Equipment fault detection method based on indicator light flicker detection

Abstract

The invention discloses an equipment fault detection method based on indicator light flicker detection, and relates to the technical field of image processing, and the method comprises the following steps: collecting videos or pictures of equipment by using a high-definition camera; carrying out picture fusion on the collected videos or pictures in a pixel space by utilizing a picture fusion method; performing indicator light detection on the fused picture by using the target monitoring model to obtain a position frame and color information of an indicator light; according to the detected indicating lamp position frame, image blocks corresponding to the indicating lamp position frame are extracted from the multiple pictures to be processed, and the flicker state and the flicker frequency of the indicating lamp are obtained; carrying out fault judgment and alarm according to a detection result; compared with the existing fault detection method, the method has the advantages that the flicker information of the index lamp of the equipment can be obtained, and the equipment fault detection or alarm basis provided based on the flicker information is more accurate, so that the accuracy of equipment fault detection can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and specifically to a device fault detection method based on indicator light flashing detection. Background Art

[0002] In the inspection scenario of an information computer room, the indicator light is one of the most prominent features for indicating the operating status of a device; generally, a red light indicates a fault, a yellow light indicates an alarm, a green light indicates normal initialization, and a blue light indicates normal operation, which is used to determine whether the device has a fault; in view of this, some researchers have proposed to identify the color of the indicator light through an image acquisition device and an image processing algorithm to determine the operating or fault status of the device; most of these methods belong to static detection, detecting its color information through an indicator light detection or recognition model, and then judging the device fault to output a warning; however, in the real inspection scenario, there is more a device that uses the flashing state of the indicator light to represent the fault level of the monitoring object. For its fault detection and fault level classification, not only the color information of the indicator light is required, but also its flashing state needs to be detected to comprehensively judge whether it has a fault. Summary of the Invention

[0003] The purpose of the present invention is to provide a device fault detection method based on indicator light flashing detection to solve the problems proposed in the prior art.

[0004] To achieve the above purpose, the present invention provides the following technical solution: A device fault detection method based on indicator light flashing detection, including the following steps:

[0005] S1. Use a high-definition camera to collect the video or picture of the device;

[0006] S2. In the pixel space, use the picture fusion method to fuse the collected video or picture;

[0007] S3. Use the target detection model to detect the indicator light on the fused picture to obtain the position box and color information of the indicator light;

[0008] S4. According to the detected indicator light position box information, extract the image block corresponding to the indicator light position box on the picture for processing to obtain the flashing state and flashing frequency of the indicator light;

[0009] S5. Make a fault judgment and give an alarm according to the detection results of the flashing frequency and flashing color of the indicator light.

[0010] Further, in step S1: The to-be-fused pictures are obtained by using a high-definition camera to collect the video or pictures of the device to be detected. If the video is collected, the shooting frame rate of the video is collected at the same time, and the video is extracted at equal intervals according to a fixed sampling frequency S to obtain N1 pictures. If pictures are collected, at least two pictures are collected, and the shooting time and time interval of the pictures are recorded.

[0011] Further, in step S2: When performing image fusion, the collected pictures can be fused into one picture by using the PCA-based image fusion algorithm and the pixel-rule-based image fusion method.

[0012] Further, when using the PCA-based image fusion algorithm, first, N2 pictures are extracted from the to-be-fused pictures, and the extracted pictures are converted into grayscale pictures. The formula for converting the extracted pictures into grayscale pictures is: Gray = R * 0.299 + G * 0.587 + B * 0.114, where Gray is used to identify the grayscale pixel value at the point (x, y), R represents the pixel value of the R channel, G represents the pixel value of the G channel, and B represents the pixel value of the B channel. After converting the pictures into grayscale pictures, the covariance matrix of the grayscale pictures is calculated. The formula for calculating the covariance matrix is:

[0013]

[0014] Then, the eigenvalues λ i and eigenvectors The eigenvalues and eigenvectors correspond one by one. The eigenvalues and eigenvectors are sorted in descending order of eigenvalues to obtain the eigenvalue set {λ1, λ1,..., λ n}, and the eigenvector set is And according to the formula The principal component of each grayscale value is calculated, where PC(x, y) represents the principal component of the grayscale picture; G(x, y) represents the grayscale value at (x, y). Based on the original pictures that were not extracted when the pictures were extracted and converted into grayscale pictures, the inverse transformation of the principal component of the grayscale value is performed, and a fused picture is obtained after the inverse transformation.

