Device for hot spot sensing

Abstract

The present invention relates to a device for hotspot sensing, which includes an input unit, a processing unit, and an output unit. The input unit is configured to provide an image of an object having a hotspot to the processing unit, wherein image data of the image includes image data of the hotspot, and wherein the image is acquired by a camera. The processing unit is configured to determine the number of pixels in the image corresponding to the size of the hotspot. The processing unit is configured to determine the average temperature of the hotspot, and the determination includes: using pixel values ​​of pixels in the image corresponding to the size of the hotspot, and the number of pixels in the image corresponding to the size of the hotspot. The processing unit is configured to determine the surrounding temperature in the image, and the determination includes: using at least one pixel in the image that is different from the pixel in the image corresponding to the size of the hotspot. The processing unit is configured to determine the corrected temperature of the hotspot, and the determination includes: using the value of the correction factor, the average temperature of the hotspot, and the surrounding temperature.

Description

Technical Field

[0001] The present invention relates to an apparatus for hot spot sensing, a hot spot sensing system, a low voltage, medium voltage or high voltage switchgear comprising such a system, and a method for hot spot sensing. Background Art

[0002] Infrared (IR) thermography (IRT), thermal imaging, and thermal video are examples of infrared imaging science. Thermal imaging cameras typically detect radiation in the long infrared range of the electromagnetic spectrum (approximately 9,000 nanometers to 14,000 nanometers or 9 μm to 14 μm) and produce an image of that radiation, called a thermogram. Since all objects with a temperature above absolute zero emit IR radiation according to the blackbody radiation law, thermal imaging can see the environment with or without visible light illumination. The amount of radiation emitted by an object increases with increasing temperature; therefore, thermal imaging allows temperature changes to be seen and also enables the determination of absolute temperature based on assumptions about the emissivity of the object's surface.

[0003] Infrared thermography is widely used for non-contact temperature monitoring of hot spots in electrical equipment. High-resolution IR cameras are usually used for sporadic manual inspections and give sufficiently accurate absolute temperature values. Such high-resolution cameras are too expensive to be permanently installed for online temperature monitoring in electrical equipment (such as, for example, detecting critical hot spots to prevent overheating of switchgear). Low-resolution IR cameras of the price range can be used for this task, but the image quality is limited. In fact, the accuracy of the temperature of the measured hot spot is reduced and the accuracy of the size of the hot spot cannot be accurately determined.

[0004] This problem must be solved. Summary of the invention

[0005] It would therefore be advantageous to have improved techniques for determining the size and temperature of hot spots in electrical equipment such as low voltage, medium voltage, and high voltage switchgear.

[0006] The objects of the invention are solved by the subject-matter of the independent claims, wherein further embodiments are incorporated in the dependent claims.

[0007] In a first aspect, there is provided an apparatus for hotspot sensing, the apparatus comprising:

[0008] - input unit;

[0009] - a processing unit; and

[0010] - output unit;

[0011] The input unit is configured to provide an image of an object having a hotspot to the processing unit. The image data of the image includes image data of the hotspot, and the image is acquired by a camera. The processing unit is configured to determine the number of pixels in the image corresponding to the size of the hotspot. The processing unit is configured to determine a maximum temperature and an average temperature of the hotspot. Determining the average temperature includes: using pixel values ​​of pixels in the image corresponding to the size of the hotspot; and including: using the number of pixels in the image corresponding to the size of the hotspot. The processing unit is configured to determine an ambient temperature in the image. Determining the ambient temperature includes: using at least one pixel in the image that is different from the pixel in the image corresponding to the size of the hotspot. The processing unit is configured to determine a corrected temperature of the hotspot. Determining the corrected temperature includes: using a value of a correction factor; including: using an average temperature of the hotspot; and including using the ambient temperature.

[0012] In this way, a low-resolution camera can be used to monitor the hot spot temperature of an electrical device. In other words, the apparatus improves the accuracy of a low-resolution IR camera used to monitor the hot spot temperature of an electrical device.

[0013] In one example, the processing unit is configured to determine pixels in the image corresponding to a maximum temperature of the hotspot. The pixels in the image corresponding to the size of the hotspot include pixels in the image corresponding to the maximum temperature of the hotspot.

[0014] In one example, the processing unit is configured to determine pixels in the image corresponding to the size of the hotspot as pixels in the image having values ​​within a threshold range of values ​​of pixels determined to have a highest temperature of the hotspot.

[0015] In one example, determining the ambient temperature includes: utilizing a number of at least one surrounding pixel in the image adjacent to a pixel in the image corresponding to a size of the hot spot; and utilizing at least one pixel value of the at least one surrounding pixel.

[0016] In one example, determining the ambient temperature includes: utilizing a number of at least one surrounding pixel in the image that is not adjacent to a pixel in the image corresponding to a size of the hot spot; and utilizing at least one pixel value of the at least one surrounding pixel.

[0017] In one example, the correction factor has multiple possible values.The value of the correction factor used to determine the corrected temperature of the hot spot is determined as a function of the number of pixels in the image that correspond to the size of the hot spot.

[0018] In other words, a size-dependent correction algorithm for hotspots is provided, which enables reducing errors in hotspot temperature measurements even when using low-resolution infrared cameras.

