Process control unit anomaly detection method and system based on indicating lamp image recognition

Abstract

The invention discloses a process control unit anomaly detection method and system based on indicator lamp image recognition, and relates to the technical field of industrial controller fault diagnosis and alarm. Preprocessing the image of the indicator lamp of the process control unit, and extracting features related to the state of the indicator lamp; inputting the characteristics related to the state of the indicating lamp into a pre-trained machine learning classification model for detection and identification; the machine learning classification model outputs a detection result which is used for indicating whether the indicator light of the process control unit is abnormal; and maintaining the indicator light of the process control unit according to the detection result. The anomaly detection method mainly depends on computer vision and image processing technologies, and the state of the indicator lamp is detected by using an image recognition algorithm so as to judge whether an anomaly exists or not. By using the anomaly detection method provided by the invention, the information concerned by the user in the operation process of the process control unit can be monitored in real time.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial controller fault diagnosis and alarm, and in particular to a process control unit abnormality detection method and system based on indicator light image recognition. Background Art

[0002] The process control unit (PCU) plays an important role in the field of industrial automation. In the numerical control system, the process control unit can be understood as an industrial computer, which has independent CPU, hard disk and other hardware components. Its main functions include processing the received data, executing control tasks, and sending the processed process information to the system bus. It provides a human-machine interface so that operators can perform tasks such as operation and program editing. In the production process, it receives various signals from the field, performs calculations and processing, obtains operating parameters, and drives the actuator to realize detection, control and protection functions.

[0003] Process control units are widely used in systems that require real-time control of continuous processes. They usually consist of central processing unit (CPU) modules, communication processor (CP) modules, Ethernet switches, repeaters, input / output modules, power supply and UPS modules, insulation monitors and other components. These components work together to achieve data processing, communication and control functions. Process control units are generally installed in the centralized control room or control room, and communicate data with the optical fiber ring network through Ethernet communication modules. It plays an important role in improving production efficiency, ensuring production safety, and optimizing resource allocation.

[0004] As we all know, process control units may experience power anomalies, operation anomalies, master-slave anomalies, and IO bus anomalies during operation. Currently, the main method for detecting anomalies in process control units is for maintenance personnel to regularly open the cabinet and observe the status of each indicator light. This not only requires a lot of manpower costs, but also may result in missed detections and false detections. Summary of the invention

[0005] The purpose of the present invention is to address the problems in the above-mentioned prior art and provide a process control unit abnormality detection method and system based on indicator light image recognition, which adopts an image recognition method based on deep learning to perform process control unit indicator light abnormality detection, improve the accuracy and robustness of detection, and reduce the missed detection and false detection of indicator light abnormality detection.

[0006] In order to achieve the above object, the present invention has the following technical solutions: A method for detecting abnormality of a process control unit based on indicator light image recognition, comprising: Acquire images of indicator lights of process control units; Preprocessing the captured image of the indicator light of the process control unit, and extracting features related to the indicator light state from the preprocessed image; wherein the features related to the indicator light state include the color, shape, size and brightness of the indicator light; Input the features related to the indicator light status into a pre-trained machine learning classification model for detection and identification; wherein the machine learning classification model adopts a support vector machine SVM model or a convolutional neural network CNN model; The machine learning classification model outputs a detection result, and the detection result is used to indicate whether there is an abnormality in the indicator light of the process control unit; Maintaining the indicator light of the process control unit according to the detection result; Wherein, maintaining the indicator light of the process control unit according to the detection result includes: When the test result indicates that the running indicator light is faulty, check whether the card type is correct in the tool software; when the test result indicates that the input and output network lights are faulty, confirm in the configuration software whether the relevant boards are created at the corresponding positions of the input and output network lights; when the test result indicates that the channel light is faulty, confirm whether the corresponding port is wired and whether the port cable is loosely connected.

[0007] As a preferred solution, the image of the indicator light of the process control unit is collected by a charge coupled device (CCD) camera.

[0008] As a preferred solution, the preprocessing of the image of the process control unit indicator light includes eliminating noise in the image, improving the contrast of the indicator light in the image, and performing color correction on the image.