[0015] Further, if the pixel-rule-based image fusion method is used, first, the to-be-fused pictures are extracted from the collected video or pictures, and all the pictures are converted into grayscale pictures: Gray = R * 0.299 + G * 0.587 + B * 0.114. Based on the maximum grayscale fusion rule, the calculation process is expressed by the formula: where F(x, y) represents the fused picture, A(x, y) represents the pixel value of the picture with the largest grayscale value among all the grayscale pictures, B(x, y) represents the pixel values of the other grayscale pictures except the largest grayscale value in the grayscale pictures, Gray A(x, y) represents the grayscale value corresponding to A(x, y), Gray B (x, y) represents the grayscale value corresponding to B(x, y).

[0016] Further, in step S3: Use the object detection model to detect the position box and color information of the indicator light on the fused picture; The object detection model is an object detection model trained using the indicator light pictures based on the yolov5 model; When training the object detection model, first collect N indicator light pictures to form indicator light data, then use the labelme tool to annotate the indicator light pictures. The annotated indicator light label categories include red lights, yellow lights, blue lights, green lights, and custom targets, such as custom white lights, and generate a dataset with pictures and labels corresponding one by one; Then use the annotated dataset to train and generate an indicator light detection model; The trained indicator light detection model can obtain the position box and color information of the indicator light in the input picture.

[0017] Further, in step S4: The indicator light flashing detection processes the corresponding image blocks extracted from the detected indicator light position box to obtain the flashing state and flashing frequency of the indicator light; The detection process includes the following steps:

[0018] S4-1. According to all M pictures to be monitored and time information obtained, and calculate the time interval between adjacent pictures as T through the time information;

[0019] S4-2. Represent the position of the detected indicator light with a box [x1, y1, x2, y2], where (x1, y1) represents the coordinate value of the upper left vertex of the box, and (x2, y2) represents the coordinate value of the lower right vertex of the box;

[0020] S4-3. According to the indicator light position box, extract the pixel blocks within the corresponding box on all pictures to be detected in chronological order, and obtain a picture set [img1, img2,..., img M for each indicator light, where the length of the picture set is M; img is the matrix representation of a 3-channel picture;

[0021] S4-4. Convert the picture set to grayscale pictures to obtain a grayscale picture set [gray1, gray2,..., gray M , where the length of the grayscale picture set is M; gray is the matrix representation of a single channel; and calculate the average grayscale of the pictures. The average grayscale value is represented by an array [mgray1, mgray2,..., mgray M , where W is the width of the position box, and H is the height of the position box;

[0022] S4-5. Calculate the eigenvalues of the average grayscale array [mgray1, mgray2,..., mgray M , and the calculated eigenvalues include the average value, variance, maximum value, and minimum value; among them, the average value is: The variance is: The maximum value is: Max = max([mgray1, mgray2,..., mgray M ); The minimum value is: Min = min([mgray1, mgray2,..., mgray M );

[0023] S4-6. Analyze the periodicity of the grayscale atlas according to the grayscale array and eigenvalues, and this periodicity represents the blinking state and frequency of the indicator light; The analysis strategy adopted is: First, judge whether the array fluctuates according to the numerical value of the set variance; if the variance of the array is greater than the set threshold, it is determined that the array fluctuates, that is, the indicator light is blinking; if the variance of the array is less than or equal to the set threshold, it is determined that the indicator light is not blinking; For the indicator lights determined to be blinking, the fast Fourier transform FFT method is used to calculate the indicator light period T2;

[0024] In addition to the above method for monitoring whether the indicator light blinks, there are two alternative methods that can be used for indicator light blink detection; Alternative one, it can be determined whether the indicator light blinks by whether the difference between the maximum value and the minimum value is greater than the established threshold; Alternative two, for the method of calculating the periodicity, the extreme point detection related method can also be used to calculate;

[0025] S4-7. According to the calculated period T2 and the picture shooting time interval T, obtain the blinking frequency of the indicator light The unit of the blinking frequency is expressed in Hz.

