A coupler multi-point temperature field anomaly early warning system

Abstract

The application relates to the technical field of temperature detection, and discloses a coupler multi-point temperature field anomaly early warning system, which comprises the following units: a collection and analysis unit, which converts resistance change signals into discrete temperature sampling values, and performs spatial mapping on the discrete temperature sampling values based on the spatial coordinates of each monitoring point in the coupler; a temperature field processing unit, which performs pixel partition calibration on infrared radiation intensity signals, fuses the spatial mapping result and a surface temperature matrix according to a spatial correspondence relationship, and constructs a multi-direction temperature gradient distribution according to all temperature differences; a temperature field early warning unit, which determines a temperature rise concentration area and a corresponding expansion trend according to the change of the multi-direction temperature gradient distribution, determines a temperature field anomaly grade based on the area proportion of the temperature rise concentration area and the expansion trend, and judges whether to adjust the temperature field anomaly grade according to the historical recurrence number of the temperature rise concentration area; and an early warning adjustment unit. The application ensures the reliability of multi-point temperature field anomaly early warning.

Description

Technical Field

[0001] This invention relates to the field of temperature detection technology, and more specifically, to a coupler multi-point temperature field anomaly early warning system. Background Technology

[0002] Laser therapy devices are widely used in clinical settings such as dermatology due to their advantages of minimal invasiveness, high precision, and rapid recovery. As a crucial component of the laser transmission link, the coupler directly determines the reliability of laser output and equipment safety.

[0003] Chinese Patent No. CN119803720A discloses a temperature detection method and system based on a fiber optic grating temperature sensor. It focuses on achieving real-time temperature detection through low-coherence light and a coupler. However, in continuous or pulsed operation, the optical elements, conductive contacts, and driving circuits inside the coupler may accumulate local heat due to factors such as current fluctuations, optical path deviation, and superposition of reflected light. This can lead to problems such as lens coating damage, fiber end face ablation, adhesive layer carbonization and aging, and structural thermal deformation. Due to the lack of global, real-time, and accurate monitoring of the internal temperature field of the coupler, and the inability of single-point temperature measurement to identify and locate local thermal effects, the reliability of the laser therapy device is insufficient.

[0004] Therefore, it is necessary to design a coupler multi-point temperature field anomaly early warning system to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a coupler multi-point temperature field anomaly early warning system, which aims to solve the problem of insufficient reliability of laser therapy devices due to the lack of global, real-time and accurate monitoring of the internal temperature field of the coupler and the inability of single-point temperature measurement to identify and locate local thermal effects.

[0006] This invention proposes a coupler multi-point temperature field anomaly early warning system, comprising:

[0007] The acquisition and analysis unit is configured to acquire resistance change signals and infrared radiation intensity signals at each monitoring point, convert the resistance change signals into discrete temperature sample values, and perform spatial mapping on the discrete temperature sample values ​​based on the spatial coordinates of each monitoring point within the coupler.

[0008] The temperature field processing unit is configured to perform pixel partitioning calibration on the infrared radiation intensity signal, determine the surface temperature matrix, fuse the spatial mapping result and the surface temperature matrix according to the spatial correspondence, obtain the temperature difference between adjacent monitoring points according to the fusion result, and construct a multi-directional temperature gradient distribution based on all temperature differences.

[0009] The temperature field early warning unit is configured to determine the concentrated temperature rise region and the corresponding expansion trend based on the changes in the multi-directional temperature gradient distribution within the sampling period, and to determine the temperature field anomaly level based on the area ratio of the concentrated temperature rise region and the expansion trend.

[0010] The early warning adjustment unit is configured to compare the temperature rise concentration area with the data in the historical temperature database, determine whether to adjust the temperature field anomaly level based on the historical recurrence frequency of the temperature rise concentration area, and determine the target temperature field anomaly level.

[0011] Furthermore, when spatially mapping the discrete temperature sampled values, the following steps are included: the acquisition and analysis unit synchronously samples the resistance change signal of each monitoring point, performs nonlinear conversion based on the resistance-temperature correspondence of the NTC thermistor, determines the discrete temperature sampled value corresponding to each monitoring point, binds the discrete temperature sampled value with the three-dimensional coordinates of each monitoring point in the coupler, determines the spatial node dataset, and performs voxel division of the spatial node dataset based on the structural contour of the coupler to determine the discrete temperature distribution matrix.

[0012] Furthermore, in determining the surface temperature matrix, the process includes: the temperature field processing unit acquiring an infrared thermal imaging image, and based on the structural contour, performing boundary cropping on the infrared thermal imaging image to determine several pixel regions, performing grayscale normalization on the infrared radiation intensity values ​​in each pixel region to determine the surface temperature value of each pixel region, and mapping the surface temperature value of each pixel region to the corresponding space based on the position of each pixel region and the spatial coordinate calibration relationship of the coupler to determine the surface temperature matrix.

