Automatic barrier gate fault identification system and method

Abstract

The invention relates to the field of barrier gate fault identification, and particularly discloses an automatic identification system and method for barrier gate faults. The method comprises the following steps: firstly, acquiring a barrier gate spring monitoring image acquired by a camera, motor power consumption data acquired by a motor power statistical device at a plurality of preset time points and spring load data values acquired by a pressure sensor at a plurality of preset time points, and then performing feature extraction and correlation analysis on the three parts by using a deep learning technology to obtain a barrier gate spring monitoring image; and finally, whether a barrier gate spring fault early warning alarm is sent out is judged through a classifier, so that potential faults are found in advance, the downtime of equipment is shortened, the situation that sudden faults affect operation is avoided, and the operation efficiency and reliability of a barrier gate are improved.

Description

Technical Field

[0001] The present application relates to the field of barrier gate fault identification, and more specifically, to an automatic barrier gate fault identification system and method. Background Art

[0002] The barrier gate is a kind of access management equipment widely used in residential areas, parking lots, highway toll stations, hospitals and other places. Its main function is to manage the passage of vehicles by controlling the lifting and lowering of the barrier gate pole when motor vehicles enter and exit. When the barrier gate pole is raised, motor vehicles are allowed to pass; when the barrier gate pole falls, motor vehicles are prevented from passing.

[0003] To ensure that the barrier can operate stably over a long period of time, especially after millions of opening and closing operations, a barrier spring group is usually installed inside the device. These springs are used to offset the weight of the barrier pole, thereby reducing the power demand of the drive motor and maintaining the structural stability of the barrier. However, the service life of the barrier spring is usually only a few hundred thousand times, while in actual applications, the barrier is required to support millions or even tens of millions of opening and closing operations. This makes spring breakage a common and serious problem. Spring breakage can cause the barrier equipment to malfunction, especially when the barrier pole is raised, and the motor needs to output more power to overcome the effect of the missing spring. In this way, manual intervention is usually required to dismantle and repair the barrier. This maintenance method has significant hysteresis and low manual lifting efficiency, which seriously affects operational efficiency.

[0004] Therefore, a system and method for automatically identifying barrier faults are desired. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an automatic identification system and method for barrier gate faults, which first obtains a barrier gate spring monitoring image collected by a camera, motor power consumption data at multiple predetermined time points collected by a motor power statistics device, and spring load data values ​​at multiple predetermined time points collected by a pressure sensor, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally uses a classifier to determine whether to issue a barrier gate spring fault warning alarm, thereby discovering potential faults in advance, reducing equipment downtime, avoiding sudden faults affecting operations, and improving the operating efficiency and reliability of the barrier gate.

[0006] According to one aspect of the present application, there is provided a system for automatically identifying a gate fault, comprising:

[0007] A barrier fault identification data acquisition module is used to acquire a barrier spring monitoring image acquired by a camera, motor power consumption data at multiple predetermined time points acquired by a motor power statistics device, and spring load data values ​​at multiple predetermined time points acquired by a pressure sensor;

[0008] A barrier spring fault identification data processing module, used for extracting a barrier spring monitoring attention feature vector and a barrier spring state association feature vector from the barrier spring monitoring image collected by the camera, the motor power consumption data at multiple predetermined time points collected by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points collected by the pressure sensor;

[0009] The barrier spring fault warning judgment module is used to judge whether to issue a barrier spring fault warning alarm based on the barrier spring monitoring attention feature vector and the barrier spring state association feature vector.

[0010] According to another aspect of the present application, a method for automatically identifying a gate fault is provided, comprising:

[0011] Acquire a gate spring monitoring image collected by a camera, motor power consumption data at multiple predetermined time points collected by a motor power statistics device, and spring load data values ​​at multiple predetermined time points collected by a pressure sensor;

[0012] Extracting a barrier spring monitoring attention feature vector and a barrier spring state association feature vector from the barrier spring monitoring image collected by the camera, the motor power consumption data at multiple predetermined time points collected by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points collected by the pressure sensor;

[0013] Based on the barrier spring monitoring attention feature vector and the barrier spring state association feature vector, it is determined whether to issue a barrier spring failure warning alarm.

[0014] Compared with the prior art, the present application provides an automatic identification system and method for barrier gate faults, which first obtains a barrier gate spring monitoring image captured by a camera, motor power consumption data at multiple predetermined time points captured by a motor power statistics device, and spring load data values ​​at multiple predetermined time points captured by a pressure sensor, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally uses a classifier to determine whether to issue a barrier gate spring fault warning alarm, thereby discovering potential faults in advance, reducing equipment downtime, avoiding sudden faults affecting operations, and improving the operating efficiency and reliability of the barrier gate. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1 It is a block diagram of a system for automatically identifying barrier gate failures according to an embodiment of the present application.

[0017] Figure 2 It is a block diagram of a barrier gate fault identification data processing module in the barrier gate fault automatic identification system according to an embodiment of the present application.

[0018] Figure 3 It is a block diagram of a barrier spring monitoring feature extraction unit in a barrier fault automatic identification system according to an embodiment of the present application.

[0019] Figure 4 It is a block diagram of a barrier spring fault warning judgment module in the barrier spring fault automatic identification system according to an embodiment of the present application.

[0020] Figure 5 The present invention is a flowchart of a method for automatically identifying a barrier gate fault according to an embodiment of the present application.