[0026] Furthermore, in step S5: Set the fault warning rule, and use the indicator light color and blinking frequency to limit the warning type in the warning rule. The warning rule is set as follows: a. The green light is always on: The system status is normal or there is a minor warning; b. The green light blinks, and the blinking frequency is F1: The HDM is initializing; c. The yellow light blinks, and the blinking frequency is F2: The system has a serious error warning; d. The red light blinks, and the blinking frequency is F3: The system has an emergency error warning; e. The blue light is always on: The blue light always on indicates that the system is normal; among them, F1, F2, and F3 respectively represent the blinking frequency thresholds set when the indicator lights of different colors blink; During the detection process, the system compares the detected indicator light color and the calculated blinking frequency F with the set warning rule to judge whether the device has a fault and issue a warning when a fault occurs.

[0027] Compared with the prior art, the beneficial effects of the present invention are:

[0028] The present invention provides a device fault detection method based on indicator light flashing detection. Compared with the previous fault detection or alarm methods based on indicator light detection, it is necessary to process videos or multiple pictures. In the process, picture fusion, indicator light detection, and flashing discrimination and frequency calculation need to be carried out successively. It can obtain information such as the color and flashing of the indicator light of the device to be detected, and then compare and analyze it with the set alarm rules for alarm. Compared with the previous fault detection methods, the present invention can obtain the flashing information of the device indicator light, and the basis for device fault detection or alarm provided thereby is more accurate, which can improve the accuracy of device fault detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a schematic flow diagram of indicator light flashing discrimination and frequency calculation for a device fault detection method based on indicator light flashing detection according to the present invention;

[0030] Figure 2 It is a schematic diagram of the front panel indicator light of the device to be monitored according to the present invention;

[0031] Figure 3 It is a schematic flow diagram of a PCA-based image fusion method for a device fault detection method based on indicator light flashing detection according to the present invention;

[0032] Figure 4 It is a schematic flow diagram of a pixel rule-based image fusion method for a device fault detection method based on indicator light flashing detection according to the present invention;

[0033] Figure 5 It is a schematic flow diagram of indicator light detection model training for a device fault detection method based on indicator light flashing detection according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0034] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0035] As Figures 1-5 shown, the present invention provides a technical solution, a device fault detection method based on indicator light flashing detection, including the following steps:

[0036] S1. Use a high-definition camera to collect videos or pictures of the device;

[0037] S2. In the pixel space, use an image fusion method to fuse the collected videos or pictures;

[0038] S3. Use the target detection model to detect the indicator lights on the fused image, and obtain the position boxes and color information of the indicator lights;

[0039] S4. According to the detected position box information of the indicator lights, extract the image blocks corresponding to the position boxes of the indicator lights on the image for processing, and obtain the flashing state and flashing frequency of the indicator lights;

[0040] S5. Perform fault judgment and alarm according to the detection results of the flashing frequency and flashing color of the indicator lights.

[0041] In step S1: Use a high-definition camera to obtain the image to be fused in two ways, namely, collecting the video or pictures of the device to be detected; if collecting video, collect the shooting frame rate of the video at the same time, and extract N1 pictures from the video at equal intervals according to a fixed sampling frequency S; if collecting pictures, at least two pictures should be collected, and record the picture shooting time and time interval.

[0042] In step S2: When performing image fusion, the image fusion algorithm based on PCA and the image fusion method based on pixel rules can be used to fuse the collected pictures into one picture.

[0043] When using the image fusion algorithm based on PCA, first, extract N2 pictures from the images to be fused, and convert the extracted pictures into grayscale images; the formula for converting the extracted pictures into grayscale images is: Gray = R * 0.299 + G * 0.587 + B * 0.114, where Gray is used to identify the grayscale pixel value at the point (x, y), R represents the pixel value of the R channel, G represents the pixel value of the G channel, and B represents the pixel value of the B channel; after converting the pictures into grayscale images, calculate the covariance matrix of the grayscale images, and the formula for calculating the covariance matrix is:

[0044]

[0045] Then, calculate the eigenvalues λ i and eigenvectors The eigenvalues and eigenvectors correspond one by one; sort the eigenvalues and eigenvectors in descending order of eigenvalues, and obtain the eigenvalue set as {λ1, λ1,..., λ n}, and the eigenvector set is And calculate the principal components of each grayscale value according to the formula where PC(x, y) represents the principal component of the grayscale image; G(x, y) represents the grayscale value at (x, y); based on the original images that were not extracted when extracting the pictures and converting them into grayscale images, perform inverse transformation on the principal components of the grayscale values, and the inverse transformation formula is: After the inverse transformation, a fused image is obtained.