[0013] Furthermore, in constructing the multi-directional temperature gradient distribution, the process includes: the temperature field processing unit superimposing and fusing the temperature discrete distribution matrix and the surface temperature matrix to determine the fused temperature field matrix; reconstructing the structure of the fused temperature field matrix based on the structural contour to determine a three-dimensional mesh; determining the fused temperature values ​​of the adjacent mesh cells based on the central mesh cells of the three-dimensional mesh; and determining the multi-directional temperature gradient distribution based on the difference between the central fused temperature value of the central mesh cell and the adjacent fused temperature values.

[0014] Furthermore, in determining the concentrated temperature rise region and its expansion trend, the process includes: the temperature field early warning unit performing a time series comparison of the multi-directional temperature gradient distribution within the sampling period to determine the amount of change; performing a connected component search based on the amount of change; identifying grid cells with a change amount greater than a change amount threshold as candidate regions; determining the centroid displacement trajectory based on the spatial centroid position of the candidate regions; and determining the concentrated temperature rise region and its expansion trend based on the centroid displacement trajectory.

[0015] Furthermore, when determining the anomaly level of the temperature field, the following is included: the expansion trend includes area expansion, area contraction and directional migration, and the temperature field early warning unit determines the area ratio of the concentrated temperature rise region to the outline area of ​​the coupler.

[0016] Furthermore, in determining the temperature field anomaly level, the method further includes: when the area proportion is greater than or equal to the area proportion threshold, and the expansion trend is area expansion or directional migration, the temperature field early warning unit determines the temperature field anomaly level as a Level 1 early warning; when the area proportion is greater than or equal to the area proportion threshold, and the expansion trend is area contraction, or when the area proportion is less than the area proportion threshold, and the expansion trend is area expansion or directional migration, the temperature field early warning unit determines the temperature field anomaly level as a Level 2 early warning; when the area proportion is less than the area proportion threshold, and the expansion trend is area contraction, the temperature field early warning unit determines the temperature field anomaly level as a Level 3 early warning.

[0017] Furthermore, when determining the target temperature field anomaly level, the method includes: when there is historical data in the historical temperature database that is the same as the area proportion and candidate region, the early warning adjustment unit determines to adjust the temperature field anomaly level and determines the target temperature field anomaly level based on the adjustment result; when there is no historical data in the historical temperature database that is the same as the area proportion and candidate region, the early warning adjustment unit determines not to adjust the temperature field anomaly level and determines the current temperature field anomaly level as the target temperature field anomaly level.

[0018] Furthermore, when adjusting the temperature field anomaly level, the following steps are included: when the temperature field early warning unit determines that the temperature field anomaly level is a level two or level three early warning, the early warning adjustment unit raises the temperature field anomaly level by one level.

[0019] Furthermore, when adjusting the temperature field anomaly level, the method further includes: when the temperature field early warning unit determines that the temperature field anomaly level is a Level 1 early warning, the early warning adjustment unit adjusts the early warning frequency of the temperature field anomaly level, and the early warning levels of Level 1, Level 2 and Level 3 are reduced sequentially.

[0020] Compared with existing technologies, the advantages of this invention are as follows: By converting signals into discrete temperature sampling values ​​and completing spatial mapping, real-time acquisition of multi-point, full-domain temperatures within the coupler is achieved. This accurately matches temperature with the spatial location of monitoring points, thereby identifying the risk of localized heat accumulation and overcoming the limitation of single-point temperature measurement in locating localized thermal effects. Pixel partitioning and surface temperature matrix construction are performed on the infrared signal. The internal spatial mapping results are fused with the surface temperature matrix, and a multi-directional temperature gradient distribution is constructed through the temperature difference between adjacent monitoring points. This achieves comprehensive integration of the coupler's internal and surface temperatures, improving the completeness and accuracy of temperature field monitoring. Based on changes in the temperature gradient distribution, the concentrated temperature rise area and its expansion trend are determined. Combined with the area proportion, the anomaly level of the temperature field is judged, enabling graded and targeted early warning. The anomaly level is optimized by combining historical temperature databases and the recurrence frequency of concentrated temperature rise areas, improving the reliability of early warnings and the operational stability of the laser therapy device. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 A functional block diagram of a coupler multi-point temperature field anomaly early warning system provided in an embodiment of the present invention;

[0023] Figure 2 A logic flowchart for determining the level of temperature field anomaly provided in an embodiment of the present invention;

[0024] Figure 3 This is a cross-sectional view of the laser therapy device provided in an embodiment of the present invention.

[0025] In the diagram: 1. Laser focusing system; 2. Light collecting mirror; 3. Laser source light outlet. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0028] See Figure 1-3 As shown in some embodiments of this application, a coupler multi-point temperature field anomaly early warning system includes:

[0029] The acquisition and analysis unit is configured to acquire resistance change signals and infrared radiation intensity signals at each monitoring point, convert the resistance change signals into discrete temperature sample values, and perform spatial mapping on the discrete temperature sample values ​​based on the spatial coordinates of each monitoring point within the coupler.

[0030] The temperature field processing unit is configured to perform pixel partitioning calibration on the infrared radiation intensity signal, determine the surface temperature matrix, fuse the spatial mapping result and the surface temperature matrix according to the spatial correspondence, obtain the temperature difference between adjacent monitoring points according to the fusion result, and construct a multi-directional temperature gradient distribution based on all temperature differences.

