The inspection data covers the running status data of wheelsets, the tightening status data of bolts, the data of possible foreign objects and the temperature data of bogies. The collection of inspection data is realized through a variety of sensors, including but not limited to temperature sensors, vibration sensors, image acquisition equipment, etc. Next, the system will preprocess the collected raw inspection data to ensure the quality and accuracy of the data. The preprocessing steps may include data cleaning, filtering, normalization and other operations. After the preprocessing is completed, the obtained inspection data will be recorded as benchmark feature data, and then the benchmark feature data will be fused in multiple dimensions. This process involves integrating the data collected by different sensors to obtain more accurate data. Comprehensive feature information, the fused baseline feature data will be further vectorized to form a fused feature vector. Based on the fused feature vector, fault identification will be performed. This process involves using the YOLOv8 algorithm to analyze the feature vector. According to the result of fault identification, the detection status of the railway bogie will be output. The state is divided into normal state and fault state. In the normal state, the detection status of the railway bogie will be recorded in the system log for subsequent query and analysis. In the fault state, the fault category of the railway bogie will be output, and the corresponding alarm signal will be issued through the alarm module according to the fault category. The alarm signal can be a visual alarm, a sound alarm or other forms of alarm to ensure that relevant personnel can understand and handle the fault in time.

[0081] like Figure 3 As shown, an electronic device, the electronic device includes:

[0082] at least one processor;

[0083] and a memory communicatively coupled to the at least one processor;

[0084] Among them, the memory stores a computer program that can be executed by at least one processor, and the computer program is executed by at least one processor so that the at least one processor can execute the above-mentioned railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm.

[0085] The processor of the above electronic device can be a general-purpose central processing unit (CPU) or a specially designed integrated circuit (IC), such as a field programmable gate array (FPGA) or an application specific integrated circuit (ASIC) to improve processing speed and efficiency. The memory can be a random access memory (RAM), a read-only memory (ROM), a flash memory or other types of non-volatile memory for storing program code and temporary data. The electronic device also includes an arithmetic unit and input-output devices. The arithmetic unit is responsible for performing various computing tasks, while the input-output devices allow users to interact with the electronic device, such as inputting instructions through a keyboard, a mouse, a touch screen, and outputting information through a display, a printer, etc.

[0086] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0087] The above is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications should also be considered as the protection scope of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention shall be implemented according to the conventional means in the art unless otherwise specified and limited.

Claims

1. Railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm, characterized by: It includes multi-dimensional data acquisition module, data preprocessing module, fusion processing module, fault identification module and alarm module; The multi-dimensional data acquisition module is used to collect detection data of the railway bogie, and the detection data includes wheelset data, bolt data, foreign matter data and temperature data; The data preprocessing module is used to preprocess the collected detection data and record the preprocessed detection data as reference feature data; The fusion processing module is used to perform multi-dimensional fusion on the reference feature data to obtain fused feature data, and to perform vectorization processing on the fused feature data to obtain a fused feature vector; The fault identification module is used to perform fault identification based on the fused feature vector, and output the detection state of the railway bogie according to the result of the fault identification, wherein the detection state includes a normal state and a fault state; In the normal state, the detection state of the railway bogie is recorded in a system log; In the fault state, the fault type of the railway bogie is output, and an alarm signal is sent through the alarm module according to the fault type.

2. The railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm according to claim 1 is characterized in that: The multi-dimensional data acquisition module includes a detection unit and a recording unit. The detection unit is used to collect wheelset data, bolt data, foreign object data and temperature data of the railway bogie. The recording unit is used to summarize and record the collected wheelset data, bolt data, foreign object data and temperature data as detection data, and store them in a database.

3. The railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm according to claim 1, characterized in that: The data preprocessing module includes a data cleaning unit and a feature extraction unit. The data cleaning unit is used to remove noise and abnormal values ​​in the detection data. The feature extraction unit is used to extract features related to fault identification from the cleaned detection data and record them as reference feature data.

4. The railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm according to claim 1, characterized in that: The fusion processing module includes a feature fusion unit and a vectorization processing unit. The feature fusion unit is used to perform multi-dimensional fusion of the baseline feature data to generate fused feature data. The vectorization processing unit is used to add feature descriptors to the fused feature data and summarize the feature descriptors to obtain a fused feature vector.

5. The railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm according to claim 1, characterized in that: The reference feature data of the same railway bogie will be aggregated into the same feature subset, and will also be stored in different independent feature subsets according to the feature types, and each independent feature subset contains only one type of reference feature data.

6. The railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm according to claim 1, characterized in that: The fault identification module includes a fault classification unit and a state output unit. The fault classification unit is used to classify the fault of the railway bogie according to the fused feature vector, and the state output unit is used to output the detection state of the railway bogie, including a normal state and a fault state.

7. The railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm according to claim 1 is characterized in that: When the state output unit is executed, the fusion feature vector of the railway bogie is collected and compared with the preset standard feature vector. The specific steps are as follows: Inputting the fused feature vector and the standard feature vector together into a preset precondition function, and recording the output result of the precondition function as an evaluation score; Comparing the evaluation score with a preset standard threshold; When the evaluation score is higher than or equal to the standard threshold, it indicates that the detection data corresponding to the railway bogie is normal, and the detection state of the railway bogie is directly output as a normal state; When the evaluation score is lower than the standard threshold, it indicates that there is an abnormality in the detection data corresponding to the railway bogie, and the wheelset data, bolt data, foreign object data and temperature data are compared one by one to determine the abnormal data, and the fault type of the railway bogie is determined based on the abnormal data.

8. The railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm according to claim 1 is characterized in that: The fault identification module further includes a fault level classification unit, which is used to classify the faults of the railway bogie according to the fault category; A plurality of fault levels are set in the fault level classification unit, and the fault levels are positively correlated with the number of railway bogie fault categories; Among them, the higher the fault level of the railway bogie, the higher the maintenance priority of the corresponding railway bogie.

9. Railway bogie intelligent fault detection method based on multi-sensor fusion and YOLOv8 algorithm, characterized by: The railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm applied to any one of claims 1 to 8 comprises the following steps: Collecting detection data of railway bogies, wherein the detection data includes wheelset data, bolt data, foreign matter data and temperature data; Preprocessing the collected detection data, and recording the preprocessed detection data as reference feature data; Perform multi-dimensional fusion on the benchmark feature data to obtain fused feature data, and perform vectorization processing on the fused feature data to obtain a fused feature vector; Perform fault identification based on the fused feature vector, and output the detection state of the railway bogie according to the result of the fault identification, wherein the detection state includes a normal state and a fault state; In the normal state, the detection state of the railway bogie is recorded in a system log; In the fault state, the fault type of the railway bogie is output, and an alarm signal is sent through the alarm module according to the fault type.

10. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively coupled to the at least one processor; Wherein, the memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the railway bogie intelligent fault detection system based on multi-sensor fusion and YOLOv8 algorithm as described in any one of claims 1 to 8.

Images