[0021] Figure 6 is a block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0023] Figure 1 FIG. 1 is a block diagram of an automatic identification system for a gate failure according to an embodiment of the present application. Figure 1As shown, the automatic identification system 100 of the barrier gate fault according to the embodiment of the present application includes: a barrier gate fault identification data acquisition module 110, which is used to obtain the barrier gate spring monitoring image collected by the camera, the motor power consumption data at multiple predetermined time points collected by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points collected by the pressure sensor; a barrier gate fault identification data processing module 120, which is used to extract the barrier gate spring monitoring attention feature vector and the barrier gate spring state association feature vector from the barrier gate spring monitoring image collected by the camera, the motor power consumption data at multiple predetermined time points collected by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points collected by the pressure sensor; a barrier spring fault early warning judgment module 130, which is used to judge whether to issue a barrier spring fault early warning alarm based on the barrier spring monitoring attention feature vector and the barrier spring state association feature vector.

[0024] In the automatic identification system 100 of the above-mentioned barrier fault, the barrier fault identification data acquisition module 110 is used to obtain the barrier spring monitoring image collected by the camera, the motor power consumption data at multiple predetermined time points collected by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points collected by the pressure sensor. It should be understood that the barrier is a traffic management device widely used in places such as residential areas, parking lots, highway toll stations and hospitals. It manages vehicle traffic by controlling the lifting and lowering of the barrier rod. When the barrier rod is raised, vehicles are allowed to pass; when the barrier rod is lowered, vehicles are blocked from passing. In order to ensure the long-term stable operation of the barrier, especially after millions of switch operations, a spring group is usually equipped. The function of these springs is to offset the weight of the barrier rod, thereby reducing the burden on the drive motor and maintaining the structural stability of the barrier. However, the service life of the barrier spring is usually only hundreds of thousands of times, while the barrier needs to support millions or even tens of millions of switch operations in actual use. This makes spring breakage a common and serious problem. When the spring breaks, the barrier device may fail, especially when the barrier pole rises, the motor needs to expend more power to compensate for the impact of the missing spring. This usually requires manual intervention to disassemble and repair the barrier. This maintenance method not only has a delayed response, but also has low efficiency, which significantly affects the operating efficiency of the equipment. Therefore, in the technical solution of the present application, by acquiring the barrier spring monitoring image captured by the camera, the motor power consumption data at multiple predetermined time points collected by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points collected by the pressure sensor, and combining with deep learning technology, it is determined whether to issue a barrier spring failure warning alarm, so as to identify potential problems early, reduce equipment downtime, avoid interference with operation caused by sudden failures, and improve the overall efficiency and reliability of the barrier.

[0025] Specifically, the camera image provides real-time visual feedback of the spring status, which helps to detect physical damage or wear in time to avoid sudden failures. The motor power data reflects the load of the drive motor at different points in time. The data from the pressure sensor provides the load of the spring in actual work, helping to detect the actual force of the spring and the compliance with the design specifications. The comprehensive analysis of these data not only helps to detect and repair equipment problems in time, but also predicts the maintenance needs of the equipment, reduces downtime, and improves the overall operation efficiency of the equipment.

[0026] In the automatic identification system 100 of the barrier fault, the barrier fault identification data processing module 120 is used to extract the barrier spring monitoring attention feature vector and the barrier spring state association feature vector from the barrier spring monitoring image collected by the camera, the motor power consumption data at multiple predetermined time points collected by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points collected by the pressure sensor. It should be understood that the extraction of these feature vectors is to achieve multi-dimensional and comprehensive monitoring and analysis of the barrier equipment. They can not only improve the accuracy of fault detection, but also optimize the operating efficiency and maintenance strategy of the equipment to ensure the stability and reliability of the equipment.

[0027] Figure 2 FIG. 1 is a block diagram of a barrier gate fault identification data processing module in a barrier gate fault automatic identification system according to an embodiment of the present application. Figure 2 As shown, in a specific embodiment of the present application, the barrier fault identification data processing module 120 includes: a barrier spring monitoring feature extraction unit 121, which is used to perform feature extraction on the barrier spring monitoring image collected by the camera to obtain a barrier spring monitoring attention feature vector; a motor power consumption feature extraction unit 122, which is used to perform feature extraction on the motor power consumption data at multiple predetermined time points collected by the motor power statistics device to obtain a motor power consumption feature vector; a spring load data feature extraction unit 123, which is used to perform feature extraction on the spring load data values ​​at multiple predetermined time points collected by the pressure sensor to obtain a spring load data feature vector; a barrier spring state feature association unit 124, which is used to associate the motor power consumption feature vector and the spring load data feature vector to obtain the barrier spring state association feature vector.

[0028] It should be understood that camera monitoring images can provide real-time visual information of the spring, but these images usually contain a lot of redundant data. Through feature extraction, these complex image data can be converted into more concise and discriminative feature vectors, focusing on key information related to spring health. For example, these feature vectors can include information such as the appearance, surface wear, and deformation of the spring. This conversion process not only improves the efficiency of data processing, but also reduces the demand for computing resources, making real-time monitoring and data analysis more efficient and accurate.

[0029] Furthermore, the motor power consumption data usually shows complex fluctuations in time series, which may be related to various factors such as the motor load, operating status, environmental factors, etc. Through feature extraction, these time series data can be converted into concise feature vectors, which can summarize the main characteristics and trends of the motor power consumption. For example, statistics such as the average power, peak power, and power fluctuation range of the motor can be extracted. These features provide an overall overview of the motor operation and help identify the difference between normal operation and abnormal behavior.