[0031] The temperature field early warning unit is configured to determine the concentrated temperature rise area and its corresponding expansion trend based on the changes in the multi-directional temperature gradient distribution within the sampling period, and to determine the temperature field anomaly level based on the area ratio of the concentrated temperature rise area and its expansion trend.

[0032] The early warning adjustment unit is configured to compare the temperature rise concentration area with the data in the historical temperature database, determine whether to adjust the temperature field anomaly level based on the historical recurrence frequency of the temperature rise concentration area, and determine the target temperature field anomaly level.

[0033] The laser therapy device includes optical components such as a laser source output port 3, a light collecting mirror 2, and a laser focusing system 1. The coupler is responsible for connecting the laser emission from the laser source output port 3, the light signal convergence of the light collecting mirror 2, and the precise focusing and transmission of the laser in the laser focusing system 1. It is the key to ensuring the laser output accuracy and the safe operation of the equipment. In the continuous or pulsed clinical operation of the laser therapy device, after the laser is emitted from the source output port and converged by the light collecting mirror 2, it is transmitted to the laser focusing system 1. During this process, the coupler needs to continuously complete the optical signal coupling operation. Its internal optical components, conductive contacts, and driving circuits are prone to local heat accumulation due to factors such as current fluctuations, optical path deviation, and the superposition of reflected light generated by the laser in the emission, collection, and focusing stages.

[0034] Heat accumulation not only directly causes damage to lens coatings and ablation of fiber end faces, but also indirectly affects the light signal collection efficiency of the light collecting mirror 2, the focusing accuracy of the laser focusing system 1, and even damages the transmission compatibility between the laser source output port 3 and the coupler. Existing temperature measurement solutions, such as fiber optic grating temperature measurement, can only achieve single-point real-time temperature measurement through low-coherence light and couplers. They lack the ability to monitor the temperature field inside the coupler in a global, real-time, and accurate manner, and cannot identify and locate local thermal effects. This makes it difficult to match the coupler temperature monitoring requirements under an integrated transmission link, resulting in insufficient overall operational reliability of the laser therapy device.

[0035] Specifically, the monitoring points are signal acquisition points arranged according to the positions of optical elements, conductive contacts, drive circuits, etc. inside the coupler. They are used to accurately capture temperature-related signals at key locations of the coupler in continuous or pulsed operation. The resistance change signal is generated by temperature-sensitive elements arranged at the monitoring points. Due to the local heat accumulation inside the coupler, its own temperature changes, which in turn generate a signal of resistance value change. The infrared radiation intensity signal is the intensity signal of infrared light radiated outward by the coupler surface due to internal heat conduction and self-heating. The infrared radiation intensity signal can indirectly reflect the temperature distribution on the coupler surface.

[0036] The acquisition and analysis unit simultaneously acquires both types of signals, converting the analog resistance change signal into discrete digital temperature sample values ​​for subsequent digital processing. Based on the spatial coordinates of each monitoring point within the coupler, the discrete temperature sample values ​​are spatially mapped. These spatial coordinates are the specific location parameters (x, y, z axes) of each monitoring point in the three-dimensional space within the coupler. Spatial mapping essentially binds each discrete temperature sample value to its corresponding monitoring point's spatial coordinates, clearly defining the specific temperature value corresponding to each three-dimensional location within the coupler, achieving precise temperature-location correspondence. By deploying multiple monitoring points and simultaneously acquiring signals, simultaneous acquisition and precise correspondence of multiple locations within the coupler are achieved, solving the technical problem of single-point temperature measurement's inability to identify and locate local thermal effects. This lays a data foundation for subsequent global temperature field monitoring and local heat accumulation identification.

[0037] The temperature field processing unit performs pixel partitioning calibration on the infrared radiation intensity signal. Pixel partitioning calibration is performed by acquiring the infrared radiation intensity signal of the coupler through an infrared thermal imaging device to form a complete infrared radiation image. The image is then cropped according to the structural contour of the coupler to divide it into multiple independent pixel regions, ensuring that the infrared radiation intensity signal of each pixel region can accurately reflect the surface temperature at the corresponding location. Based on the actual spatial position of each pixel region on the coupler surface, they are arranged sequentially to form a two-dimensional matrix, namely the surface temperature matrix. The surface temperature matrix reflects the temperature of each pixel region after calibration, and it intuitively and clearly presents the temperature distribution of the coupler surface. According to the spatial correspondence, the temperature field processing unit fuses the temperature-position spatial mapping result obtained by the acquisition and analysis unit with the surface temperature matrix, and performs one-to-one spatial matching between the internal discrete temperature data (reflecting the internal temperature) and the surface temperature matrix data (reflecting the surface temperature), integrating them to form full-domain temperature data containing information on the internal and surface temperatures of the coupler.

