Flushing and cooling method, system and equipment with automatic route identification function

Abstract

The invention provides a flushing cooling method, system and equipment with an automatic route recognition function. The method comprises the steps that an equipment model is constructed through three-dimensional laser scanning, and a surface heat conductivity coefficient is obtained in combination with a material recognition technology. A temperature gradient and thermal load distribution diagram is generated through multi-sensor data fusion, and the heat dissipation requirement is analyzed and the flushing priority is determined based on a thermodynamic model. A dynamic planning algorithm is adopted to optimize the flushing path and sequence, then pressure, flow and time parameters of all flushing points are adjusted in a self-adaptive mode according to thermal load distribution, and a personalized flushing scheme is formed. The temperature change is monitored in real time in the execution process, and parameters are dynamically adjusted to guarantee the cooling efficiency. According to the method, intelligent identification of the thermal state of the equipment surface, automatic planning of the flushing path and accurate control of the flushing parameters are achieved, the flushing cooling accuracy and efficiency are remarkably improved, and the method is suitable for efficient heat dissipation maintenance of industrial equipment.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a flushing and cooling method, system and equipment with automatic route identification. Background Art

[0002] Industrial equipment washdown and cooling technology plays a vital role in modern manufacturing, directly impacting equipment operational stability, production efficiency, and product quality. With the continuous advancement of industrial automation, the demand for intelligent and precise washdown and cooling systems is becoming increasingly stringent. This technology has become a key component in ensuring the safe operation of industrial equipment.

[0003] Traditional flushing and cooling methods rely heavily on manual experience to establish fixed flushing routes and parameters, without the ability to dynamically adjust to the specific characteristics of individual equipment. This static approach results in suboptimal flushing results, potentially leaving areas of insufficient flushing and wasting resources, making it difficult to meet the diverse and personalized cooling needs of modern industrial equipment.

[0004] The current regional challenges faced in this field stem from the complexity of identifying the structural characteristics of equipment. Different industrial equipment has significant differences in geometry, internal structure, material properties, etc., and it is difficult for the system to accurately identify these structural features and formulate appropriate flushing strategies accordingly. This difficulty in identification further leads to the technical difficulties of heat load distribution analysis, because the heat generation and heat dissipation requirements of various parts of the equipment are not uniform during operation, and the system needs to be able to accurately analyze the heat distribution patterns and changing trends. The inaccuracy of heat load distribution analysis directly leads to the blindness of flushing route planning. The system cannot determine the optimal flushing sequence, key flushing areas, and the required flushing intensity and duration for each area, which ultimately affects the overall cooling effect.

[0005] Therefore, how to build a system that can automatically identify the structural characteristics of equipment, accurately analyze the heat load distribution, and intelligently plan the flushing route based on this, and realize the automated customization of the flushing and cooling plan, has become a key issue in the development of current industrial equipment flushing and cooling technology. Summary of the Invention

[0006] To achieve automated customization of flushing and cooling plans for equipment, the present invention provides, in its first aspect, a flushing and cooling method with automatic route identification, comprising: S101. 3D laser scanning technology is used to comprehensively scan the equipment surface to obtain a 3D model of the equipment; S102. Based on the 3D model, material properties of the equipment surface are analyzed using a material identification method, and the thermal conductivity of the equipment surface is obtained based on the material property information; S103. The temperature distribution of various parts of the equipment during operation is monitored in real time, and the measurement results of multiple sensors are integrated using temperature data fusion technology to obtain a temperature gradient distribution map of the equipment surface; S104. According to the temperature gradient distribution map and thermal conductivity, the heat load calculation model is used to analyze the heat load data, heat load density distribution data and heat dissipation requirements of each area, and the flushing priority is determined based on the heat load distribution map and heat dissipation requirements; S105. Based on the priority, the shortest path and flushing order between each flushing point are calculated by the dynamic programming method to obtain the flushing route; S106. According to the flushing route and the heat load density distribution data of each area, the parameter adaptive adjustment algorithm is used to determine the flushing pressure, flow rate and duration of each flushing point to obtain a personalized flushing parameter configuration table; S107. Based on the flushing route and the flushing parameter configuration table, automated flushing operations are performed, and the temperature changes in each area are monitored in real time. If the temperature drop rate is lower than the expected value, the flushing parameters are dynamically adjusted.