[0030] Furthermore, the spring load data may show complex fluctuation patterns at different time points, which are related to the actual working state of the spring, load changes, spring aging, etc. Through feature extraction, these data can be converted into representative feature vectors, such as average load, load fluctuation amplitude, maximum and minimum load, etc. These feature vectors can summarize the main trends and characteristics of the spring load, making data analysis more concise and efficient.

[0031] In particular, correlating the motor power consumption feature vector and the spring load data feature vector to obtain the barrier spring state correlation feature vector is to comprehensively consider the working state of the motor and the spring to achieve a more comprehensive understanding of the overall performance of the equipment. By correlating these two types of features, the relationship between them can be revealed, thereby more accurately evaluating the operating status of the barrier equipment.

[0032] Figure 3 FIG. 4 is a block diagram of a barrier spring monitoring feature extraction unit in a barrier fault automatic identification system according to an embodiment of the present application. Figure 3As shown, in a specific embodiment of the present application, the barrier spring monitoring feature extraction unit 121 includes: a barrier spring monitoring target detection subunit 1211, which is used to pass the barrier spring monitoring image collected by the camera through a barrier spring monitoring target detection network model to obtain a barrier spring region of interest; a barrier spring monitoring spatial attention encoding subunit 1212, which is used to pass the barrier spring region of interest through a barrier spring monitoring spatial attention feature encoder to obtain a barrier spring monitoring attention feature map; a barrier spring monitoring attention pooling subunit 1213, which is used to perform global mean pooling on the barrier spring monitoring attention feature map to obtain a barrier spring monitoring attention feature vector.

[0033] It should be understood that in actual applications, the images captured by the camera usually include a panoramic view of the barrier device, which contains a large amount of background information and other irrelevant objects. The barrier spring may be just a small area in the image. Through the target detection network model, the region of interest (ROI) of the spring, that is, the location and range of the spring, can be automatically identified from these images. The target detection network model uses a deep learning algorithm to learn and identify the characteristics and position of the spring, and automatically annotate the bounding box, so as to achieve precise positioning of the spring area. Determining the region of interest can also greatly reduce the complexity of subsequent data processing. Specifically, the barrier spring monitoring target detection network model is used to process the barrier spring monitoring image captured by the camera with the following target detection formula to obtain the barrier spring region of interest;

[0034] Among them, the target detection formula is:

[0035] ROI=(cls(ψ det ,B),Regr(ψ det ,B))

[0036] Among them, ψ det is the monitoring image of the gate spring collected by the camera, B is the anchor frame, ROI is the region of interest of the gate spring, cls(ψ det ,B) indicates classification, Regr(ψ det ,B) indicates regression.

[0037] Furthermore, the barrier spring region of interest is passed through the barrier spring monitoring spatial attention feature encoder, which optimizes feature extraction, allowing the system to more effectively focus on the important areas in the spring image and ignore irrelevant background information, thereby improving the accuracy and efficiency of detection. Among them, the barrier spring monitoring spatial attention feature encoder is a model that uses deep learning technology, which can dynamically adjust the weight of the focus area in the image. By encoding the region of interest with spatial attention, the system can generate an attention feature map that emphasizes the key features of the spring area and reduces interference with other irrelevant areas. The core of this mechanism is to focus more computing resources and analysis capabilities on the most critical parts through the "attention" mechanism, thereby improving the model's perception of the spring state. Specifically, the convolutional coding part of the barrier spring monitoring spatial attention feature encoder is used to perform deep convolutional coding on the barrier spring region of interest to obtain an initial convolutional feature map; the initial convolutional feature map is input into the spatial attention part of the barrier spring monitoring spatial attention feature encoder to obtain a spatial attention map; the spatial attention map is activated by a Softmax function to obtain a spatial attention feature map; and the spatial attention feature map and the initial convolutional feature map are multiplied by the position points to obtain the barrier spring monitoring attention feature map. More specifically, the initial convolutional feature map is input into the spatial attention part of the barrier spring monitoring spatial attention feature encoder to obtain a spatial attention map, including: performing average pooling and maximum pooling along the channel dimension on the initial convolutional feature map to obtain an average feature matrix and a maximum feature matrix; cascading and channel-adjusting the average feature matrix and the maximum feature matrix to obtain a channel feature matrix; and using the convolutional layer of the spatial attention feature map to perform convolutional coding on the channel feature matrix to obtain a spatial attention map.

[0038] Furthermore, global mean pooling can provide a more comprehensive summary of the overall characteristics of the spring state. Through averaging, the feature vector can reflect the global information in the feature map, not just the local details. This method can help the system extract representative features from the entire spring monitoring image, so that subsequent analysis can focus on the overall health of the spring, rather than relying solely on local anomalies in the image.

[0039] In a specific embodiment of the present application, the motor power consumption feature extraction unit 122 includes: constructing the motor power consumption data at multiple predetermined time points collected by the motor power statistics device into a motor power consumption input vector; passing the motor power consumption input vector through a motor power consumption timing feature extractor to obtain the motor power consumption feature vector.

[0040] It should be understood that motor power consumption usually changes over time, and these changes reflect the operating status and load conditions of the motor. By organizing the data at multiple time points into vectors, the time series data can be converted into a format suitable for processing by machine learning models, allowing the model to identify and learn the dynamic patterns of motor power consumption. This input form of time series data helps capture trends, periodic changes, and potential abnormal behaviors of motor power.