[0009] As a preferred scheme, before inputting the features related to the indicator light state into a pre-trained machine learning classification model for detection and identification, the method also includes: obtaining an image dataset of a normal indicator light state and an image dataset of an abnormal indicator light state; training the machine learning classification model based on the image dataset of the normal indicator light state and the image dataset of the abnormal indicator light state to obtain a first machine learning classification model; performing model evaluation on the first machine learning classification model using a cross-validation method to obtain a target machine learning classification model; wherein the target machine learning classification model includes a mapping relationship between features related to the indicator light state and indicator light failure results.

[0011] As a preferred scheme, the inputting of the features related to the indicator light status into a pre-trained machine learning classification model for detection and identification includes: the machine learning classification model compares according to a set threshold, and judges whether the current status of the indicator light is normal or abnormal based on the mapping relationship between the features related to the indicator light status obtained in pre-training and the indicator light fault results.

[0012] As a preferred solution, the method also includes: when the machine learning classification model outputs that there is an abnormality in the process control unit indicator light, triggering an alarm mechanism, and alerting the remote operation and maintenance platform through a wireless network; transmitting the collected process control unit indicator light image to the storage unit of the remote operation and maintenance platform, and recording the time, location and abnormal situation of the process control unit indicator light abnormality.

[0013] A process control unit abnormality detection system based on indicator light image recognition, comprising: An indicator light image acquisition module is used to acquire images of the indicator lights of the process control unit; An indicator light image feature extraction module, used to preprocess the acquired image of the indicator light of the process control unit, and extract features related to the indicator light state from the preprocessed image; wherein the features related to the indicator light state include the color, shape, size and brightness of the indicator light; A machine learning classification model judgment module is used to input the features related to the indicator light state into a pre-trained machine learning classification model for detection and identification; wherein the machine learning classification model adopts a support vector machine SVM model or a convolutional neural network CNN model; A machine learning classification result output module, used for the machine learning classification model to output a detection result, wherein the detection result is used to indicate whether there is an abnormality in the indicator light of the process control unit; An abnormal maintenance module, used to maintain the indicator light of the process control unit according to the detection result; Wherein, maintaining the indicator light of the process control unit according to the detection result includes: When the test result indicates that the running indicator light is faulty, check whether the card type is correct in the tool software; when the test result indicates that the input and output network lights are faulty, confirm in the configuration software whether the relevant boards are created at the corresponding positions of the input and output network lights; when the test result indicates that the channel light is faulty, confirm whether the corresponding port is wired and whether the port cable is loosely connected.

[0014] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the process control unit abnormality detection method based on indicator light image recognition is implemented.

[0015] Compared with the prior art, the present invention has at least the following beneficial effects: First, the image of the indicator light of the process control unit is captured by an image acquisition device. These images contain enough information to distinguish different indicator light states. The image of the indicator light of the process control unit is input into the microprocessor, and the image is preprocessed, feature analyzed and identified in the microprocessor, and the process control unit is diagnosed with reference to the reference image in the database; if the process control unit fails, an alarm is sent to the remote operation and maintenance platform through a wireless network and the collected image is transmitted. The use of image recognition technology for process control unit diagnosis can effectively improve the real-time and accuracy of the monitoring system and solve the interference of the analysis results affected by human factors. The process control unit indicator light anomaly detection method based on image recognition of the present invention mainly relies on computer vision and image processing technology. By capturing the image of the indicator light of the process control unit, the state of the indicator light is detected by the image recognition algorithm to determine whether there is an abnormality. Using the anomaly detection method proposed by the present invention, the information that users are concerned about during the operation of the process control unit can be monitored in real time, and then used to assist on-duty personnel and maintenance personnel to carry out their respective work, reduce the difficulty of work, reduce the probability of errors, and effectively improve work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 It is a main flow chart of a method for detecting abnormality of a process control unit based on indicator light image recognition according to an embodiment of the present invention.

[0018] Figure 2 It is a structural diagram of a process control unit abnormality detection system based on indicator light image recognition according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, ordinary technicians in this field can also obtain other embodiments without making creative work.