[0007] Preferably, the step S101 includes: Use 3D laser technology to fully detect the surface of the equipment, collect original geometric shape information and spatial coordinate data, and obtain the initial scanning data set; The point cloud data processing algorithm is used to denoise the initial scan data set, remove noise interference, and obtain a cleaned point cloud data set; Perform registration operations on the cleaned point cloud dataset, adjust the data consistency of different scanning angles, and determine a unified point cloud registration result; Based on the unified point cloud registration results, a 3D model of the equipment is constructed, key structural feature information is extracted, and preliminary 3D model data is obtained; By optimizing the preliminary 3D model data and correcting the geometric deviation in the 3D model, a refined 3D model data set is obtained; If the structural feature information of some areas in the refined 3D model data set is incomplete, the missing parts are supplemented by an interpolation algorithm to obtain the final 3D model data.

[0008] Preferably, the step S102 includes: Based on the three-dimensional model data of the device, an initial model is generated by a modeling tool, and a mesh is performed on the surface of the initial model to obtain surface geometric feature data; Based on the surface geometric feature data, a material identification method is used to analyze the material properties of each grid area on the surface to obtain material property information of each area, and a material distribution map is output based on the material property information; According to the material distribution map and combined with the pre-established material thermal conductivity database, the thermal conductivity of each area is obtained.

[0009] Preferably, the step S103 includes: Use infrared thermal imaging sensors to collect real-time data from various parts of the equipment during operation to obtain a set of original temperature data; Based on the original temperature data set, data preprocessing technology is used to denoise and calibrate the collected information to obtain a corrected temperature data set; For the corrected temperature data set, the measurement results of multiple sensors are integrated to construct a unified temperature distribution matrix; Based on the temperature distribution matrix, the temperature data of different sensors are weighted according to their location importance and sensor accuracy to obtain the overall reference temperature, and the temperature of the uncovered area is estimated using an interpolation algorithm to calculate the complete temperature gradient distribution map of the device surface.

[0010] Preferably, the step S104 includes: The temperature distribution data of each area is obtained through the temperature gradient distribution diagram, and the heat load distribution data and heat load density distribution data of each area are obtained using the pre-established heat load calculation model; Based on the heat load distribution results and the thermal conductivity in the material property database, the heat dissipation requirements of each area are obtained; If the temperature of a region in the heat dissipation demand exceeds a preset threshold, the region is determined to be a high heat load region, and its location information is recorded and a heat load distribution map is generated; Prioritize washdown areas based on heat load profiles and cooling requirements.

[0011] Preferably, the step S105 includes: Based on the priority, determine the area with the highest heat load as the starting flushing point to obtain initial flushing position data; Based on the initial flushing position data, the heat load distribution information of the adjacent area is obtained. If the heat load of the adjacent area is lower than that of the current area and meets the preset heat load threshold, it is marked as the next flushing candidate point, and the candidate flushing point set is determined; For the candidate flushing point set, the dynamic programming method is used to calculate the path distance between each point, obtain all possible paths from the current flushing point to the candidate point, and determine the shortest path as the next flushing route; Based on the shortest path data, obtain the specific location information of the next flushing point and determine whether all areas have been included in the flushing route; If there are areas that are not included in the flushing route, the new area with the highest heat load is obtained as the new flushing point and a new flushing order is determined; According to the new flushing order, the path optimization calculation is repeated to obtain the shortest path connection method of the remaining areas and obtain the complete flushing route.

[0012] Preferably, the step S106 includes: Based on the preset flushing scheme and combined with the heat load density distribution data, if the heat load density of a certain area exceeds the preset regional threshold, the flushing pressure and flushing flow rate are adjusted through an adaptive algorithm to obtain the optimized flushing parameters; Analyze the duration requirements for the optimized flushing parameters. If the heat load density continues to be higher than the regional threshold, extend the flushing time to determine the final flushing duration configuration. Using the final flushing time configuration and combining it with the flushing intensity requirements, the intensity adjustment value of each flushing point is obtained to generate intensity correction data; By combining intensity correction data with configuration parameters and using a unified formatting tool, a personalized flushing parameter configuration table can be obtained.