[0041] Furthermore, the time series feature extractor is able to compress high-dimensional time series data into a smaller and more representative feature vector. The original time series data may contain a large amount of noise and redundant information. Directly processing this data may increase the computational complexity and be easily affected by noise. By converting the data into a feature vector through a feature extractor, unnecessary noise and redundant information can be removed, key information can be retained, and subsequent data processing and analysis can be simplified. This simplification helps to improve the efficiency and accuracy of the analysis while reducing the consumption of computing resources. Specifically, the fully connected layer of the motor power consumption time series feature extractor is used to fully connect the motor power consumption input vector to extract the high-dimensional implicit features of the eigenvalues ​​at each position in the motor power consumption input vector; and the one-dimensional convolutional layer of the motor power consumption time series feature extractor is used to one-dimensionally encode the motor power consumption input vector to extract the high-dimensional implicit correlation features of the correlation between the eigenvalues ​​at each position in the motor power consumption input vector.

[0042] In a specific embodiment of the present application, the spring load data feature extraction unit 123 includes: constructing the spring load data values ​​at multiple predetermined time points collected by the pressure sensor into a spring load data input vector; passing the spring load data input vector through a spring load data sequence encoder to obtain the spring load data feature vector.

[0043] It should be understood that the load data provided by the pressure sensor usually changes over time, and these changes reflect the dynamic load of the spring. Organizing the data values ​​at multiple time points into an input vector preserves the time series information, allowing the analysis and model to learn the laws and patterns of load changes. This structured data form can better capture the trends, periodic changes and sudden events of the load, which is conducive to in-depth analysis and prediction.

[0044] Furthermore, the original load data input vector contains load measurements at multiple time points, which may be complex and noisy. Sequence encoders, such as recurrent neural networks (RNNs) or long short-term memory networks (LSTMs), can capture long-term dependencies and dynamic change patterns in time series. By processing the data through the encoder, the most representative features of the spring load state can be extracted, which often reflect the overall operating trends and potential problems of the spring. The sequence encoder compresses high-dimensional time series data into a low-dimensional feature vector, which can significantly reduce the complexity of the data. Specifically, the spring-loaded data sequence encoder is used to: use each layer of the spring-loaded data sequence encoder to perform the following on the input data in the forward pass of the layer: perform convolution processing based on a one-dimensional convolution kernel on the input data to obtain a convolution feature map; perform pooling processing based on a feature matrix on the convolution feature map to obtain a pooled feature map; perform nonlinear activation on the pooled feature map to obtain an activation feature map; wherein the input of the first layer of the spring-loaded data sequence encoder is the spring-loaded data input vector, and the output of the last layer of the spring-loaded data sequence encoder is the spring-loaded data feature vector.

[0045] In a specific embodiment of the present application, the barrier spring state feature association unit 124 includes: using SpringInitializr to create a new SpringBoot project, and adding SpringWeb, SpringCloud, SpringBootDevTools and Lombok dependencies; configuring necessary properties in the project's application.properties file, including the application name, server port and Eureka service registration and discovery URL; creating a controller class FeatureVectorController, defining an HTTP POST request processing method in the class, receiving a request body containing a motor power consumption feature vector and a spring load data feature vector, and calling a service class method for calculation; enabling the service discovery client function of the SpringBoot application and SpringCloud in the main application class FeatureVectorApplication; using the Maven command to package the project into a JAR file, and running the generated JAR file through the command to deploy and verify the feature vector calculation service.

[0046] Specifically, some codes of the application.properties configuration file are as follows.

[0047]

[0048]

[0049] Among them, the controller code of FeatureVectorController.java is as follows.

[0050]

[0051]

[0052] Among them, part of the code of the FeatureVectorApplication.java main application class is shown below.

[0053]

[0054]

[0055] Finally, navigate to the project folder in your terminal and compile and run the code with the following commands:

[0056] mvncleanpackage

[0057] java-jartarget / your-project-name.jar

[0058] Among them, SpringWeb provides powerful functions for building Web applications and supports RESTful API development; SpringCloud is a set of tools for building distributed systems, providing service discovery, configuration management and other functions; SpringBootDevTools is a set of development tools designed to improve development efficiency, including functions such as automatic restart and LiveReload; Lombok is a Java library that simplifies code through annotations and reduces the writing of boilerplate code, such as automatically generating getter and setter methods. Together, these dependencies support fast and efficient SpringBoot application development, especially when building a microservice architecture.

[0059] The application.properties configuration file defines the basic information of the application, including the application name, server port, and URL for Eureka service registration and discovery. These configurations ensure that the microservice can be correctly registered with the service registry and discovered and called by other services. The FeatureVectorService.java service class is responsible for handling the calculation logic of the feature vector. It contains two main methods: transpose is used to transpose the spring load data feature vector, and multiply is used to calculate the dot product of two vectors. These two methods together implement the product calculation of the feature vector. The FeatureVectorController.java controller class defines an HTTP POST request processing method that receives a request body containing the motor power consumption feature vector and the spring load data feature vector. This method calls the calculateProduct method in FeatureVectorService for calculation and returns the result to the client. This design enables the calculation of the feature vector to be called through the RESTful API, which is convenient for integration with other systems. The FeatureVectorApplication.java main application class enables the service discovery client function of SpringBoot applications and SpringCloud. Through the @EnableDiscoveryClient annotation, the application can automatically register with the Eureka service registry to achieve dynamic discovery and management of services. After deploying and running the microservice, you can send JSON data containing feature vectors to the / feature-vector / calculate endpoint via HTTP POST request. The microservice will receive this data, call the internal logic to calculate the product of the feature vectors, and return the result in JSON format. This design makes the feature vector calculation process highly modular and extensible, which is easy to integrate and use in complex distributed systems. Through the powerful functions of SpringCloud, the microservice can not only efficiently handle the calculation of feature vectors, but also work with other microservices to jointly build a resilient and scalable distributed system. In this way, SpringBoot's automatic configuration and simplified development process allow developers to focus on the implementation of business logic, greatly improving development efficiency and code quality.