[0020] With the enhancement of computer computing power and the rapid development of data storage technology, image recognition technology has made breakthrough progress. At present, image recognition technology is mainly based on deep learning and convolutional neural networks, which can process a large amount of image data and extract feature information from it. Image recognition technology can not only identify basic elements such as objects, scenes, and texts, but also perform advanced image analysis such as face recognition and lighting recognition. In the future, image recognition technology will also be more widely used in fields such as transportation, energy, and smart industry. In addition, with the continuous development of deep learning technology, image recognition methods based on deep learning have performed well in processing complex and varied image data. Therefore, it is possible to consider applying deep learning technology to abnormal detection of indicator lights in process control units to improve the accuracy and robustness of detection.

[0021] The embodiment of the present invention proposes a method for detecting anomalies of process control units based on indicator light image recognition. By adopting a low-power domestic microprocessor and an image acquisition module equipped with AI performance, image acquisition and intelligent recognition of the indicator lights of the process control unit are realized. For the abnormal detection results of the indicator lights obtained by research and judgment, not only local alarms can be realized, but also the abnormal detection results can be transmitted to the remote operation and maintenance platform, thereby accurately and quickly realizing abnormal detection of the process control unit.

[0022] See also Figure 1 The process control unit abnormality detection method based on indicator light image recognition according to an embodiment of the present invention includes: S1, collecting images of the indicator lights of the process control unit; S2, preprocessing the collected image of the indicator light of the process control unit, and extracting features related to the indicator light state from the preprocessed image; wherein the features related to the indicator light state include the color, shape, size and brightness of the indicator light; S3, inputting the features related to the indicator light status into a pre-trained machine learning classification model for detection and identification; wherein the machine learning classification model adopts a support vector machine SVM model or a convolutional neural network CNN model; S4, the machine learning classification model outputs a detection result, and the detection result is used to indicate whether there is an abnormality in the indicator light of the process control unit; S5. Maintain the indicator light of the process control unit according to the detection result; Wherein, maintaining the indicator light of the process control unit according to the detection result includes: When the test result indicates that the running indicator light is faulty, check whether the card type is correct in the tool software; when the test result indicates that the input and output network lights are faulty, confirm in the configuration software whether the relevant boards are created at the corresponding positions of the input and output network lights; when the test result indicates that the channel light is faulty, confirm whether the corresponding port is wired and whether the port cable is loosely connected.

[0023] In a possible implementation, step S1 collects an image of the indicator light of the process control unit through a CCD (Charge Coupled Device) camera. The charge coupled device CCD is composed of a number of charge coupled units, and the image is captured and transmitted through the storage and transfer of charges. The basic unit of the charge coupled device CCD is a MOS (metal-oxide-semiconductor) capacitor. When light is irradiated to the surface of the CCD, electron-hole pairs are generated inside the semiconductor, and the electrons are collected in the potential well to form signal charges, the number of which is proportional to the intensity of the incident light. Through the control of the clock pulse signal, these signal charges can be directed to move inside the CCD and finally converted into electrical signals for output. CCD components are usually installed on photosensitive chips, and the density and number of their installation determine the optical resolution. When light and images pass through the lens and are projected onto the surface of the CCD, the CCD generates current and converts the sensed content into digital data for storage. Therefore, the more pixels the CCD has and the larger the size of a single pixel, the clearer the collected image will be. CCD is highly sensitive to light and can capture clear images in low-light environments. Compared with other sensors, CCD can maintain good imaging effects in strong light environments and is not prone to overexposure. Due to the structural characteristics of CCD, its imaging distortion is small and can restore more realistic images. CCD cameras are small in size, easy to carry and install; at the same time, their service life is relatively long. CCD cameras can also maintain stable imaging effects in vibration environments and are suitable for various complex environments.

[0024] In a possible implementation manner, the preprocessing of the image of the process control unit indicator light in step S2 includes eliminating noise in the image, improving the contrast of the indicator light in the image, and performing color correction on the image.