[0013] Preferably, the step S107 includes: The intelligent control system obtains flushing parameter data from the flushing parameter configuration table, initializes the automated operation process based on the obtained parameters, and obtains a preliminary operation execution plan; According to the preliminary operation execution plan, the automated flushing operation is started. At the same time, the temperature change data of each area is monitored in real time through the sensor network to obtain real-time records of temperature changes; Based on the real-time records of temperature changes, the temperature drop rate of each area is calculated. If the calculated drop rate is lower than the preset expected value, the parameter adjustment mechanism is triggered; Through the parameter adjustment mechanism, the flushing parameter data is dynamically updated, and the updated parameters are used to reconfigure the automated operation process to obtain the adjusted operation execution plan; According to the adjusted operation execution plan, the flushing operation is continuously performed, and the updated temperature change data is obtained through the sensor network to determine whether the temperature drop rate reaches the expected value; If the updated temperature drop rate still does not reach the expected value, the historical temperature change data and flushing parameter data are processed through a regression analysis model to determine the optimal adjustment parameters.

[0014] On the second aspect, the present application provides a flushing and cooling system with automatic route identification, the system comprising: a scanning module, which uses three-dimensional laser scanning technology to perform an all-round scan of the surface of the equipment to obtain a three-dimensional model of the equipment; a material analysis module, which uses a material identification method to analyze the material property information of the equipment surface based on the three-dimensional model, and obtains the thermal conductivity of the equipment surface based on the material property information; a real-time monitoring module, which monitors the temperature distribution of each part of the equipment in real time during operation, and uses temperature data fusion technology to integrate the measurement results of multiple sensors to obtain a temperature gradient distribution map of the equipment surface; a priority determination module, which uses heat load calculation according to the temperature gradient distribution map and thermal conductivity. The model analyzes the heat load data, heat load density distribution data and heat dissipation requirements of each area, and determines the priority of flushing; the flushing route acquisition module calculates the shortest path and flushing order between each flushing point through a dynamic programming method based on the priority to obtain the flushing route; the flushing parameter acquisition module uses a parameter adaptive adjustment algorithm to determine the flushing pressure, flow rate and duration of each flushing point according to the flushing route and the heat load density distribution data of each area, and obtains a personalized flushing parameter configuration table; the dynamic adjustment module performs automated flushing operations based on the flushing route and the flushing parameter configuration table, monitors the temperature changes in each area in real time, and dynamically adjusts the flushing parameters if the temperature drop rate is lower than the expected value.

[0015] In a third aspect, the present application provides a flushing and cooling device with automatic route identification, the device comprising: a memory, a processor, and a flushing and cooling program with automatic route identification stored in the memory and runnable on the processor, the flushing and cooling program with automatic route identification being configured to implement the flushing and cooling method with automatic route identification as described above.

[0016] The technical solution provided by the embodiment of the invention may have the following beneficial effects: The invention discloses a flushing and cooling method with automatic route identification. This method uses three-dimensional laser scanning and infrared thermal imaging technology to obtain equipment geometry and temperature distribution data. It then analyzes surface heat conduction characteristics using a material identification algorithm. A heat load calculation model identifies high-heat-load areas, employs a path optimization algorithm to plan the optimal flushing route, and utilizes a parameter adaptive adjustment algorithm to determine personalized flushing parameter configurations. This method achieves precise positioning and intelligent cooling of hot spots on the surfaces of industrial equipment. It can dynamically adjust flushing strategies based on real-time temperature changes, effectively improving cooling efficiency and equipment heat dissipation performance, extending equipment life, and reducing energy consumption. This provides a new technical solution for the intelligent operation and maintenance of industrial equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The present invention is a flow chart of a flushing and cooling method with automatic route identification.

[0018] Figure 2 Schematic diagram of a specific flow of step S104 of the present invention.

[0019] Figure 3 Schematic diagram of a specific flow of step S107 of the present invention. DETAILED DESCRIPTION

[0020] To further understand the content of the present invention, the present invention is described in detail with reference to the accompanying drawings and examples. The present application is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are intended only to illustrate the relevant invention and are not intended to limit the invention. It should also be noted that, for ease of description, only portions relevant to the invention are shown in the accompanying drawings.