[0038] Based on the fused global temperature data, the temperature difference between adjacent monitoring points is calculated, which reflects the temperature variation at adjacent locations. These temperature differences are used to construct multi-directional temperature gradient distributions within the coupler in different directions (e.g., horizontal, vertical, and oblique), clearly reflecting the direction and concentration of heat diffusion. Pixel partitioning and spatial correspondence fusion solve the problem that single internal or surface temperature measurements cannot comprehensively reflect the coupler's temperature field distribution. Furthermore, by constructing multi-directional temperature gradient distributions, the direction of temperature change and areas of temperature concentration are accurately identified, avoiding the shortcomings of single temperature data in determining heat diffusion trends and identifying localized heat accumulation.

[0039] The sampling period is a pre-set, fixed signal acquisition and analysis cycle used to periodically and continuously monitor the dynamic changes in the coupler's temperature field. It can be dynamically adjusted according to the duration of laser treatment. The temperature field early warning unit dynamically analyzes the changes in the multi-directional temperature gradient distribution within the continuous sampling period, capturing the numerical and directional changes in the gradient distribution. This allows for the determination of the temperature change patterns and trends, specifically the concentrated temperature rise areas and their corresponding expansion trends. The concentrated temperature rise areas are regions with large temperature gradient changes and relatively concentrated spatial locations; these areas are where heat accumulation and potential safety hazards may occur inside the coupler. The expansion trend refers to the changes in the area and location of the concentrated temperature rise areas within the continuous sampling period, such as gradual expansion or migration towards the coupler's optical elements.

[0040] The temperature field early warning unit determines the anomaly level of the temperature field based on the area ratio and expansion trend of the concentrated temperature rise region. The area ratio is the ratio of the actual area of ​​the concentrated temperature rise region to the effective working area inside the coupler as a whole. The larger the ratio, the wider the scope of the potential hazard. By dynamically analyzing the changes in temperature gradient distribution, the unit accurately locates the concentrated temperature rise region and determines its development trend, issuing targeted early warnings of different levels, thus achieving precise early warning of temperature field anomalies. The historical temperature database is a database pre-stored by the system containing temperature field data, information on concentrated temperature rise regions, temperature field anomaly levels, hazard handling results, and other relevant data from all past sampling periods, providing historical reference for current early warning results.

[0041] The early warning adjustment unit compares the spatial location, size, and temperature gradient characteristics of the currently identified concentrated temperature rise area with all historical data in the historical temperature database. It determines whether the concentrated temperature rise area has appeared in previous monitoring periods. The historical recurrence count is the number of times the concentrated temperature rise area has appeared in all past sampling periods. A higher recurrence count indicates that the heat accumulation in this area is a recurring potential hazard, rather than an occasional fluctuation. Based on the historical recurrence count, the early warning adjustment unit determines whether to adjust the previously determined temperature field anomaly level. If the concentrated temperature rise area appears multiple times, it indicates a recurring temperature hazard, and the temperature field anomaly level is adjusted accordingly to improve the accuracy and reliability of the early warning. Finally, the target temperature field anomaly level is determined, and the level is corrected based on historical data, effectively reducing the probability of false alarms and missed alarms. This allows for timely and accurate identification of coupler temperature field anomalies, avoiding problems such as lens coating damage, fiber end face ablation, adhesive layer carbonization and aging, and structural thermal deformation caused by temperature field anomalies, thereby improving the reliability and safety of the laser therapy device.

[0042] Synchronously acquiring resistance change signals from each monitoring point provides a data foundation for the accuracy of subsequent temperature conversion and spatial mapping, avoiding deviations in temperature-location correspondence caused by asynchronous sampling. NTC thermistors are thermistors with a negative temperature coefficient, meaning their resistance gradually decreases as temperature increases, and the relationship between resistance and temperature is not linear. The resistance-temperature correspondence is a fixed correlation between the resistance value of the NTC thermistor and the corresponding temperature value, calibrated experimentally. Nonlinear conversion is necessary because the resistance of the NTC thermistor has a nonlinear relationship with temperature. The acquisition and analysis unit needs to convert the acquired resistance change signal (resistance value) into the corresponding temperature value through nonlinear conversion. Discrete temperature sampling values ​​are multiple independent temperature values ​​obtained after converting the continuously changing resistance change signal. Each discrete temperature sampling value corresponds to the temperature of a monitoring point at a certain moment. By utilizing the temperature sensitivity of the NTC thermistor and the nonlinear conversion relationship, the accuracy of temperature conversion is ensured.

[0043] By binding discrete temperature sample values ​​with the three-dimensional coordinates of each monitoring point within the coupler, a spatial node dataset is determined. This allows for the precise location of each monitoring point within the coupler. A one-to-one correspondence is established between the discrete temperature sample value and the three-dimensional coordinates of each monitoring point, ensuring that each temperature value clearly corresponds to its specific spatial location within the coupler. The spatial node dataset is a collection of all data bound to three-dimensional coordinates and discrete temperature sample values. Each data point contains a correspondence between three-dimensional coordinates and discrete temperature sample values, and can be viewed as a discrete spatial node of the temperature field within the coupler, laying the foundation for constructing a global temperature field.