[0025] Image denoising is a key technology in image preprocessing. Its purpose is to restore a noise-free image from a noisy image to improve image quality and the accuracy of subsequent image analysis. Common image denoising methods include: Mean filtering: Reduces noise by replacing pixel values ​​with the average of their neighboring pixel values, but may result in loss of image details.

[0026] Median filtering: Replaces each pixel value with the median of its neighboring pixel values, which is very effective in removing salt and pepper noise.

[0027] Gaussian filtering: Uses a Gaussian function as a weight to reduce noise. It is suitable for removing Gaussian noise, and the degree of smoothing can be controlled by adjusting the standard deviation.

[0028] Bilateral filtering: retains edge information while denoising. It is suitable for scenes where you want to keep edges clear while denoising.

[0029] Wavelet denoising: By decomposing the image into sub-bands of different frequencies and then performing threshold processing on these sub-bands to remove noise, the important features of the image can be well preserved.

[0030] Non-local mean denoising: It uses the information of similar regions in the image to remove noise, which is suitable for retaining image details.

[0031] The specific denoising method you choose depends on the type of noise in the image and the requirements after denoising.

[0032] Secondly, the influence of light intensity on image quality is the main factor, because the light intensity of the light source cannot guarantee uniform illumination of the target object, and there is always a part of the target object that receives lower light intensity than other parts. Therefore, improving the contrast of the image can make the target object in the image clearer and more identifiable. Common contrast enhancement methods include: Histogram equalization: Enhance the contrast of the image by stretching the grayscale range of the image to make the histogram distribution of the image more uniform.

[0033] Contrast stretching: Enhance the contrast of an image manually or automatically by adjusting the image's brightness and contrast parameters.

[0034] In addition, the purpose of color correction is to adjust the color of the image. For the image of the indicator light of the process control unit, color correction can ensure that the color of the indicator light is accurate, thereby avoiding misjudgment. Common color correction methods include: Color correction based on supervised learning: Color correction of images is performed using a standard color palette. The colors in the image are adjusted based on the color values ​​on the standard color palette using algorithms such as polynomial regression, artificial neural networks, or SVR (Support Vector Regression).

[0035] White balance adjustment: Adjust the ratio of the red, green and blue color channels in the image to make the white part of the image look more natural.

[0036] In a possible implementation manner, step S2 extracts features related to the indicator light state including the indicator light's color, shape, size, brightness, etc. These features can intuitively reflect the state of the process control unit.

[0037] The color of an indicator light is often used to indicate different states or conditions. For example, green may indicate normal operation or readiness, yellow or orange may indicate a warning or attention-required state, and red often indicates an emergency stop or malfunction. Color is an intuitive and easy-to-understand signal that quickly attracts the operator's attention.

[0038] While the shape of an indicator light may not convey status information as directly as color, it is still an important feature. Different shapes may be used to distinguish different types of indicators or to represent different functions. For example, a round indicator light may be used to indicate power status, while a square or rectangular indicator light may be used to indicate a specific operating mode or fault type.

[0039] The size of an indicator light may vary depending on the application, but it can also provide clues about its importance. For example, a larger indicator light may indicate a more critical status or an area that requires more attention. Additionally, size may affect the visibility of the indicator light at a distance or in different lighting conditions.

[0040] Brightness is another important feature that can be used to emphasize the status of an indicator light or to attract the operator's attention. For example, in an emergency, a red indicator light may flash at a higher brightness to alert the operator to pay attention and take action. In addition, brightness can be adjusted based on ambient lighting conditions to ensure that the indicator light is clearly visible in a variety of environments.

[0041] When processing images related to the status of indicator lights, these features can be used for image recognition, feature extraction, and status monitoring, which can automatically monitor and record the status of equipment, reduce human errors, and improve production efficiency.

[0042] In a possible implementation, the machine learning classification model in step S3 adopts a support vector machine (SVM) model or a convolutional neural network (CNN) model.