[0019] In one example, multiple possible values ​​of the correction factor are determined based on multiple calibration images of the one or more hot spots acquired by the camera and / or one or more cameras substantially the same as the camera, and based on one or more measured temperatures or reference temperatures of the one or more hot spots.

[0020] In one example, a thermocouple is used to provide the measured temperature.

[0021] In one example, determining the multiple possible values ​​of the correction factor includes determining, for each calibration image in a plurality of calibration images, a number of pixels in the calibration image that correspond to the size of the hot spot. Determining the multiple possible values ​​of the correction factor also includes determining, for each calibration image in the plurality of calibration images, an average temperature of the hot spot in the calibration image, the determining including: using pixel values ​​of pixels in the calibration image that correspond to the size of the hot spot and the number of pixels in the calibration image that correspond to the size of the hot spot. Determining the multiple possible values ​​of the correction factor also includes determining, for each calibration image in the plurality of calibration images, an ambient temperature in the calibration image, the determining including: using at least one pixel in the calibration image other than the pixel in the calibration image that corresponds to the size of the hot spot.

[0022] In one example, the pixels in the calibration image corresponding to the size of the hot spot include pixels in the calibration image corresponding to the highest temperature of the hot spot. The pixels in the calibration image corresponding to the size of the hot spot are determined to be pixels in the calibration image having values ​​within a threshold range of values ​​of pixels determined to have the highest temperature of the hot spot in the calibration image.

[0023] In one example, determining a plurality of possible values ​​of the correction factor includes determining an error value for each calibration image in a plurality of calibration images. The determining includes: calculating a first value as a measured temperature or a reference temperature of the hot spot subtracted from an average temperature of the hot spot in the calibration image; and includes: calculating a second value as a surrounding temperature of the hot spot in the calibration image subtracted from an average temperature of the hot spot in the calibration image. The error value is then determined as a ratio of the first value to the second value.

[0024] In one example, determining the corrected temperature includes: multiplying a correction factor by a difference between an average temperature of the hot spot in the image and an ambient temperature in the image; and includes: subtracting the resulting product from the average temperature of the hot spot in the image.

[0025] In one example, after the processing unit receives the image from the input unit and before any further processing, the processing unit is configured to transform the image into an interpolated image for further processing.

[0026] Thus, the temperature error can be further reduced by increasing the resolution of the original image through interpolation methods before applying the correction.

[0027] In a second aspect, there is provided a hotspot sensing system, the system comprising:

[0028] - an apparatus according to the first aspect; and

[0029] -camera.

[0030] The camera is configured to acquire an image of an object having a hot spot.

[0031] In a third aspect, a low voltage, medium voltage or high voltage switchgear is provided, comprising at least one system according to the second aspect. The object having the hot spot is a part of the switchgear.

[0032] In a fourth aspect, there is provided a method for hotspot sensing, the method comprising:

[0033] b) providing an image of the object having the hotspot to a processing unit, wherein image data of the image includes image data of the hotspot, and wherein the image is acquired by a camera;

[0034] e) determining, by the processing unit, the number of pixels in the image corresponding to the size of the hotspot;

[0035] f) determining, by a processing unit, an average temperature of the hot spot, the determining comprising: using pixel values ​​of pixels in the image corresponding to the size of the hot spot and the number of pixels in the image corresponding to the size of the hot spot;

[0036] g) determining, by a processing unit, an ambient temperature in the image, the determining comprising: utilizing at least one pixel in the image that is different from a pixel in the image that corresponds to a size of the hot spot;

[0037] h) determining, by the processing unit, a corrected temperature of the hot spot, the determining comprising: using the value of the correction factor, the average temperature of the hot spot, and the ambient temperature.

[0038] In one example, the method comprises step d): determining, by a processing unit, pixels in the image corresponding to the maximum temperature of the hotspot, and wherein in step e), the pixels in the image corresponding to the size of the hotspot include pixels in the image corresponding to the maximum temperature of the hotspot.

[0039] In one example, step e) includes: the processing unit determining pixels in the image corresponding to the size of the hotspot as pixels in the image having values ​​within a threshold range of values ​​of pixels determined to have the highest temperature value of the hotspot.

[0040] In one example, step g) includes: utilizing at least one number of surrounding pixels in the image adjacent to a pixel in the image corresponding to a size of the hotspot and at least one pixel value of the at least one surrounding pixel.

[0041] In one example, step g) includes: utilizing the number of at least one surrounding pixel in the image that is not adjacent to the pixel in the image corresponding to the size of the hotspot and at least one pixel value of the at least one surrounding pixel.

[0042] In one example, in step h), the correction factor has a plurality of possible values, and wherein step h) comprises determining the value of the correction factor as a function of the number of pixels in the image corresponding to the size of the hotspot.

[0043] In one example, the method includes step a): determining multiple possible values ​​of the correction factor based on multiple calibration images of one or more hot spots acquired by the camera and / or one or more cameras substantially identical to the camera, and based on one or more measured temperatures or reference temperatures of the one or more hot spots.

[0044] In one example, a thermocouple is used to provide the measured temperature.

[0045] In one example, step a) includes: for each calibration image in a plurality of calibration images: determining the number of pixels in the calibration image corresponding to the size of the hotspot; determining the average temperature of the hotspot in the calibration image, including using pixel values ​​of pixels in the calibration image corresponding to the size of the hotspot and the number of pixels in the calibration image corresponding to the size of the hotspot; and determining the ambient temperature in the calibration image, including using at least one pixel in the calibration image that is different from the pixel in the calibration image corresponding to the hotspot.