[0044] Based on the structural contour of the coupler, the spatial node dataset is divided into voxels. The structural contour of the coupler is the external shape of the coupler itself, the distribution contour of its internal cavities and components. Voxel division divides the three-dimensional space inside the coupler into multiple uniform three-dimensional cubic units (i.e., voxels) according to the structural contour of the coupler. Each voxel corresponds to a small spatial region inside the coupler. During the division, the spatial nodes corresponding to all monitoring points can fall into the corresponding voxel. After voxel division, the discrete temperature sampling values ​​in each voxel are integrated to determine the average value of the discrete temperature sampling values ​​of all spatial nodes in the voxel. Then, the average values ​​corresponding to all voxels are arranged in sequence according to the spatial position of the voxels in the coupler to form a discrete temperature distribution matrix. The discrete temperature distribution matrix intuitively presents the discrete temperature distribution of each spatial region inside the coupler, so that the temperature distribution can accurately match the internal space of the coupler, further improving the globality and accuracy of temperature field monitoring.

[0045] Infrared thermal imaging images are images captured by infrared thermal imaging equipment that reflect the distribution of infrared radiation intensity on the surface of the coupler. The grayscale value of each pixel in the infrared thermal imaging image corresponds to the infrared radiation intensity at the corresponding position on the coupler surface. The temperature field processing unit performs cropping processing on the acquired infrared thermal imaging image according to the structural contour of the coupler to remove background areas unrelated to the coupler and avoid interference from infrared radiation signals from irrelevant backgrounds in subsequent temperature calculations. The cropped infrared thermal imaging image containing only the coupler is divided into multiple independent, continuous small pixel blocks according to the image pixel size. Each small pixel block is a pixel region, and each pixel region corresponds to a small spatial region on the coupler surface. The infrared radiation intensity values ​​in each pixel region are grayscale normalized and converted. The infrared radiation intensity value is the intensity value of infrared light radiated outward from the corresponding position on the coupler surface corresponding to each pixel in the infrared thermal imaging image.

[0046] Because the range of infrared radiation intensity values ​​varies across different pixel regions, and is affected by environmental interference and the accuracy of the equipment itself, the infrared radiation intensity values ​​may deviate. By using a normalization algorithm to convert the infrared radiation intensity values ​​of all pixel regions to a uniform numerical range, the influence of numerical differences and interference can be eliminated, allowing the surface temperature value to accurately correspond to the actual temperature value of the pixel region. The surface temperature value corresponding to each pixel region is associated with the three-dimensional coordinates of the corresponding position on the coupler surface, mapping the surface temperature value to the actual space of the coupler. The matrix formed by arranging the pixel regions according to their spatial distribution on the coupler surface is the surface temperature matrix. The surface temperature matrix can intuitively and clearly present the temperature distribution at various spatial positions on the coupler surface, further improving the comprehensiveness and accuracy of coupler temperature field monitoring.

[0047] The discrete temperature distribution matrix is ​​a matrix determined by voxelizing the spatial node dataset based on the structural contour of the coupler. It is used to present the discrete temperature distribution of each spatial region inside the coupler. The surface temperature matrix is ​​a matrix determined by mapping the surface temperature value of each pixel region to the corresponding space based on the position of each pixel region and the spatial coordinate calibration relationship of the coupler. It is used to intuitively and clearly present the temperature distribution of each spatial position on the coupler surface. The internal temperature data of the coupler contained in the discrete temperature distribution matrix is ​​superimposed and integrated with the surface temperature data of the coupler contained in the surface temperature matrix. That is, the spatial position correspondence between each voxel in the discrete temperature distribution matrix (corresponding to a certain three-dimensional spatial region inside the coupler) and each pixel region in the surface temperature matrix (corresponding to a certain three-dimensional spatial region on the coupler surface) is clarified, ensuring that the temperature data in the two matrices can accurately correspond to the same spatial position of the coupler.

[0048] The fused temperature field matrix integrates all temperature data from the coupler's interior and surface, comprehensively reflecting the coupler's global temperature distribution. Based on the coupler's structural contour, the fused temperature field matrix undergoes spatial structural adjustments to ensure a precise match between its spatial distribution and the coupler's actual structure. This prevents misalignment between temperature data and the actual structure. For example, if the coupler's actual structure includes an internal optical lens mounting cavity, fiber optic interface grooves, and external shell protrusions, the initial state of the fused temperature field matrix after superposition and fusion is a regular three-dimensional matrix with a uniform cubic grid distribution, which cannot match the coupler's irregular structure. In this case, the temperature field processing unit spatially adjusts the fused temperature field matrix according to the coupler's structural contour. It adapts and shrinks the matrix grid corresponding to the optical lens mounting cavity area to fit the actual cavity shape and size, deletes redundant temperature data grids in the fiber optic interface groove area that do not correspond to the actual structure, and extends and expands the matrix grid in the external shell protrusion area, ensuring that each grid cell accurately corresponds to the actual structural area of ​​the coupler.