[0043] Support vector machine is a binary classification model, whose basic model is a linear classifier with the largest margin defined in the feature space. When the data is linearly separable, the optimal classification hyperplane is found by solving a convex quadratic programming optimization algorithm to maximize the margin between the two types of samples. In SVM, several training sample points that are closest to the hyperplane and meet certain conditions are called support vectors. These points are crucial to determining the position of the hyperplane. A hyperplane is a straight line (in two-dimensional space) or a plane (in three-dimensional space) used to divide data categories, and a hyperplane in a higher-dimensional space. The goal of SVM is to find a hyperplane that maximizes the distance (margin) between the two types of sample points and the hyperplane. Such a hyperplane is called the maximum margin hyperplane. Support vectors are sample points closest to the hyperplane, and they determine the position of the hyperplane.

[0044] Convolutional neural network is an important model in the field of deep learning, especially in image processing. CNN can better capture local features in the image by using convolution kernels for local perception. This local perception enables CNN to effectively extract detail information when processing images and has good spatial invariance. In CNN, the parameters of the convolution kernel are shared, which means that the same convolution kernel can extract features at different positions of the entire input image. This parameter sharing reduces the number of network parameters, reduces the risk of overfitting, and makes the model more generalizable. CNN can reduce the size of feature maps and the amount of calculation through pooling operations, while retaining the main feature information. This helps to improve the efficiency and robustness of the model. CNN uses multiple layers of convolution and pooling operations to gradually extract abstract features in the image from low to high levels. This spatial hierarchical structure enables CNN to process features at different levels and capture information at different scales in the image. Since convolution and pooling operations in CNN can be calculated in parallel, it has good high performance on hardware accelerators such as GPUs.

[0045] The CNN model mainly includes the following layers: Input layer: receives input data, such as images. For image data, the input layer is usually a three-dimensional array, including the width, height, and number of channels of the image (such as 3 channels for RGB images).

[0046] Convolutional layer: extracts features of input data through convolution operation. Each neuron in the convolutional layer is only connected to the local area of ​​the previous layer, and convolution calculation is performed using weight parameters. The output of the convolutional layer is usually called a feature map.

[0047] Activation function: Perform nonlinear transformation on the output of the convolutional layer to increase the nonlinear expression ability of the model. Common activation functions include ReLU, sigmoid, tanh, etc.

[0048] Pooling layer: downsamples the feature map output by the convolutional layer to reduce the size and number of parameters of the feature map. Common pooling methods include maximum pooling and average pooling.

[0049] Fully connected layer: globally connects the features extracted by the convolutional layer and the pooling layer, and outputs the final classification or regression result. Each neuron in the fully connected layer is connected to all neurons in the previous layer.

[0050] Furthermore, the machine learning classification model of the embodiment of the present invention is trained using a known image dataset of normal and abnormal indicator light states.

[0051] Optionally, before inputting the features related to the indicator light state into a pre-trained machine learning classification model for detection and identification, the method further includes: Acquire an image dataset of a normal indicator light state and an image dataset of an abnormal indicator light state; The machine learning classification model is trained based on the image dataset of the normal indicator light state and the image dataset of the abnormal indicator light state to obtain a first machine learning classification model; The first machine learning classification model is evaluated using a cross-validation method to obtain a target machine learning classification model; wherein the target machine learning classification model includes a mapping relationship between features related to the indicator light state and indicator light failure results. Optionally, when the embodiment of the present invention inputs features related to the indicator light state into a pre-trained machine learning classification model, the machine learning classification model compares according to a set threshold value, and determines whether the current state of the indicator light is normal or abnormal based on the mapping relationship between the features related to the indicator light state obtained in advance and the indicator light failure results.

[0052] Optionally, the method also includes: when the machine learning classification model outputs an abnormality in the process control unit indicator light, triggering an alarm mechanism, and alerting the remote operation and maintenance platform through a wireless network; transmitting the collected process control unit indicator light image to a storage unit of the remote operation and maintenance platform, recording the time, location and abnormal situation of the process control unit indicator light abnormality, and notifying relevant personnel to handle it, so that the indicator light failure can be handled in time.