[0046] In one example, in step a), the pixels in the calibration image corresponding to the size of the hotspot include pixels corresponding to the highest temperature of the hotspot in the calibration image, and wherein step a) includes determining the pixels in the calibration image corresponding to the size of the hotspot as pixels in the calibration image whose values ​​are within a threshold range of the values ​​of the pixels determined to have the highest temperature of the hotspot in the calibration image.

[0047] In one example, step a) includes: determining an error value for each calibration image of a plurality of calibration images, the determination including: calculating a first value as an average temperature of the hot spot in the calibration image minus a measured temperature or a reference temperature of the hot spot; and calculating a second value as an average temperature of the hot spot in the calibration image minus a surrounding temperature of the hot spot in the calibration image, and wherein step a) includes: determining the error value as a ratio of the first value to the second value.

[0048] In one example, step h) comprises: multiplying the correction factor by the difference between the average temperature of the hot spot in the image and the surrounding temperature in the image, and subtracting the difference from the average temperature of the hot spot in the image.

[0049] In one example, the method comprises step c): transforming the image into an interpolated image by a processing unit for further processing.

[0050] The above aspects and examples will be apparent from and elucidated with reference to the embodiments described hereinafter. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Exemplary embodiments are described below with reference to the following drawings:

[0052] Figure 1 A detailed workflow of an example utilizing an apparatus, system, and method for hotspot sensing is shown;

[0053] Figure 2 An exemplary ideal hotspot is shown;

[0054] FIG3 shows an example of hotspot determination for an ideal case in a) and a non-ideal case in b);

[0055] Figure 4 The evaluation pathway for ambient temperature is shown;

[0056] Figure 5 The evaluation pathway for ambient temperature is shown;

[0057] Figure 6 An example of a correction function as a function of the number of pixels in a hotspot is shown;

[0058] Figure 7 An example of spline interpolation of non-ideal hot spots is shown; and

[0059] Figure 8 Shows Figure 7 Example of hotspot determination for a spline interpolated image. DETAILED DESCRIPTION

[0060] Figures 1 to 8The invention relates to the operation of an apparatus, system and method for hotspot sensing. In one example, an apparatus for hotspot sensing is provided. The apparatus includes an input unit, a processing unit, and an output unit. The input unit is configured to provide an image of an object having a hotspot to the processing unit. The image data of the image includes the image data of the hotspot, and the image is acquired by a camera. The processing unit is configured to determine the number of pixels in the image corresponding to the size of the hotspot. The processing unit is configured to determine the maximum temperature and the average temperature of the hotspot. Determining the average temperature includes: using the pixel value of the pixel corresponding to the size of the hotspot in the image; and including: using the number of pixels corresponding to the size of the hotspot in the image. The processing unit is configured to determine the ambient temperature in the image. Determining the ambient temperature includes: using at least one pixel in the image that is different from the pixel corresponding to the size of the hotspot in the image. The processing unit is configured to determine the corrected temperature of the hotspot. Determining the corrected temperature includes: using the value of the correction factor; including: using the average temperature of the hotspot; and including using the ambient temperature.

[0061] The device has utility in applications where an infrared camera measurement device can be installed to measure the temperature of a hot spot. Examples are for low voltage, medium voltage and high voltage switchgear.

[0062] In one example, the camera used to acquire the image is a low resolution camera. In one example, the camera has a sensor with 32×32 pixels. In one example, the camera has a sensor with 128×128 pixels. In one example, the camera has a sensor with 256×256 pixels. In one example, the camera has a sensor with 512×512 pixels. In one example, the camera has a sensor with a phase of 1024×1024.

[0063] In one example, the camera used to acquire the image operates in the long infrared range. In one example, the camera used to acquire the image operates in the 9 μm to 14 μm range.

[0064] According to an example, the processing unit is configured to determine pixels in the image corresponding to a maximum temperature of the hotspot. The pixels in the image corresponding to the size of the hotspot include pixels in the image corresponding to the maximum temperature of the hotspot.

[0065] According to an example, the processing unit is configured to determine pixels in the image corresponding to the size of the hotspot as pixels in the image having values ​​within a threshold range of values ​​of pixels determined to have the highest temperature of the hotspot.

[0066] According to one example, determining the ambient temperature includes: utilizing the number of at least one surrounding pixel in the image adjacent to a pixel in the image corresponding to a size of the hot spot; and utilizing at least one pixel value of the at least one surrounding pixel.

[0067] According to one example, determining the ambient temperature includes: utilizing the number of at least one surrounding pixel in the image that is not adjacent to a pixel in the image corresponding to a size of the hot spot; and utilizing at least one pixel value of the at least one surrounding pixel.

[0068] According to one example, the correction factor has a plurality of possible values.The value of the correction factor used to determine the corrected temperature of the hotspot is determined as a function of the number of pixels in the image corresponding to the size of the hotspot.

[0069] According to one example, multiple possible values ​​of the correction factor are determined based on multiple calibration images of the one or more hot spots acquired by the camera and / or one or more cameras substantially identical to the camera, and based on one or more measured temperatures or reference temperatures of the one or more hot spots.