[0046] If the pixel - rule - based image fusion method is adopted, first, extract the images to be fused from the collected videos or pictures, and convert all the images into grayscale images: Gray = R * 0.299+G * 0.587 + B * 0.114. Calculate based on the maximum - grayscale fusion rule. The calculation process is expressed by the formula: where F(x, y) represents the fused image, A(x, y) represents the pixel value of the image with the maximum grayscale value among all grayscale images, B(x, y) represents the pixel values of other grayscale images except the maximum grayscale value in the grayscale images, Gray A (x, y) represents the grayscale value corresponding to A(x, y), Gray B (x, y) represents the grayscale value corresponding to B(x, y).

[0047] In step S3: Use the object - detection model to detect the position box and color information of the indicator light on the fused image; the object - detection model is an object - detection model trained with indicator - light images based on the yolov5 model. When training the object - detection model, first collect N indicator - light images to form indicator - light data, and then use the labelme tool to annotate the indicator - light images. The annotated indicator - light label categories include red lights, yellow lights, blue lights, green lights, and custom targets, such as custom white lights, and generate a dataset with one - to - one correspondence between images and labels. Then use the annotated dataset to train and generate an indicator - light detection model. The trained indicator - light detection model can obtain the position box and color information of the indicator light in the input image.

[0048] In step S4: The indicator - light flicker detection processes the corresponding image blocks extracted from the detected indicator - light position box to obtain the flicker state and flicker frequency of the indicator light; the detection process includes the following steps:

[0049] S4 - 1: According to all M images to be monitored and time information obtained, calculate the time interval T between adjacent images through the time information;

[0050] S4 - 2: Represent the position of the detected indicator light with a box [x1, y1, x2, y2], where (x1, y1) represents the coordinate value of the upper - left vertex of the box, and (x2, y2) represents the coordinate value of the lower - right vertex of the box;

[0051] S4 - 3: According to the indicator - light position box, extract the pixel blocks within the corresponding box on all images to be detected in chronological order, and obtain a picture set [img1, img2,..., img M for each indicator light, where the length of the picture set is M; img is the matrix representation of a 3 - channel image;

[0052] S4-4. Convert the picture set into grayscale images to obtain a grayscale image set [gray1, gray2,..., gray M , where the length of the grayscale image set is M; gray is represented by a single-channel matrix; and calculate the average grayscale of the pictures. The average grayscale values are represented by an array [mgray1, mgray2,..., mgray M , where W is the width of the position box and H is the height of the position box;

[0053] S4-5. Calculate the eigenvalues of the average grayscale array [mgray1, mgray2,..., mgray M . The calculated eigenvalues include the mean, variance, maximum value, and minimum value; among them, the mean is: The variance is: The maximum value is: Max = max([mgray1, mgray2,..., mgray M ); The minimum value is: Min = min([mgray1, mgray2,..., mgray M );

[0054] S4-6. Analyze the periodicity of the grayscale image set based on the grayscale array and eigenvalues. This periodicity represents the blinking state and frequency of the indicator light; The analysis strategy adopted is: First, judge whether the array fluctuates according to the value of the set variance; If the variance of the array is greater than the set threshold, it is determined that the array fluctuates, that is, the indicator light is blinking; If the variance of the array is less than or equal to the set threshold, it is determined that the indicator light is not blinking; For the indicator lights determined to be blinking, the fast Fourier method is used to calculate the indicator light period T2;

[0055] In addition to the above method for monitoring whether the indicator light blinks, there are two alternative methods that can be used for indicator light blinking detection; Alternative method 1: It can be determined whether the indicator light blinks by whether the difference between the maximum value and the minimum value is greater than the established threshold; Alternative method 2: For the method of calculating periodicity, the extreme point detection related method can also be used to calculate;

[0056] S4-7. Finally, according to the calculated period T2 and the picture shooting time interval T, obtain the blinking frequency of the indicator light The unit of the blinking frequency is expressed in Hz.