[0060] In the automatic identification system 100 of the barrier fault, the barrier spring fault early warning judgment module 130 is used to determine whether to issue a barrier spring fault early warning alarm based on the barrier spring monitoring attention feature vector and the barrier spring state association feature vector. It should be understood that the barrier spring monitoring attention feature vector is usually composed of key features such as dynamic load changes, stress and fatigue generated by the spring during operation. These features may include the amplitude and frequency of load fluctuations and other dynamic behaviors related to the spring state. By analyzing these feature vectors, the normal and abnormal modes of the spring under different working conditions can be identified, thereby indicating that the spring has potential faults.

[0061] Figure 4 FIG. 1 is a block diagram of a barrier spring fault early warning judgment module in a barrier spring fault automatic identification system according to an embodiment of the present application. Figure 4 As shown, in a specific embodiment of the present application, the barrier spring fault warning judgment module 130 includes: a barrier spring feature fusion unit 131, which is used to perform feature fine-grained dimension correction on the barrier spring monitoring attention feature vector and the barrier spring state association feature vector based on back-propagation displacement compensation to obtain a barrier spring fault judgment classification feature vector; a barrier spring fault warning alarm judgment classification unit 132, which is used to pass the barrier spring fault judgment classification feature vector through a classifier to obtain a classification result, and the classification result is used to determine whether to issue a barrier spring fault warning alarm.

[0062] It should be understood that by fusing the barrier spring monitoring attention feature vector and the barrier spring state association feature vector, a comprehensive barrier spring fault judgment classification feature vector can be obtained. This fusion method integrates multiple aspects of information on spring dynamic behavior and system impact, thereby improving the accuracy and classification ability of fault detection and providing a more reliable basis for maintenance and management.

[0063] In particular, in the technical solution of the present application, the barrier spring monitoring attention feature vector comes from image data processing, and features are extracted through target detection and spatial attention mechanism. These features are mainly focused on the description of visual information and spatial position, and often contain rich spatial structure information and detail features. The barrier spring state association feature vector is generated by motor power and spring load data. These data are extracted and sequence encoded through time series features, and mainly focus on dynamic changes and correlations in time series, emphasizing the patterns and trends that change over time, which are significantly different from the spatial features in image data. This difference leads to a shift in fine-grained dimensions when the two features are simply degraded. This shift in the fine-grained dimension of the feature may cause the local structure collapse of the fused barrier spring fault judgment classification feature vector, that is, in the feature space after feature fusion, the structural information of some areas is destroyed, causing the classifier to lose local consistency when processing these features, thereby affecting the final classification accuracy. Therefore, in the technical solution of the present application, the barrier spring monitoring attention feature vector and the barrier spring state association feature vector are corrected based on the fine-grained dimension of the feature based on back-propagation displacement compensation to obtain the barrier spring fault judgment classification feature vector.

[0064] Among them, the barrier spring monitoring attention feature vector and the barrier spring state-related feature vector are subjected to feature fine-grained dimension correction based on back-propagation displacement compensation to obtain a barrier spring fault judgment classification feature vector, including: calculating the position difference between the barrier spring monitoring attention feature vector and the barrier spring state-related feature vector to obtain a displacement compensation differential feature vector; calculating the position point multiplication between the barrier spring monitoring attention feature vector and the barrier spring state-related feature vector to obtain a displacement compensation position point multiplication feature vector; calculating the position addition between the barrier spring monitoring attention feature vector and the barrier spring state-related feature vector to obtain a displacement compensation position addition feature vector; adding the displacement compensation differential feature vector and the displacement compensation feature vector to obtain a displacement compensation position addition feature vector. The position point product feature vector and the displacement compensation position addition feature vector are input into a one-dimensional convolution layer and maximum pooling is performed to obtain a displacement compensation convolution coding pooling feature vector; the square root of the two norm of the displacement compensation convolution coding pooling feature vector is weighted and added with the one norm of the displacement compensation convolution coding pooling feature vector to obtain a displacement compensation factor; the displacement compensation position point product feature vector and the square root of the length of the barrier spring monitoring attention feature vector are added by position to obtain a displacement compensation intermediate vector; the displacement compensation factor is used as the first weighting coefficient and the second weighted hyperparameter is used as the second weighting coefficient to calculate the position weighted sum of the displacement compensation position addition feature vector and the displacement compensation intermediate vector to obtain a barrier spring fault judgment classification feature vector.

[0065] Among them, the barrier spring monitoring attention feature vector and the barrier spring state association feature vector are corrected by feature fine-grained dimension based on back-propagation displacement compensation to obtain the barrier spring fault judgment classification feature vector, including:

[0066]

[0067]

[0068]

[0069] Among them, V1 represents the barrier spring monitoring attention feature vector, V2 represents the barrier spring state association feature vector, represents positional subtraction, ⊙ represents positional multiplication, represents positional addition, conv1D represents a one-dimensional convolutional layer, MaxPool represents maximum pooling, V μ represents the displacement-compensated convolutional coding pooling feature vector, ·1 represents the first norm, ·2 represents the second norm, α represents the first weighted hyperparameter, τ represents the displacement compensation factor, L represents the length of the gate spring monitoring attention feature vector, β represents the second weighted hyperparameter, V f Represents the classification feature vector of gate spring fault judgment.