[0070] According to one example, a thermocouple is used to provide the measured temperature.

[0071] According to one example, determining multiple possible values ​​of the correction factor includes, for each calibration image in a plurality of calibration images, determining a number of pixels in the calibration image that correspond to the size of the hotspot. Determining multiple possible values ​​of the correction factor also includes, for each calibration image in the plurality of calibration images, determining an average temperature of the hotspot in the calibration image, the determining including: using pixel values ​​of pixels in the calibration image that correspond to the size of the hotspot and the number of pixels in the calibration image that correspond to the size of the hotspot. Determining multiple possible values ​​of the correction factor also includes, for each calibration image in the plurality of calibration images, determining an ambient temperature in the calibration image, the determining including: using at least one pixel in the calibration image that is different from the pixel in the calibration image that corresponds to the size of the hotspot.

[0072] According to one example, the pixels in the calibration image corresponding to the size of the hot spot include pixels corresponding to the highest temperature of the hot spot in the calibration image. The pixels in the calibration image corresponding to the size of the hot spot are determined to be pixels in the calibration image whose values ​​are within a threshold range of the values ​​of the pixels determined to have the highest temperature of the hot spot in the calibration image.

[0073] According to one example, determining a plurality of possible values ​​of the correction factor includes determining an error value for each calibration image in a plurality of calibration images. The determining includes: calculating a first value as a measured temperature or a reference temperature of the hot spot minus an average temperature of the hot spot in the calibration image; and including: calculating a second value as a surrounding temperature of the hot spot in the calibration image minus an average temperature of the hot spot in the calibration image. The error value is then determined as a ratio of the first value to the second value.

[0074] According to one example, determining the corrected temperature includes: multiplying the correction factor by the difference between the average temperature of the hot spot in the image and the surrounding temperature in the image; and includes: subtracting the resulting product from the average temperature of the hot spot in the image.

[0075] According to an example, after the processing unit receives the image from the input unit and before any further processing, the processing unit is configured to transform the image into an interpolated image for further processing.

[0076] Thus, the device may operate in an offline mode, wherein the camera acquires images which are then transmitted to a device that analyzes the images to determine information related to hot spots as described above.

[0077] However, the device may be linked to a camera to form a system which can acquire images in real time and analyse the images to provide accurate quantification of hot spots in electrical equipment such as switchgear in real time.

[0078] As described above, the workflow is related to hotspot quantification. Therefore, in one example, a method for hotspot sensing is provided, which includes in its basic steps:

[0079] b) providing an image of the object having the hotspot to a processing unit, wherein image data of the image includes image data of the hotspot, and wherein the image is acquired by a camera;

[0080] e) determining, by the processing unit, the number of pixels in the image corresponding to the size of the hotspot;

[0081] f) determining, by a processing unit, an average temperature of the hot spot, the determining comprising: using pixel values ​​of pixels in the image corresponding to the size of the hot spot and the number of pixels in the image corresponding to the size of the hot spot;

[0082] g) determining, by a processing unit, an ambient temperature in the image, the determining comprising: utilizing at least one pixel in the image that is different from a pixel in the image that corresponds to a size of the hot spot;

[0083] h) determining, by the processing unit, a corrected temperature of the hot spot, the determining comprising: using the value of the correction factor, the average temperature of the hot spot, and the ambient temperature.

[0084] According to one example, the method comprises step d): determining, by a processing unit, pixels in the image corresponding to the maximum temperature of the hotspot, and wherein in step e), the pixels in the image corresponding to the size of the hotspot include pixels in the image corresponding to the maximum temperature of the hotspot.

[0085] According to an example, step e) comprises: the processing unit determines pixels in the image corresponding to the size of the hotspot as pixels in the image whose values ​​are within a threshold range of values ​​of pixels determined to have the highest temperature value of the hotspot.

[0086] According to one example, step g) comprises: utilizing the number of at least one surrounding pixel in the image adjacent to the pixel in the image corresponding to the size of the hotspot and at least one pixel value of the at least one surrounding pixel.

[0087] According to one example, step g) comprises: utilizing the number of at least one surrounding pixel in the image that is not adjacent to the pixel in the image corresponding to the size of the hotspot and at least one pixel value of the at least one surrounding pixel.

[0088] According to an example, in step h), the correction factor has a plurality of possible values, and wherein step h) comprises determining the value of the correction factor as a function of the number of pixels in the image corresponding to the size of the hotspot.

[0089] According to one example, the method includes step a): determining multiple possible values ​​of the correction factor based on multiple calibration images of one or more hot spots acquired by a camera and / or one or more cameras substantially identical to the camera, and based on one or more measured temperatures or reference temperatures of the one or more hot spots.

[0090] According to one example, a thermocouple is used to provide the measured temperature.

[0091] According to one example, step a) includes: for each calibration image in a plurality of calibration images: determining the number of pixels in the calibration image corresponding to the size of the hotspot; determining the average temperature of the hotspot in the calibration image, including using pixel values ​​of pixels in the calibration image corresponding to the size of the hotspot, and the number of pixels in the calibration image corresponding to the size of the hotspot; and determining the ambient temperature in the calibration image, including using at least one pixel in the calibration image that is different from the pixel in the calibration image corresponding to the hotspot.