[0049] The 3D mesh divides the 3D space inside and on the surface of the coupler into multiple uniform, interconnected 3D mesh units according to the fused temperature field matrix after structural reconstruction. Each mesh unit corresponds to a spatial region inside or on the surface of the coupler, and each mesh unit contains a corresponding fused temperature value. The fused temperature value is the temperature data corresponding to a specific spatial location of the coupler after superimposing and fusing the temperature discrete distribution matrix and the surface temperature matrix. The central mesh unit is selected as a reference point by choosing a center in the 3D mesh, and the fused temperature value of the center mesh unit is the fused temperature value corresponding to the center mesh unit. All meshes spatially adjacent to the central mesh unit are considered fused temperature values. The fusion temperature value corresponding to the grid cell is the adjacent fusion temperature value. The temperature field processing unit uses the central grid cell as a reference to determine the fusion temperature values ​​of all its adjacent grid cells one by one. Based on the difference between the central fusion temperature value and each adjacent fusion temperature value, and combined with the different spatial directions of the adjacent grid cells relative to the central grid cell (such as horizontal, vertical, oblique, etc.), these differences are correlated with the corresponding spatial directions to finally determine the multi-directional temperature gradient distribution. The multi-directional temperature gradient distribution can reflect the rate and trend of change of the coupler in different directions in the full-domain three-dimensional space, further improving the comprehensiveness of coupler temperature field monitoring.

[0050] Time series alignment compares all multi-directional temperature gradient distributions one by one in chronological order within the sampling period. This comprehensively and continuously captures the dynamic changes of the multi-directional temperature gradient distribution within the sampling period. After comparing the multi-directional temperature gradient distribution corresponding to each grid cell in the 3D mesh with the multi-directional temperature gradient distribution of the same grid cell in the sampling period, the resulting changes can intuitively reflect the changes in the temperature gradient of each grid cell, that is, the strength of the temperature change in the coupler space region corresponding to that grid cell. This clearly distinguishes grid cells with stable, slightly fluctuating, and drastically changing temperature gradients.

[0051] Connected domain search involves a global search of the changes in all mesh cells within a 3D mesh. It identifies clusters of mesh cells whose changes are correlated and spatially adjacent (including horizontal, vertical, and diagonal adjacency). A connected domain is a continuous and complete region composed of spatially adjacent mesh cells with consistent change characteristics. For example, if there are multiple adjacent mesh cells near the optical lens mounting cavity inside the coupler, and these mesh cells all exhibit significant abnormal changes and are spatially connected without interruption, then these adjacent mesh cells are considered a complete and continuous connected domain. This avoids misclassifying mesh cells that are originally adjacent and belong to the same abnormal region as multiple independent and scattered regions. It also excludes isolated mesh cells with small fluctuations. If the change exceeds a threshold, it indicates that the temperature gradient change of the mesh cell or the mesh cell formed by the connected domain has exceeded the normal operating range. This suggests a possible abnormal temperature rise and heat accumulation in the corresponding spatial region of the coupler. Therefore, all mesh cells with changes exceeding the threshold are filtered out, and the region formed by these mesh cells is the candidate region.

[0052] By statistically analyzing the three-dimensional spatial coordinates of all grid cells within the candidate region, the spatial centroid position of the candidate region is determined based on these coordinates. The spatial centroid position accurately represents the overall spatial position of the candidate region, avoiding location deviations caused by irregular shapes (such as long strips or irregular blocks). Within the sampling period, the temperature field early warning unit records the spatial centroid position in each sampling interval and connects these different spatial centroid positions sequentially according to time, forming a continuous trajectory reflecting the changes in the spatial position of the candidate region. For example, if a candidate region corresponds to the vicinity of the internal drive circuit of the coupler, and its spatial centroid position is located at the edge of the drive circuit, then at the next sampling moment, it moves towards the center of the drive circuit, and at the third sampling moment, it continues to move towards the center. By sequentially connecting the centroid positions of these three sampling cycles, a trend of candidate regions migrating towards the center of the drive circuit can be formed. Candidate regions with rapid temperature increases, high temperature values, and relatively concentrated spatial locations are identified as temperature rise concentration regions. If the centroid displacement trajectory shows a continuous movement in a specific direction, it indicates that the temperature rise concentration region is migrating in that direction. If the centroid displacement trajectory remains basically unchanged, but the connected domain of the candidate region continues to expand, it indicates that the area of ​​the temperature rise concentration region is expanding. If the centroid displacement trajectory remains basically unchanged, but the connected domain of the candidate region continues to shrink, it indicates that the area of ​​the temperature rise concentration region is shrinking. By determining the specific spatial location and range of the temperature rise concentration region, problems such as damage and aging caused by heat accumulation can be avoided, thereby improving the reliability of multi-point temperature field early warning for the coupler.

[0053] Area expansion refers to the continuous expansion of the connected domain of the concentrated temperature rise area, corresponding to a continuous increase in the coverage area of ​​the abnormal heating area of ​​the coupler. Area contraction refers to the continuous shrinking of the connected domain of the concentrated temperature rise area, corresponding to a continuous decrease in the coverage area of ​​the abnormal heating area of ​​the coupler. Directional migration refers to the continuous movement of the spatial centroid of the concentrated temperature rise area along a specific direction, corresponding to a shift in the spatial position of the abnormal heating area of ​​the coupler. For example, if the concentrated temperature rise area gradually moves from the fiber optic interface of the coupler towards the internal optical lens, then this change is directional migration. If the area of ​​the concentrated temperature rise area gradually expands from its original small region, covering more surrounding grid cells, then this change is area expansion. If the area of ​​the concentrated temperature rise area gradually shrinks and the number of grid cells decreases, then this change is area contraction.