[0070] In the technical solution of the present application, in the process of fusing the barrier spring monitoring attention feature vector and the barrier spring state association feature vector, the feature fine-grained displacement compensation between the barrier spring monitoring attention feature vector and the barrier spring state association feature vector is taken into consideration, which may cause the fused barrier spring fault judgment classification feature vector to have local structural collapse. Based on this, in the technical solution of the present application, the barrier spring monitoring attention feature vector and the barrier spring state association feature vector are corrected for feature fine-grained dimensions based on back-propagation displacement compensation.

[0071] Specifically, based on the restricted expression of the foreground-background structure distinction of the high-dimensional manifold of the gate spring monitoring attention feature vector and the gate spring state-associated feature vector by the amplitude of the semantic difference vector between the gate spring monitoring attention feature vector and the gate spring state-associated feature vector, the key feature fine-grained correspondence between the gate spring monitoring attention feature vector and the gate spring state-associated feature vector is modeled, and then the active fine-grained correspondence adjustment propagation is performed through the back-propagation displacement compensation information between the corresponding eigenvalues ​​of the gate spring monitoring attention feature vector and the gate spring state-associated feature vector, so as to avoid the imbalance at the feature vector level of the gate spring monitoring attention feature vector and the gate spring state-associated feature vector in a focusing manner, thereby synchronizing the micro-dimensions of the two feature vectors, ensuring that in the fusion process, the information between the feature vectors can be effectively aligned and combined, thereby avoiding the abnormal accumulation of local structures.

[0072] Furthermore, the barrier spring fault judgment classification feature vector has integrated the multi-dimensional information of the dynamic behavior of the spring and its impact on the system. This feature vector contains data obtained by fusing monitoring attention features and state association features, which together describe the health status of the spring and its impact on the performance of the equipment. In order to transform these complex feature information into specific judgments and decisions, the classifier plays a key role. The classifier can map the feature vector to specific categories, such as normal state and fault state, so as to accurately classify the fault. The classification result is used to determine whether to issue a fault warning alarm. The timely issuance of warning alarms helps to reduce equipment downtime, reduce maintenance costs, and ensure the safety and normal operation of equipment.

[0073] In another specific embodiment of the present application, an intelligent control panel system integrating multiple functions is provided. First, the tidal lamp has a two-way indication function, which can automatically adjust the light state according to the license plate recognition situation: when the license plate is recognized or the entry / exit payment is processed, the light is displayed as green, and the reverse indicator light is red; if the license plate is not recognized, the light automatically returns to green. In terms of the anti-following function, the system can automatically close the gate according to the configured time when the gate is not closed, and intelligently judge the following situation, memorize the number of gate openings, and ensure the smooth passage of vehicles. The remote control gate opening function allows the gate to be started by a remote control signal and the gate opening record is pushed to the cloud background, and the IMS system is linked for early warning. In queue mode, the gate bar will open immediately and remain in the normally open state to meet the needs of vehicle queuing. The real-time monitoring function of the gate status supports the update of the gate bar switch status and temperature detection. In the case of power outages, the system has special processing capabilities to automatically open the gate bar to deal with sudden equipment failures. In addition, the remote restart function can solve the abnormal problems of the identification device or radar by restarting after power outage. The system also intelligently handles invalid orders. If the license plate does not enter or leave the parking lot after the vehicle recognition barrier is raised, the system will automatically handle such invalid orders. Finally, the automatic barrier drop function supports timed detection of the barrier status. If the barrier exceeds the time limit, a voice reminder will be issued and the barrier will automatically drop. These functions together ensure the efficiency and reliability of barrier management.

[0074] In summary, the embodiment of the present application first obtains the barrier spring monitoring image collected by the camera, the motor power consumption data at multiple predetermined time points collected by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points collected by the pressure sensor, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally uses a classifier to determine whether to issue a barrier spring fault warning alarm, so as to discover potential faults in advance, reduce equipment downtime, avoid sudden faults affecting operations, and improve the operating efficiency and reliability of the barrier.

[0075] As described above, the automatic identification system 100 for barrier gate failure according to the embodiment of the present application can be implemented in various terminal devices. In one example, the automatic identification system 100 for barrier gate failure can be integrated into the terminal device as a software module and / or a hardware module. For example, the automatic identification system 100 for barrier gate failure can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the automatic identification system 100 for barrier gate failure can also be one of the many hardware modules of the terminal device.

[0076] Alternatively, in another example, the automatic identification system 100 for barrier gate failure and the terminal device may be separate devices, and the automatic identification system 100 for barrier gate failure may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0077] Figure 5 FIG. 1 is a flow chart of a method for automatically identifying a barrier fault according to an embodiment of the present application. Figure 5 As shown, according to the automatic identification method of barrier gate fault in the embodiment of the present application, it includes: S110, acquiring a barrier gate spring monitoring image captured by a camera, motor power consumption data at multiple predetermined time points captured by a motor power statistics device, and spring load data values ​​at multiple predetermined time points captured by a pressure sensor; S120, extracting a barrier spring monitoring attention feature vector and a barrier spring state association feature vector from the barrier spring monitoring image captured by the camera, the motor power consumption data at multiple predetermined time points captured by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points captured by the pressure sensor; S130, judging whether to issue a barrier spring fault early warning alarm based on the barrier spring monitoring attention feature vector and the barrier spring state association feature vector.

[0078] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned method for automatically identifying a gate fault have been described in the above reference. Figures 1 to 4 The description of the automatic identification system of the gate failure has been introduced in detail, and therefore, its repeated description will be omitted.

[0079] Below, reference Figure 6 To describe an electronic device according to an embodiment of the present application.