[0092] According to one example, in step a), the pixels corresponding to the size of the hotspot in the calibration image include pixels corresponding to the highest temperature of the hotspot in the calibration image, and wherein step a) includes determining the pixels corresponding to the size of the hotspot in the calibration image as pixels in the calibration image whose values ​​are within a threshold range of the values ​​of the pixels determined to have the highest temperature of the hotspot in the calibration image.

[0093] According to one example, step a) includes: determining an error value for each calibration image of a plurality of calibration images, the determination including: calculating a first value as an average temperature of the hot spot in the calibration image minus a measured temperature or a reference temperature of the hot spot; and calculating a second value as an average temperature of the hot spot in the calibration image minus a surrounding temperature of the hot spot in the calibration image, and wherein step a) includes: determining the error value as a ratio of the first value to the second value.

[0094] According to one example, step h) comprises: multiplying the correction factor by the difference between the average temperature of the hot spot in the image and the surrounding temperature in the image, and subtracting the difference from the average temperature of the hot spot in the image.

[0095] According to an example, the method comprises the step c) transforming the image into an interpolated image by a processing unit for further processing.

[0096] Thus, the apparatus, systems, and methods described above solve the problem of accurately determining precise absolute temperature values ​​for small hot spots detected by a low-resolution IR camera, which is otherwise not possible using a low-resolution camera but would require a high-resolution and expensive IR camera.

[0097] The device, system and method overcome the following problems

[0098] [1] Lower resolution makes it more difficult to distinguish hotspot temperatures from background temperatures. The hotspot pixel with the highest measured temperature can represent the background temperature to some extent. This effect is more pronounced the greater the temperature difference between the hotspot and the background. Therefore, the accuracy of low-resolution IR cameras is further reduced, especially for small hotspots near a cold background.

[0099] [2] Using a wide-angle lens results in a fisheye effect in low-resolution IR cameras, which means that objects of the same size at the same temperature are resolved with a varying number of pixels depending on their position in the image. The further the object is from the center, the less accurate the measured temperature becomes due to the reduced number of pixels.

[0100] [1] and [2] are characteristics of low resolution IR cameras that can usually only be mitigated by selecting a high quality high resolution IR camera, which has made permanent installation unattractive until now. However, the technology currently described enables low resolution cameras to be effectively used in permanently installed locations and cost-effectively used therein for monitoring electrical equipment.

[0101] As discussed above, the above problems have been solved by using a correction algorithm that compensates for the temperature error of the hotspot during the image processing step. The correction algorithm enables improving the temperature accuracy from the low-resolution IR camera, and the accuracy is further improved by applying an interpolation method.

[0102] Now, refer again Figures 1 to 8 Additional details of specific embodiments are described.

[0103] Temperature Correction Algorithm - Overview

[0104] In the case where there are several hotspots in the global image, the partitioning algorithm divides the measured image into its components. Each component represents an independent hotspot. In the following, the solution is explained for one hotspot.

[0105] The temperature correction algorithm consists of four steps, such as Figure 1 As shown in the flowchart:

[0106] [1] Detect the image pixel with the highest temperature. The algorithm finds the image pixel showing the highest absolute temperature value.

[0107] [2] Determine the size and temperature of the hotspot. The size of the hotspot is determined by counting the number of image pixels that meet predefined hotspot conditions near the maximum temperature. The mean of all detected pixels gives the hotspot temperature.

[0108] [3] Determine the ambient temperature / background temperature. The higher the difference between the ambient / background and hotspot temperatures, the more the algorithm compensates.

[0109] [4] Correction function

[0110] (a) Derivation - Measurements are performed to estimate the temperature error associated with the hotspot size. The correction function is derived by curve fitting the error.

[0111] (b) Application - The correction function of the algorithm gives as output the corrected hotspot temperature after steps [1] to [3].

[0112] Interpolation Methods - Overview

[0113] Interpolation methods can be applied to the original image to achieve better spatial resolution of hotspot areas and sizes.

[0114] The solution is designed for online temperature monitoring of hot spots using a low-resolution IR camera, where the algorithm can be implemented in the sensor’s microcontroller or in a central data aggregation unit.

[0115] Temperature Sensing - More Details

[0116] Figure 2The ideal hotspot is shown. The pixel with the highest temperature is in the middle of the image, but the hotspot definition also includes the neighboring area of ​​the pixel. The hotspot effect is obvious in the transition zone, while the background is not affected. It has been established that for the number of hotspot pixels N ≥ 9, the maximum temperature is very consistent with the actual temperature, where N is the total number of detected hotspot pixels. Therefore, in the derived correction function, when N ≥ 9, the number of corrections is reduced. Figure 2 In the example, the pixel representing the highest temperature is surrounded by pixels with approximately the same temperature. This is the ideal situation. The distribution of hotspot pixels can also be taken into account for error compensation, but for simplicity, this can be ignored. Therefore, in the following, in addition to providing a compensation function that varies as a function of the number of pixels in the hotspot, a compensation function that takes the distribution of hotspot pixels into account can also be determined through additional experimental data.

[0117] Figure 3 schematically outlines the determination of hot spots in ideal and non-ideal situations:

[0118] 1. The first step is to identify the highest temperature of the complete image as T max In this paper, the full image only represents the hotspots, which, as mentioned before, can be separated from the larger image.