[0054] The temperature field early warning unit determines the area ratio between the concentrated temperature rise region and the coupler's outline area. In other words, the area ratio is determined after the expansion trend is determined. The concentrated temperature rise region is the total area of ​​the actual spatial area of ​​the coupler corresponding to all grid cells covered by the concentrated temperature rise region. The coupler's outline area refers to the total area of ​​the entire spatial area covered by the coupler's own external structure. The area ratio can intuitively reflect the coverage of the concentrated temperature rise region in the overall structure of the coupler. The larger the ratio, the wider the impact range of the abnormal heating region on the coupler. When the area ratio of the concentrated temperature rise region is greater than or equal to the area ratio threshold, and the expansion trend is area expansion or directional migration, it indicates that the impact range of the concentrated temperature rise region has exceeded the reasonable range, and the abnormal heating region is still expanding or migrating, continuously increasing the threat to key components such as optical elements and conductive contacts inside the coupler. The temperature field early warning unit determines the temperature field anomaly level to be a Level 1 warning.

[0055] When the area proportion of the concentrated temperature rise region is greater than or equal to the area proportion threshold, and the expansion trend is area shrinkage, it indicates that although the influence range of the concentrated temperature rise region exceeds the reasonable range, the abnormal heating area is continuously shrinking and the threat is gradually decreasing. Alternatively, when the area proportion is less than the area proportion threshold, but the expansion trend is area expansion or directional migration, it indicates that although the influence range of the concentrated temperature rise region is within the reasonable range, the abnormal heating area is continuously expanding or migrating, and the threat may escalate further. In both cases, the temperature field early warning unit determines the temperature field anomaly level as a Level II early warning.

[0056] When the area of ​​the concentrated temperature rise region is less than the area ratio threshold and the expansion trend is area shrinkage, it indicates that the influence range of the concentrated temperature rise region is within a reasonable range, and the abnormal heating area is continuously shrinking, and the threat is constantly decreasing. The temperature field early warning unit determines the temperature field anomaly level as a level three early warning. Based on the severity, scope of influence and development trend of the temperature anomaly, different levels of early warning are issued, further improving the reliability of the coupler's multi-point temperature field early warning.

[0057] The historical temperature database is a pre-stored database of coupler temperature-related data monitored in all past sampling periods. If the historical temperature database contains historical monitoring data with the same area percentage, candidate region location, and size as the currently monitored data, meaning that the current temperature anomaly (area percentage, candidate region) has occurred in the past, the early warning adjustment unit will determine to adjust the temperature field anomaly level. This avoids oversights caused by applying judgments based on occasional fluctuations to recurring problems. By adjusting the anomaly level, the targeting of the early warning is improved, preventing the risk of coupler damage due to neglecting recurring problems. When the historical temperature database does not contain historical data with the same area percentage and candidate region, it indicates that the current temperature anomaly is a first occurrence or an occasional occurrence without any past reference. In this case, the early warning adjustment unit will not adjust the temperature field anomaly level and will directly use the currently determined temperature field anomaly level as the target temperature field anomaly level. This achieves adaptive correction of the temperature field anomaly level. Through data-driven adaptive adjustment, the reliability of multi-point temperature field early warning for the coupler is ensured.

[0058] The warning frequency is the number of times the system issues warning alerts based on the current temperature field anomaly level. A higher warning frequency indicates a more frequent and denser series of warnings per unit time, providing more frequent reminders to staff. The warning levels decrease sequentially from Level 1 to Level 3, meaning the severity is ranked as follows: Level 1 is the highest, followed by Level 2, and then Level 3 is the lowest. Specifically, Level 1 corresponds to the most serious temperature anomaly hazard, Level 2 to a moderate one, and Level 3 to the least serious. When the temperature field warning unit determines the temperature field anomaly level to be Level 2 or Level 3, it indicates that the identified temperature anomaly hazard is moderate or minor. However, since identical historical data exists in the historical temperature database, this suggests… If the temperature anomaly is a recurring hazard rather than a minor, occasional fluctuation, the temperature anomaly level will be raised by one level. Specifically, a Level 2 warning will be raised to Level 1, and a Level 3 warning to Level 2. When the temperature anomaly level determined by the temperature field warning unit is Level 1, since Level 1 is already the highest level and cannot be further raised, the warning adjustment unit will adjust the warning frequency. By shortening the warning interval, the number of warning reminders per unit time will be increased, continuously reminding staff of the most serious recurring temperature anomaly hazard. This avoids underestimating or missing recurring hazards and ensures that the highest-level hazards receive warning resources, further improving the accuracy and relevance of temperature anomaly warnings.