[0080] Figure 6 is a block diagram of an electronic device according to an embodiment of the present application. Figure 6 , which shows a schematic diagram of the structure of an electronic device suitable for implementing a terminal device or a server in an embodiment of the present application.

[0081] like Figure 6As shown, the electronic device includes one or more processors, a communication unit, etc. The one or more processors are, for example: one or more central processing units (CPUs), and / or one or more graphics processing units (GPUs), etc. The processor can perform various appropriate actions and processes according to executable instructions stored in a read-only memory (ROM) or executable instructions loaded from a storage part to a random access memory (RAM). The communication unit may include but is not limited to a network card, and the network card may include but is not limited to an IB (Infiniband) network card. The processor may communicate with a read-only memory and / or a random access memory to execute executable instructions, and is connected to the communication unit through a bus and communicates with other target devices through the communication unit, thereby completing the operation corresponding to any method provided in the embodiments of the present application, for example, the primary shard manager shards the features in the feature queue; for any shard, the primary shard manager stores the any shard in the corresponding shard feature library, and records the mapping relationship between the any shard and the corresponding shard feature library in a mapping table; in response to a failure of the primary shard manager, a shard manager is reselected from at least one slave shard manager as the new primary shard manager; the new primary shard manager obtains shard information of the shard currently being written from all shard feature libraries, and recreates the mapping relationship of the shard currently being written based on the shard information of the shard currently being written in all shard feature libraries.

[0082] In addition, various programs and data required for the operation of the device can also be stored in the RAM. The CPU, ROM and RAM are connected to each other via a bus. In the case of RAM, the ROM is an optional module. The RAM stores executable instructions, or writes executable instructions into the ROM at runtime, and the executable instructions enable the processor to perform operations corresponding to any of the above methods of the present invention. The input / output (I / O) interface is also connected to the bus. The communication unit can be integrated or can be configured to have multiple submodules (e.g., multiple IB network cards) and be linked on the bus.

[0083] The following components are connected to the I / O interface: an input section including a keyboard, a mouse, etc.; an output section including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section including a hard disk, etc.; and a communication section including a network interface card such as a LAN card, a modem, etc. The communication section performs communication processing via a network such as the Internet. A drive is also connected to the I / O interface as needed. Removable media, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., are installed on the drive as needed so that a computer program read therefrom is installed into the storage section as needed.

[0084] If necessary, Figure 6The architecture shown is only an optional implementation method. In the specific practice, the above can be modified according to actual needs. Figure 6 The number and type of components can be selected, deleted, added or replaced; different functional components can be set up by separate settings or integrated settings, for example, the GPU and CPU can be set up separately or the GPU can be integrated on the CPU, the communication unit can be set up separately or integrated on the CPU or GPU, etc. These replaceable implementations all fall within the protection scope disclosed in the present invention.

[0085] In addition, an embodiment of the present invention further provides a computer program, including computer-readable code. When the computer-readable code runs on a device, a processor in the device executes instructions for implementing each step in the data processing method of any of the above embodiments of the present invention.

[0086] In addition, an embodiment of the present invention further provides a computer-readable storage medium for storing computer-readable instructions, which, when executed, implement the operations of each step in the data processing method of any of the above embodiments of the present invention.

[0087] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system embodiment, since it basically corresponds to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0088] The method and apparatus of the present invention may be implemented in many ways. For example, the method and apparatus of the present invention may be implemented by software, hardware, firmware, or any combination of software, hardware, firmware. The above order of steps for the method is only for illustration, and the steps of the method of the present invention are not limited to the order specifically described above, unless otherwise specifically stated. In addition, in some embodiments, the present invention may also be implemented as a program recorded in a recording medium, which includes machine-readable instructions for implementing the method according to the present invention. Thus, the present invention also covers a recording medium storing a program for executing the method according to the present invention.

[0089] The description of the present invention is given for the purpose of illustration and description, and is not intended to be exhaustive or to limit the invention to the disclosed form. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present invention and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for particular uses.

Claims

1. An automatic identification system for gate faults, characterized in that: include: A barrier fault identification data acquisition module is used to acquire a barrier spring monitoring image acquired by a camera, motor power consumption data at multiple predetermined time points acquired by a motor power statistics device, and spring load data values ​​at multiple predetermined time points acquired by a pressure sensor; A barrier spring fault identification data processing module, used for extracting a barrier spring monitoring attention feature vector and a barrier spring state association feature vector from the barrier spring monitoring image collected by the camera, the motor power consumption data at multiple predetermined time points collected by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points collected by the pressure sensor; The barrier spring fault warning judgment module is used to judge whether to issue a barrier spring fault warning alarm based on the barrier spring monitoring attention feature vector and the barrier spring state association feature vector.

2. The automatic identification system for gate failure according to claim 1 is characterized in that: The gate fault identification data processing module includes: A barrier spring monitoring feature extraction unit, used for performing feature extraction on the barrier spring monitoring image acquired by the camera to obtain a barrier spring monitoring attention feature vector; A motor power consumption feature extraction unit, used for extracting features from the motor power consumption data at a plurality of predetermined time points collected by the motor power statistics device to obtain a motor power consumption feature vector; A spring load data feature extraction unit, used for performing feature extraction on the spring load data values ​​at a plurality of predetermined time points collected by the pressure sensor to obtain a spring load data feature vector; The barrier spring state characteristic association unit is used to associate the motor power consumption characteristic vector with the spring load data characteristic vector to obtain the barrier spring state association characteristic vector.