[0119] 2. Define hotspot conditions to detect all image pixels belonging to the hotspot, such as

[0120] T HS，i >(1-p)T max

[0121] The small parameter p gives the maximum temperature T max Maximum deviation of. Temperature T HS,i is the hotspot pixel with index i. In medium voltage (MV) switchgear, a typical value of p is about 0.05. This means that max All pixels with a maximum 5% difference enclose the connected hotspot. In an ideal case, with T max The pixel is located at the center of the hotspot, since it is most likely to be unaffected by the surroundings. However, if T max If one of the pixel's neighbor pixels does not meet the hot spot condition, temperature correction is used to obtain a more accurate temperature measurement.

[0122] 3. By applying the hotspot condition, all hotspot pixels T can be detected HS,i , where i = 1...N, where N represents the total number of hotspot pixels. Then, the hotspot temperature T is obtained based on the average temperature of all detected pixels. HS :

[0123]

[0124] Where T HS ≤T max Established.

[0125] 4. Finally, the number N and the average temperature T HS The size and temperature of the hot spot are described.

[0126] For the evaluation of the ambient temperature, two cases are distinguished. The first approach takes into account the background temperature which is not affected by the hot spots. The second approach focuses on the temperature in the direct (or adjacent) area surrounding the detected pixel with the highest temperature.

[0127] Figure 4 The evaluation of the background temperature is schematically depicted:

[0128] 1. The background temperature is represented by image pixels that are close to the hotspot but not affected. Therefore, the background region encloses the connected boundaries around the hotspot and is used to determine the background temperature.

[0129] 2. The average temperature of the background T BG Given by:

[0130]

[0131] Where N BG is the number of all detected pixels, T BG,i It refers to the temperature of pixel i in the background area.

[0132] 3. Average background temperature T BG Used in the correction function to take into account the effect on the hot spot temperature.

[0133] It has been established that the influence of the ambient temperature can also be estimated close to the hot spot. In this approach, the ambient temperature is affected by the hot spot, and the difference from the highest measured temperature gives a quantity related to the accuracy of the hot spot temperature measurement. If the difference becomes zero, the ambient temperature of the hot spot is increasingly coinciding with the highest temperature, and the error becomes negligible. Figure 5 Qualitatively shows the surrounding temperature T of the hot spot HS,s Determination of:

[0134] 1. The surrounding temperature T of the eight image pixels around the pixel with the highest temperature that produces the hot spot HS,s .

[0135] 2. The surrounding temperature of the hot spot T HS,s It can be determined by the following formula:

[0136]

[0137] Where T HS,si is the temperature of pixel i in the neighboring area of ​​pixel i with the highest temperature.

[0138] 3. In the correction function, the surrounding temperature T of the hot spot HS,s Indicates the amount of correction, which is required due to the difference in maximum temperature from its immediate surroundings.

[0139] The core of the algorithm is the correction functions which describe the hotspot size-dependent errors of the temperature measured by the infrared camera. They can be derived from reference measurements or basic measurements, in which thermocouples are additionally installed for comparison purposes. For the purpose of a generally valid correction function, it is convenient to additionally vary the object (hotspot) and background temperatures:

[0140] 1. Maximum temperature T max Or hot spot temperature T HS Can be used to measure the hot spot temperature T HS,meas , to derive the correction function. Note that the temperature correction algorithm should be applied to select the same measured temperature.

[0141] 2. Actual temperature T HS,real The measured temperature T HS,meas The error can be written as:

[0142] T HS,meas -T HS，real =f corr (N)(T HS，meas -T SR )

[0143] where f corr The correction function depending on the number of hotspot pixels N is described. The correction function f corr Additionally multiply by T HS,meas and T SR The difference is calculated to reflect the effect of the surrounding temperature on the hot spot temperature.

[0144] 3. Two approaches are introduced to consider the surroundings of the hotspot, which lead to different correction functions fcorr. Surrounding temperature T SR It can be expressed as the background temperature T SR =T BG Or the temperature around the hot spot T SR =T HS,s .

[0145] However, the background temperature can be controlled in the measurements used to derive the correction function. The resulting hotspot surrounding temperature implicitly depends on the background temperature.

[0146] 4. By evaluating the measurements, the correction function f can be plotted for the number N of different hotspot and surrounding temperatures corr The curve fitting is performed to have a global mathematical function that reveals the best fit of the measurement error of N for the considered range of hot spots and the surrounding temperature range, see Figure 6 ,Should Figure 6 The measurement error is shown versus the number of hotspot pixels N and the correction function from the curve fit (referred to herein as the linear function). The different types of points represent measurements with different hotspot and background temperatures.

[0147] The final step in the temperature correction algorithm is to calculate the corrected temperature:

[0148] As explained above, the correction function f may be provided corr , the correction function is actually generated by the curve fitting of the error. If the measured hot spot T HS,meas and ambient temperature T SR And the number of hotspot pixels N is known from the previous step, then the corrected temperature can be directly calculated according to the following formula:

[0149] T corr =T HS，meas -f corr (N)(T HS，meas -T SR ).

[0150] Thus, in this way, the temperature correction technique described above allows to improve the accuracy of hotspot temperature detection at very low cost, since only a simple algorithm needs to be implemented in the sensor's microcontroller or in a central data aggregation unit. Neither additional complex optics nor high quality IR cameras with better or higher resolution are required.