[0057] In step S5: Set the fault warning rules. In the warning rules, the warning types are defined by the color and blinking frequency of the indicator light. The warning rules are set as follows: a. The green light is always on: The system status is normal or there is a minor warning; b. The green light blinks with a blinking frequency of F1: The HDM is initializing; c. The yellow light blinks with a blinking frequency of F2: A serious error warning occurs in the system; d. The red light blinks with a blinking frequency of F3: An emergency error warning occurs in the system; e. The blue light is always on: The blue light always on indicates that the system is normal; During the detection process, the system compares the detected color of the indicator light and the calculated blinking frequency F with the set warning rules to determine whether a fault has occurred in the device and issue a warning when a fault occurs.

[0058] Embodiment 1: If the yellow light blinks with a blinking frequency of 1 Hz: Then a serious error warning occurs in the system; If the red light blinks with a blinking frequency of 1 Hz: An emergency error warning occurs in the system.

[0059] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above-described exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be encompassed within the present invention. Any reference signs in the claims should not be regarded as limiting the claims involved.

Claims

1. A device fault detection method based on the detection of the flashing of an indicator light, characterized in that: It includes the following steps: S1. Use a high-definition camera to collect videos or pictures of the device; S2. In the pixel space, use an image fusion method to fuse the collected videos or pictures; S3. Use an object detection model to detect the indicator lights on the fused picture, and obtain the position box and color information of the indicator lights; S4. According to the detected indicator light position box information, extract the image block corresponding to the indicator light position box on the picture for processing, and obtain the blinking state and blinking frequency of the indicator light; S5. Make a fault judgment and give an alarm according to the detection results of the color and blinking frequency of the indicator light.

2. The device fault detection method based on indicator light flashing detection according to claim 1, characterized in that: In step S1: Use a high-definition camera to collect the pictures to be fused in two ways: videos or pictures of the device to be detected; If collecting a video, collect the shooting frame rate of the video at the same time, and extract N1 pictures from the video at equal intervals according to a fixed sampling frequency S; if collecting pictures, the number of collected pictures is at least two, and record the picture shooting time and time interval.

3. A device fault detection method based on indicator light flashing detection according to claim 1, characterized in that: In step S2: When performing image fusion, use the PCA-based image fusion algorithm and the pixel rule-based image fusion method to fuse the collected pictures into one picture.

4. The device fault detection method based on indicator light flashing detection according to claim 3, characterized in that: When using the PCA-based image fusion algorithm, first, extract N2 pictures from the pictures to be fused, and convert the extracted pictures into grayscale pictures; the formula for converting the extracted pictures into grayscale pictures is: Gray = R * 0.299 + G * 0.587 + B * 0.114, where Gray is used to identify the grayscale pixel value at the point (x, y), R represents the pixel value of the R channel, G represents the pixel value of the G channel, and B represents the pixel value of the B channel; after converting the pictures into grayscale pictures, calculate the covariance matrix of the grayscale pictures, and the formula for calculating the covariance matrix is: Then, calculate the eigenvalues λ of the covariance matrix i and eigenvectors The eigenvalues and eigenvectors are in one-to-one correspondence; sort the eigenvalues and eigenvectors in descending order of eigenvalues to obtain the set of eigenvalues as {λ1, λ1,..., λ n}, and the set of eigenvectors is and calculate the principal component of each gray value according to the formula where PC(x, y) represents the principal component of the grayscale image; G(x, y) represents the gray value at (x, y); based on the original image that was not extracted when the picture was converted to a grayscale image, perform an inverse transformation on the principal component of the gray value, and after the inverse transformation, a fused image is obtained.

5. The device fault detection method based on indicator light flashing detection according to claim 3, wherein: If the image fusion method based on pixel rules is adopted, first extract the pictures to be fused from the collected videos or pictures, and convert all the pictures into grayscale images: Gray = R * 0.299 + G * 0.587 + B * 0.