[0053] The present invention proposes a method for detecting abnormalities of indicator lights based on image recognition in response to abnormalities that may occur during the operation of the process control unit. By adopting a low-power domestic microprocessor and image acquisition module equipped with AI performance, image acquisition and intelligent recognition of the indicator lights of the process control unit are realized. For the abnormal detection results of the indicator lights obtained through research and judgment, not only local alarms can be realized, but also the abnormal detection results can be transmitted to the remote operation and maintenance platform. Using the above-mentioned method of the present invention, the information that users are concerned about during the operation of the process control unit can be monitored in real time, which can be used to assist on-duty personnel and maintenance personnel in carrying out their respective work, reduce the difficulty of work, reduce the probability of errors, and improve work efficiency.

[0054] See also Figure 2 Another embodiment of the present invention further provides a process control unit abnormality detection system 200 based on indicator light image recognition, comprising: The indicator light image acquisition module 201 is used to acquire the image of the indicator light of the process control unit; The indicator light image feature extraction module 202 is used to pre-process the image of the indicator light of the process control unit, and extract features related to the indicator light state from the pre-processed image; wherein the features related to the indicator light state include the color, shape, size and brightness of the indicator light; The machine learning classification model judgment module 203 is used to input the features related to the indicator light state into a pre-trained machine learning classification model for detection and identification; wherein the machine learning classification model adopts a support vector machine SVM model or a convolutional neural network CNN model; The machine learning classification result output module 204 is used for the machine learning classification model to output the detection result, and the detection result is used to indicate whether the indicator light of the process control unit is abnormal; An abnormal maintenance module 205 is used to maintain the indicator light of the process control unit according to the detection result; Wherein, maintaining the indicator light of the process control unit according to the detection result includes: When the test result indicates that the running indicator light is faulty, check whether the card type is correct in the tool software; when the test result indicates that the input and output network lights are faulty, confirm in the configuration software whether the relevant boards are created at the corresponding positions of the input and output network lights; when the test result indicates that the channel light is faulty, confirm whether the corresponding port is wired and whether the port cable is loosely connected.

[0055] Another embodiment of the present invention further provides an electronic device, comprising: A memory storing at least one instruction; and a processor executing the instruction stored in the memory to implement the process control unit abnormality detection method based on indicator light image recognition.

[0056] Another embodiment of the present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the process control unit abnormality detection method based on indicator light image recognition is implemented.

[0057] Exemplarily, the instructions stored in the memory may be divided into one or more modules / units, which are stored in a computer-readable storage medium and executed by the processor to complete the process control unit abnormality detection method based on indicator light image recognition according to the embodiment of the present invention. The one or more modules / units may be a series of computer-readable instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the server.

[0058] The electronic device may be a computing device such as a smart phone, a notebook, a PDA, and a cloud server. The electronic device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the electronic device may also include more or fewer components, or a combination of certain components, or different components, for example, the electronic device may also include an input / output device, a network access device, a bus, etc.

[0059] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0060] The memory may be an internal storage unit of the server, such as a hard disk or memory of the server. The memory may also be an external storage device of the server, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the server. Furthermore, the memory may include both an internal storage unit of the server and an external storage device. The memory is used to store the computer-readable instructions and other programs and data required by the server. The memory may also be used to temporarily store data that has been output or is to be output.

[0061] It should be noted that the information interaction, execution process and other contents between the above-mentioned module units are based on the same concept as the method embodiment. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0062] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0063] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the camera device / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0064] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0065] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A process control unit abnormality detection method based on indicator light image recognition, characterized in that: include: Acquire images of indicator lights of process control units; Preprocessing the captured image of the indicator light of the process control unit, and extracting features related to the indicator light state from the preprocessed image; wherein the features related to the indicator light state include the color, shape, size and brightness of the indicator light; Input the features related to the indicator light status into a pre-trained machine learning classification model for detection and identification; wherein the machine learning classification model adopts a support vector machine SVM model or a convolutional neural network CNN model; The machine learning classification model outputs a detection result, and the detection result is used to indicate whether there is an abnormality in the indicator light of the process control unit; Maintaining the indicator light of the process control unit according to the detection result; Wherein, maintaining the indicator light of the process control unit according to the detection result includes: When the test result indicates that the running indicator light is faulty, check whether the card type is correct in the tool software; when the test result indicates that the input and output network lights are faulty, confirm in the configuration software whether the relevant boards are created at the corresponding positions of the input and output network lights; when the test result indicates that the channel light is faulty, confirm whether the corresponding port is wired and whether the port cable is loosely connected.