[0151] Interpolation methods can be used to improve image quality. Figure 7 An exemplary spline interpolation of a non-ideal hot spot is shown.

[0152] like Figure 8 As outlined, interpolation increases the resolution of an image, which is similar to Figure 2 This improves the characterization of the size and temperature of the hotspot as well as the surrounding / background temperature compared to the original image in b).

[0153] Interpolation methods are applied for deriving the correction function as well as for the temperature correction algorithm itself. In both cases, interpolation is performed beforehand and all further steps are performed using the better quality interpolated image.

[0154] Interpolation produces a smooth linear function between the data points in each pixel of the original image. In practice, non-adjacencies caused by, for example, a hot cable passing through a cold background can only be captured by the thermal image at the boundaries of the original image pixels. In general, interpolation does not allow detection of the exact boundaries of non-adjacencies in the original image.

[0155] The main benefit is that the interpolation produces more accurate results. In particular, the temperature correction algorithm becomes less sensitive to changes in the number of hotspot pixels, thereby reducing oscillatory behavior during online temperature monitoring.

[0156] Although the present invention has been described and illustrated in detail in the drawings and the foregoing description, such description and illustration are to be considered illustrative or exemplary rather than restrictive. The present invention is not limited to the disclosed embodiments. Other variations to the disclosed embodiments may be understood and effected by those skilled in the art in practicing the claimed invention by studying the drawings, the disclosure, and the appended claims.

Claims

1. A device for hotspot sensing, the device comprising: - input unit; - Processing unit; as well as - output unit; wherein the input unit is configured to provide an image of an object having a hotspot to the processing unit, wherein the image data of the image includes image data of the hotspot, and wherein the image is acquired by a camera; wherein the processing unit is configured to determine the number of pixels in the image corresponding to the size of the hotspot; wherein the processing unit is configured to determine an average temperature of the hot spot, the determining comprising: utilizing pixel values ​​of the pixels in the image corresponding to the size of the hot spot, and the number of the pixels in the image corresponding to the size of the hot spot; wherein the processing unit is configured to determine the ambient temperature in the image, the determining comprising: utilizing at least one pixel in the image that is different from the pixel in the image corresponding to the size of the hot spot; and wherein the processing unit is configured to determine a corrected temperature of the hot spot, the determination comprising: utilizing a value of a correction factor, the average temperature of the hot spot, and the surrounding temperature, and wherein the determination of the corrected temperature comprises: multiplying the correction factor by a difference between the average temperature of the hot spot in the image and the surrounding temperature in the image, and subtracting the resulting product from the average temperature of the hot spot in the image.

2. The apparatus according to claim 1, wherein the processing unit is configured to determine pixels in the image corresponding to a maximum temperature of the hot spot, and wherein the pixels in the image corresponding to the size of the hot spot include: The pixel in the image corresponding to the highest temperature of the hot spot.

3. An apparatus according to claim 2, wherein the processing unit is configured to determine the pixel in the image corresponding to the size of the hotspot as: the pixel in the image whose value is within a threshold range of the value of the pixel determined to have the highest temperature of the hotspot.

4. The device according to any one of claims 1 to 3, wherein the determination of the ambient temperature comprises: The number of at least one surrounding pixel in the image adjacent to the pixel in the image corresponding to the size of the hot spot and at least one pixel value of the at least one surrounding pixel are utilized.

5. The device according to any one of claims 1 to 3, wherein the determination of the ambient temperature comprises: The number of at least one surrounding pixel in the image that is not adjacent to the pixel in the image corresponding to the size of the hot spot and at least one pixel value of the at least one surrounding pixel are utilized.

6. An apparatus according to any one of claims 1 to 3, wherein the correction factor has multiple possible values, and wherein the value of the correction factor used to determine the corrected temperature of the hotspot is determined as a function of the number of pixels in the image corresponding to the size of the hotspot.

7. An apparatus according to claim 6, wherein the multiple possible values ​​of the correction factor are determined based on multiple calibration images of one or more hot spots acquired by the camera and / or one or more cameras substantially identical to the camera, and based on one or more measured temperatures or reference temperatures of the one or more hot spots.

8. The apparatus of claim 7, wherein a thermocouple is used to provide the measured temperature.

9. The apparatus according to any one of claims 7 to 8, wherein the determination of the plurality of possible values ​​of the correction factor comprises: For each calibration image in the plurality of calibration images: determining a number of pixels in the calibration image that corresponds to a size of the hotspot; Determining the average temperature of the hot spot in the calibration image includes utilizing pixel values ​​of the pixels in the calibration image corresponding to the size of the hot spot and the number of pixels in the calibration image corresponding to the size of the hot spot; and determining the ambient temperature in the calibration image includes utilizing at least one pixel in the calibration image that is different from the pixel in the calibration image corresponding to the size of the hot spot.

10. The apparatus of claim 9, wherein the pixels in the calibration image corresponding to the size of the hotspot include: a pixel in the calibration image corresponding to the maximum temperature of the hotspot, and wherein the pixel in the calibration image corresponding to the size of the hotspot is determined to be: the pixel in the calibration image whose value is within a threshold range of the value of the pixel determined to have the maximum temperature of the hotspot in the calibration image.