3. The automatic identification system for gate failure according to claim 2 is characterized in that: The gate spring monitoring feature extraction unit comprises: A barrier spring monitoring target detection subunit, used for passing the barrier spring monitoring image acquired by the camera through a barrier spring monitoring target detection network model to obtain a barrier spring region of interest; A barrier spring monitoring spatial attention encoding subunit, used for passing the barrier spring region of interest through a barrier spring monitoring spatial attention feature encoder to obtain a barrier spring monitoring attention feature map; The barrier spring monitoring attention pooling subunit is used to perform global mean pooling on the barrier spring monitoring attention feature map to obtain a barrier spring monitoring attention feature vector.

4. The automatic identification system for gate failure according to claim 3 is characterized in that: The motor power consumption feature extraction unit comprises: constructing the motor power consumption data at a plurality of predetermined time points collected by the motor power statistics device into a motor power consumption input vector; The motor power consumption input vector is passed through a motor power consumption timing feature extractor to obtain the motor power consumption feature vector.

5. The automatic identification system for gate failure according to claim 4 is characterized in that: The spring load data feature extraction unit comprises: constructing the spring load data values ​​at a plurality of predetermined time points collected by the pressure sensor into a spring load data input vector; The spring load data input vector is passed through a spring load data sequence encoder to obtain the spring load data feature vector.

6. The automatic identification system for gate failure according to claim 5 is characterized in that: The barrier spring state characteristic association unit comprises: Create a new Spring Boot project using SpringInitializr and add Spring Web, SpringCloud, Spring Boot DevTools, and Lombok dependencies; Configure the necessary properties in the project's application.properties file, including the application name, server port, and the URL for Eureka service registration and discovery; Create a controller class FeatureVectorController ler, define an HTTP POST request processing method in this class, receive the request body containing the motor power consumption feature vector and the spring load data feature vector, and call the service class method for calculation; Enable the service discovery client function of Spring Boot application and Spring Cloud in the main application class FeatureVectorApplication; Use the Maven command to package the project into a JAR file and run the generated JAR file through the command to deploy and verify the feature vector calculation service.

7. The automatic identification system for gate failure according to claim 6 is characterized in that: The gate spring fault early warning judgment module includes: A barrier spring feature fusion unit is used to perform feature fine-grained dimension correction on the barrier spring monitoring attention feature vector and the barrier spring state association feature vector based on back-propagation displacement compensation to obtain a barrier spring fault judgment classification feature vector; The barrier spring fault warning alarm judgment classification unit is used to pass the barrier spring fault judgment classification feature vector through a classifier to obtain a classification result, and the classification result is used to determine whether to issue a barrier spring fault warning alarm.

8. The automatic identification system for gate failure according to claim 7 is characterized in that: The barrier spring feature fusion unit comprises: Calculating the position difference between the barrier spring monitoring attention feature vector and the barrier spring state association feature vector to obtain a displacement compensation difference feature vector; Calculate the position point multiplication between the barrier spring monitoring attention feature vector and the barrier spring state association feature vector to obtain a displacement compensation position point multiplication feature vector; Calculate the positional addition between the barrier spring monitoring attention feature vector and the barrier spring state association feature vector to obtain a displacement compensation positional addition feature vector; Inputting the displacement-compensated differential feature vector, the displacement-compensated positional point-product feature vector and the displacement-compensated positional addition feature vector into a one-dimensional convolutional layer and performing maximum pooling to obtain a displacement-compensated convolutional coding pooling feature vector; The square root value of the two-norm of the displacement-compensated convolutional coding pooling feature vector is weighted and added to the one-norm of the displacement-compensated convolutional coding pooling feature vector to obtain a displacement compensation factor; The displacement compensation point-by-position feature vector and the square root of the length of the barrier spring monitoring attention feature vector are added according to the position to obtain a displacement compensation intermediate vector; Using the displacement compensation factor as the first weighting coefficient and the second weighted hyperparameter as the second weighting coefficient, the position-weighted sum of the displacement compensation addition feature vector and the displacement compensation intermediate vector is calculated to obtain a barrier spring fault judgment classification feature vector.

9. A method for automatically identifying a gate fault, characterized in that: include: Acquire a gate spring monitoring image collected by a camera, motor power consumption data at multiple predetermined time points collected by a motor power statistics device, and spring load data values ​​at multiple predetermined time points collected by a pressure sensor; Extracting a barrier spring monitoring attention feature vector and a barrier spring state association feature vector from the barrier spring monitoring image collected by the camera, the motor power consumption data at multiple predetermined time points collected by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points collected by the pressure sensor; Based on the barrier spring monitoring attention feature vector and the barrier spring state association feature vector, it is determined whether to issue a barrier spring failure warning alarm.

10. The automatic identification method of gate fault according to claim 9, characterized in that: Extracting a barrier spring monitoring attention feature vector and a barrier spring state association feature vector from the barrier spring monitoring image collected by the camera, the motor power consumption data at multiple predetermined time points collected by the motor power statistics device, and the spring load data values ​​at multiple predetermined time points collected by the pressure sensor, including: Performing feature extraction on the barrier spring monitoring image captured by the camera to obtain a barrier spring monitoring attention feature vector; Performing feature extraction on the motor power consumption data at a plurality of predetermined time points collected by the motor power statistics device to obtain a motor power consumption feature vector; Performing feature extraction on the spring load data values ​​at a plurality of predetermined time points collected by the pressure sensor to obtain a spring load data feature vector; The motor power consumption characteristic vector and the spring load data characteristic vector are associated to obtain the barrier spring state associated characteristic vector.

Images