[0059] In summary, the beneficial effects of this invention are as follows: By converting the signal into discrete temperature sampling values ​​and completing spatial mapping, real-time acquisition of multi-point, full-domain temperature within the coupler is achieved. This accurately matches the temperature with the spatial location of the monitoring points, thereby identifying the risk of localized heat accumulation and overcoming the limitation of single-point temperature measurement in locating localized thermal effects. Pixel partitioning and surface temperature matrix construction are performed on the infrared signal. The internal spatial mapping results are fused with the surface temperature matrix, and a multi-directional temperature gradient distribution is constructed through the temperature difference between adjacent monitoring points. This achieves comprehensive integration of the coupler's internal and surface temperatures, improving the completeness and accuracy of temperature field monitoring. Based on changes in the temperature gradient distribution, the concentrated temperature rise area and its expansion trend are determined. Combined with the area proportion, the anomaly level of the temperature field is judged, enabling graded and targeted early warning. The anomaly level is optimized by combining historical temperature databases and the recurrence frequency of concentrated temperature rise areas, improving the reliability of early warning and the operational stability of the laser therapy device.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A coupler multi-point temperature field anomaly early warning system, comprising: The acquisition and analysis unit is configured to acquire resistance change signals and infrared radiation intensity signals at each monitoring point, convert the resistance change signals into discrete temperature sample values, and perform spatial mapping on the discrete temperature sample values ​​based on the spatial coordinates of each monitoring point within the coupler. The temperature field processing unit is configured to perform pixel partitioning calibration on the infrared radiation intensity signal, determine the surface temperature matrix, fuse the spatial mapping result and the surface temperature matrix according to the spatial correspondence, obtain the temperature difference between adjacent monitoring points according to the fusion result, and construct a multi-directional temperature gradient distribution based on all temperature differences. The temperature field early warning unit is configured to determine the concentrated temperature rise region and its corresponding expansion trend based on the changes in the multi-directional temperature gradient distribution within the sampling period, and to determine the temperature field anomaly level based on the area ratio of the concentrated temperature rise region and the expansion trend. Specifically, the temperature field early warning unit performs time-series comparison of the multi-directional temperature gradient distribution within the sampling period to determine the amount of change, performs connected component search based on the amount of change, identifies grid cells with changes greater than a threshold as candidate regions, determines the centroid displacement trajectory based on the spatial centroid position of the candidate regions, and determines the concentrated temperature rise region and its expansion trend based on the centroid displacement trajectory. The expansion trend includes area expansion, area contraction, and directional migration. The temperature field early warning unit determines the area ratio of the concentrated temperature rise region to the coupler's contour area. When the area ratio is greater than or equal to the area ratio threshold, and the expansion trend is area expansion or directional migration, it is determined as a Level 1 early warning; when the area ratio is greater than or equal to the area ratio threshold, and the expansion trend is area contraction, or when the area ratio is less than the area ratio threshold, and the expansion trend is area expansion or directional migration, it is determined as a Level 2 early warning; when the area ratio is less than the area ratio threshold, and the expansion trend is area contraction, it is determined as a Level 3 early warning. The early warning adjustment unit is configured to compare the temperature rise concentration area with the data in the historical temperature database, determine whether to adjust the temperature field anomaly level based on the historical recurrence frequency of the temperature rise concentration area, and determine the target temperature field anomaly level, including: when there is historical data with the same area proportion and candidate area, determining to adjust the anomaly level and determining the target temperature field anomaly level based on the adjustment result; when there is no historical data with the same area proportion and candidate area, determining not to adjust the anomaly level and determining the current temperature field anomaly level as the target temperature field anomaly level.

2. The coupler multi-point temperature field anomaly early warning system according to claim 1, characterized in that, When spatially mapping the discrete temperature sampled values, the following steps are included: the acquisition and analysis unit synchronously samples the resistance change signal of each monitoring point, performs nonlinear conversion based on the resistance-temperature correspondence of the NTC thermistor, determines the discrete temperature sampled value corresponding to each monitoring point, binds the discrete temperature sampled value with the three-dimensional coordinates of each monitoring point in the coupler, determines the spatial node dataset, divides the spatial node dataset into voxels based on the structural contour of the coupler, and determines the discrete temperature distribution matrix.

3. The coupler multi-point temperature field anomaly early warning system according to claim 2, characterized in that, The process of determining the surface temperature matrix includes: the temperature field processing unit acquiring an infrared thermal imaging image, and performing boundary cropping on the infrared thermal imaging image based on the structural contour to determine several pixel regions; performing grayscale normalization on the infrared radiation intensity values ​​in each pixel region to determine the surface temperature value of each pixel region; and mapping the surface temperature value of each pixel region to the corresponding space based on the position of each pixel region and the spatial coordinate calibration relationship of the coupler to determine the surface temperature matrix.

4. The coupler multi-point temperature field anomaly early warning system according to claim 3, characterized in that, When constructing a multi-directional temperature gradient distribution, the process includes: the temperature field processing unit superimposing and fusing the discrete temperature distribution matrix and the surface temperature matrix to determine a fused temperature field matrix; reconstructing the structure of the fused temperature field matrix based on the structural contour to determine a three-dimensional mesh; determining the fused temperature values ​​of adjacent mesh cells based on the central mesh cells of the three-dimensional mesh; and determining the multi-directional temperature gradient distribution based on the difference between the central fused temperature value of the central mesh cell and the adjacent fused temperature values.

Images