114. Calculate based on the maximum grayscale fusion rule, and the calculation process is expressed by the formula: Among them, F(x, y) represents the fused picture, A(x, y) represents the pixel value of the picture with the largest grayscale value among all grayscale images, B(x, y) represents the pixel values of other grayscale images except the largest grayscale value in the grayscale images, Gray(x, y) represents the grayscale value corresponding to A(x, y), and Gray B (x, y) represents the grayscale value corresponding to B(x, y).

6. The device fault detection method based on indicator light flashing detection according to claim 1, wherein: In step S3: Use an object detection model to detect the position box and color information of the indicator lights on the fused picture; the indicator light detection model is an object detection model trained using indicator light pictures based on the yolov5 model; when training the object detection model, first collect N indicator light pictures to form indicator light data, and then use the labelme tool to label the indicator light pictures. The labeled indicator light tag categories include red lights, yellow lights, blue lights, green lights, and custom targets, and generate a dataset with one-to-one correspondence between pictures and labels; then use the labeled dataset to train and generate an indicator light detection model.

7. A device fault detection method based on indicator light flashing detection according to claim 1, characterized in that: In step S4: The indicator light blinking detection processes the extracted image block corresponding to the detected indicator light position box to obtain the blinking state and blinking frequency of the indicator light; the detection process includes the following steps: S4-1. According to all the M pictures to be monitored and the time information obtained, calculate the time interval between adjacent pictures as T through the time information; S4-2. Represent the position of the detected indicator light with a box [x1, y1, x2, y2], where (x1, y1) represents the coordinate value of the upper left vertex of the box, and (x2, y2) represents the coordinate value of the lower right vertex of the box; S4-3. According to the position frame of the indicator light, extract the pixel blocks within the corresponding frame on all the pictures to be detected in chronological order, and obtain a picture set [img1, img2,..., img M for each indicator light, where the length of the picture set is M; img is the matrix representation of a 3-channel picture; S4-4. Convert the image set to grayscale images to obtain a grayscale image set [gray1, gray2,..., gray M , where the length of the grayscale image set is M; gray is represented by a single-channel matrix; and calculate the average grayscale of the images. The average grayscale value is represented by an array [mgray1, mgray2,..., mgray M , where W is the width of the position box, and H is the height of the position box; S4-5. Calculate the eigenvalues of the average grayscale array [mgray1, mgray2,..., mgray M , and the calculated eigenvalues include the average value, variance, maximum value, and minimum value; among them, the average value is: The variance is: The maximum value is: Max = max([mgray1, mgray2,..., mgray M ); The minimum value is: Min = min([mgray1, mgray2,..., mgray M ); S4-6. Analyze the periodicity of the grayscale image set based on the grayscale array and eigenvalue, where the periodicity represents the blinking state and frequency of the indicator light; the analysis strategy adopted is as follows: First, set a variance threshold, and compare the set variance value with the set variance threshold to determine whether the array fluctuates; if the array variance is greater than the set threshold, it is determined that the array fluctuates, that is, the indicator light is blinking; if the array variance is less than or equal to the set threshold, it is determined that the indicator light is not blinking; then for the indicator light determined to be blinking, use the fast Fourier method to calculate the indicator light period T2. S4-7. Obtain the blinking frequency of the indicator light based on the calculated period T2 and the picture taking time interval T 8. A device fault detection method based on indicator light flashing detection according to claim 1, characterized in that: In step S5: Set the fault warning rules, and use the indicator light color and blinking frequency to limit the warning type in the warning rules. The warning rules are set as follows: a. Green light is always on: The system status is normal or there is a minor warning; b. Green light blinks, and the blinking frequency is F1: HDM is initializing. c. Yellow blinks, and the blinking frequency is F2: A serious error warning occurs in the system. d. Red blinks, and the blinking frequency is F3: An emergency error warning occurs in the system; e. Blue light is always on: The blue light always on indicates that the system is normal. During the detection process, the system compares the detected indicator light color and the calculated blinking frequency F with the set warning rules to determine whether the device has a fault and issues a warning when a fault occurs.

Images