2. The process control unit abnormality detection method based on indicator light image recognition according to claim 1 is characterized in that: The images of the indicator lights of the process control unit are collected by a charge coupled device (CCD) camera.

3. The process control unit abnormality detection method based on indicator light image recognition according to claim 1 is characterized in that: The preprocessing of the image of the process control unit indicator light includes eliminating noise, improving contrast and color correction.

4. According to the method for detecting abnormality of a process control unit based on indicator light image recognition in claim 1, before inputting the features related to the indicator light state into a pre-trained machine learning classification model for detection and recognition, the method further comprises: Acquire an image dataset of a normal indicator light state and an image dataset of an abnormal indicator light state; The machine learning classification model is trained based on the image dataset of the normal indicator light state and the image dataset of the abnormal indicator light state to obtain a first machine learning classification model; The first machine learning classification model is evaluated using a cross-validation method to obtain a target machine learning classification model; wherein the target machine learning classification model includes a mapping relationship between features related to the indicator light status and the indicator light failure results.

5. The process control unit abnormality detection method based on indicator light image recognition according to claim 4 is characterized in that: The inputting the features related to the indicator light status into a pre-trained machine learning classification model for detection and identification includes: The machine learning classification model compares according to a set threshold value and determines whether the current state of the indicator light is normal or abnormal based on the mapping relationship between the features related to the indicator light state obtained in advance and the indicator light failure results.

6. The process control unit abnormality detection method based on indicator light image recognition according to claim 1 is characterized in that: The method further comprises: When the output process of the machine learning classification model shows that the indicator light of the control unit is abnormal, an alarm mechanism is triggered, and an alarm is sent to the remote operation and maintenance platform through the wireless network; The collected process control unit indicator light image is transmitted to the storage unit of the remote operation and maintenance platform, and the time, location and abnormal situation of the abnormality of the process control unit indicator light are recorded.

7. A process control unit abnormality detection system based on indicator light image recognition, characterized in that: include: An indicator light image acquisition module is used to acquire images of the indicator lights of the process control unit; An indicator light image feature extraction module, used to preprocess the acquired image of the indicator light of the process control unit, and extract features related to the indicator light state from the preprocessed image; wherein the features related to the indicator light state include the color, shape, size and brightness of the indicator light; A machine learning classification model judgment module is used to input the features related to the indicator light state into a pre-trained machine learning classification model for detection and identification; wherein the machine learning classification model adopts a support vector machine SVM model or a convolutional neural network CNN model; A machine learning classification result output module, used for the machine learning classification model to output a detection result, wherein the detection result is used to indicate whether there is an abnormality in the indicator light of the process control unit; An abnormal maintenance module, used to maintain the indicator light of the process control unit according to the detection result; Wherein, maintaining the indicator light of the process control unit according to the detection result includes: When the test result indicates that the running indicator light is faulty, check whether the card type is correct in the tool software; when the test result indicates that the input and output network lights are faulty, confirm in the configuration software whether the relevant boards are created at the corresponding positions of the input and output network lights; when the test result indicates that the channel light is faulty, confirm whether the corresponding port is wired and whether the port cable is loosely connected.

8. The process control unit abnormality detection method based on indicator light image recognition according to claim 7 is characterized in that: The images of the indicator lights of the process control unit are collected by a charge coupled device (CCD) camera.

9. The process control unit abnormality detection method based on indicator light image recognition according to claim 7 is characterized in that: The preprocessing of the image of the process control unit indicator light includes eliminating noise, improving contrast and color correction.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for detecting abnormality of a process control unit based on indicator light image recognition as claimed in any one of claims 1 to 6 is implemented.

Images