11. The apparatus according to any one of claims 8 or 10, wherein the determination of the plurality of possible values ​​of the correction factor comprises: For each calibration image of the plurality of calibration images, an error value is determined, the determination comprising calculating a first value as the measured temperature or the reference temperature of the hot spot subtracted from the average temperature of the hot spot in the calibration image, and calculating a second value as the ambient temperature of the hot spot in the calibration image subtracted from the average temperature of the hot spot in the calibration image, and wherein the error value is determined as a ratio of the first value to the second value.

12. An apparatus according to any one of claims 1 to 3, 7 to 8 and 10, wherein after the processing unit receives the image from the input unit and before any further processing, the processing unit is configured to transform the image into an interpolated image for further processing.

13. A system for hotspot sensing, the system comprising: - a device according to any one of claims 1 to 12; as well as -camera; Wherein the camera is configured to acquire the image of the object having a hot spot.

14. A low voltage, medium voltage or high voltage switchgear comprising at least one system according to claim 13, wherein the object having a hot spot is a part of the switchgear.

15. A method for hotspot sensing, the method comprising: b) providing an image of the object having the hotspot to a processing unit, wherein image data of the image includes image data of the hotspot, and wherein the image is acquired by a camera; e) determining, by the processing unit, the number of pixels in the image corresponding to the size of the hotspot; f) determining, by the processing unit, an average temperature of the hot spot, the determining comprising: using pixel values ​​of the pixels in the image corresponding to the size of the hot spot and the number of the pixels in the image corresponding to the size of the hot spot; g) determining, by the processing unit, the ambient temperature in the image, the determining comprising: utilizing at least one pixel in the image that is different from the pixel in the image corresponding to the size of the hot spot; and h) determining, by the processing unit, a corrected temperature of the hot spot, the determination comprising using a value of a correction factor, the average temperature of the hot spot, and the ambient temperature, and wherein step h) comprises multiplying the correction factor by a difference between the average temperature of the hot spot in the image and the ambient temperature in the image, and subtracting the resulting product from the average temperature of the hot spot in the image.

16. The method according to claim 15, wherein the method comprises step d): determining, by the processing unit, pixels in the image corresponding to the highest temperature of the hot spot, and wherein in step e), the pixels in the image corresponding to the size of the hot spot comprise: The pixel in the image corresponding to the highest temperature of the hot spot.

17. The method according to claim 16, wherein step e) comprises: The processing unit determines the pixels in the image corresponding to the size of the hot spot as pixels in the image having values ​​within a threshold range of values ​​of the pixels determined to have the highest temperature value of the hot spot.

18. The method according to any one of claims 15 to 17, wherein step g) comprises: The number of at least one surrounding pixel in the image adjacent to the pixel in the image corresponding to the size of the hot spot and at least one pixel value of the at least one surrounding pixel are utilized.

19. The method according to any one of claims 15 to 17, wherein step g) comprises: The number of at least one surrounding pixel in the image that is not adjacent to the pixel in the image corresponding to the size of the hot spot and at least one pixel value of the at least one surrounding pixel are utilized.

20. The method according to any one of claims 15 to 17, wherein in step h), the correction factor has a plurality of possible values, and wherein step h) comprises: The value of the correction factor is determined as a function of the number of pixels in the image that correspond to the size of the hot spot.

21. A method according to claim 20, wherein the method comprises step a): determining the multiple possible values ​​of the correction factor based on a plurality of calibration images of one or more hot spots acquired by the camera and / or one or more cameras substantially identical to the camera, and based on one or more measured temperatures or reference temperatures of the one or more hot spots.

22. The method of claim 21, wherein a thermocouple is used to provide the measured temperature.

23. The method according to any one of claims 21 to 22, wherein step a) comprises: For each calibration image of the plurality of calibration images: determining the number of pixels in the calibration image corresponding to the size of the hotspot; Determining the average temperature of the hot spot in the calibration image includes utilizing pixel values ​​of the pixels in the calibration image corresponding to the size of the hot spot and the number of pixels in the calibration image corresponding to the size of the hot spot; and determining the ambient temperature in the calibration image includes utilizing at least one pixel in the calibration image that is different from the pixel in the calibration image corresponding to the hot spot.

24. The method of claim 23, wherein in step a), the pixels in the calibration image corresponding to the size of the hotspot include: pixels in the calibration image corresponding to the maximum temperature of the hotspot, and wherein step a) comprises determining the pixels in the calibration image corresponding to the size of the hotspot as pixels in the calibration image whose values ​​are within a threshold range of the values ​​of the pixels determined to have the maximum temperature of the hotspot in the calibration image.

25. The method according to any one of claims 21 to 22 and 24, wherein step a) comprises: For each calibration image of the plurality of calibration images, an error value is determined, the determination comprising: calculating a first value as the average temperature of the hotspot in the calibration image minus the measured temperature or the reference temperature of the hotspot, and calculating a second value as the average temperature of the hotspot in the calibration image minus the ambient temperature of the hotspot in the calibration image, and wherein step a) comprises: determining the error value as a ratio of the first value to the second value.

26. The method according to any one of claims 15 to 17, 21 to 22 and 24, wherein the method comprises step c): transforming the image into an interpolated image for further processing by the processing unit.

Images