[0021] like Figure 1 In this embodiment, a flushing and cooling method with automatic route identification may specifically include: S101. Use 3D laser scanning technology to perform a full-scale scan of the surface of the equipment to obtain a 3D model of the equipment; In this embodiment, a comprehensive detection of the device surface is performed using three-dimensional laser technology to collect original geometric shape information and spatial coordinate data to obtain an initial scan data set. A point cloud data processing algorithm is used to denoise the initial scan data set to remove noise interference and obtain a cleaned point cloud data set. Perform registration operations on the cleaned point cloud dataset, adjust the data consistency of different scanning angles, and determine a unified point cloud registration result; based on the unified point cloud registration result, build a 3D model of the equipment, extract key structural feature information, and obtain preliminary 3D model data; optimize the preliminary 3D model data and correct the geometric deviations in the 3D model to obtain a refined 3D model dataset; if the structural feature information of certain areas in the refined 3D model dataset is incomplete, supplement the missing parts through interpolation algorithm to obtain the final 3D model data.

[0022] For example, when using 3D laser scanning technology to perform an all-round scan of the surface of industrial equipment, a high-precision laser scanner such as the FARO Focus S350 can be used. Its scanning range can reach 350 meters, the accuracy is ±1mm, and the scanning rate is as high as 976,000 points per second. In actual operation, the equipment is placed in the center of the factory workshop, and the scanner sets multiple scanning points around the equipment to ensure that all surfaces of the equipment are covered. The scanning time is about 2 hours, and about 500 million points of raw point cloud data are generated. Subsequently, denoising is performed through the point cloud data processing algorithm. A statistical filtering-based method is used, and the neighborhood radius is set to 0.05 meters. Outliers that deviate from the average distance by more than 2 times the standard deviation are eliminated. After filtering, the point cloud data is reduced to about 480 million points, and the noise points are reduced by about 4%, thereby improving the data quality. Next, registration processing is performed using an iterative closest point algorithm. Based on the initially aligned point cloud data, the maximum number of iterations is set to 50, and the convergence threshold is 0.001 meters. Ultimately, accurate stitching of multi-viewpoint point cloud data is achieved, with the registration error controlled within 0.002 meters, forming a unified 3D model of the device. On this basis, structural feature parameters are extracted, and the curvature distribution of the device surface is calculated using a curvature analysis algorithm. The curvature threshold is set to 0.1, and approximately 1,000 key geometric feature points of the device are identified. The main dimensional parameters are calculated, such as the device height of 2.5 meters and the width of 1.8 meters, with an error range of ±0.005 meters. Finally, a triangulation algorithm is used to construct a 3D mesh model, generating approximately 2 million triangular facets to ensure that the model surface smoothness meets industrial design requirements, and the surface reconstruction error is less than 0.003 meters.

[0023] Through the above process, a complete technical chain is formed from scanning to modeling. At the same time, the model data can be connected with the factory's digital management system to realize the intelligent application of equipment maintenance and space planning, ensuring data consistency and business collaboration.

[0024] S102. Analyzing material property information of the device surface using a material identification method based on the three-dimensional model, and obtaining a thermal conductivity coefficient of the device surface based on the material property information; In this embodiment, an initial model is generated by a modeling tool based on the three-dimensional model data of the device, and the surface of the initial model is meshed to obtain surface geometric feature data; based on the surface geometric feature data, a material identification method is used to perform material property analysis on each mesh area of ​​the surface to obtain material property information of each area, and a material distribution map is output based on the material property information; based on the material distribution map, combined with a pre-established material thermal conductivity database, the thermal conductivity of each area is obtained.

[0025] For example, based on the acquired three-dimensional model of the device, high-precision scanning technology is first used to extract the geometric features and texture information of the device surface. Assuming the model data comes from a laser scanner, the resolution reaches 0.01 mm, ensuring that all surface details are captured completely. Next, material identification methods are used, such as a convolutional neural network (CNN) model based on deep learning. The pre-trained dataset contains the spectral reflectance characteristics of metal alloys, ceramics, and composite materials, achieving an identification accuracy of 95%. By analyzing the reflectance spectrum data of the device surface, assuming that the reflectance peak in a certain area is at 600 nanometers, combined with database comparison, it is determined to be aluminum alloy with a thermal conductivity of approximately 200 W / (m·K); another area with a reflectance peak at 800 nanometers is determined to be alumina ceramic with a thermal conductivity of only 30 W / (m·K).

[0026] S103, real-time monitoring of the temperature distribution of various parts of the equipment during operation, integrating the measurement results of multiple sensors using temperature data fusion technology to obtain a temperature gradient distribution map of the equipment surface; In this embodiment, infrared thermal imaging sensors are used to collect real-time data from various parts of the equipment when it is in operation, and a set of original temperature data is obtained. Based on the original temperature data set, data preprocessing technology is used to denoise and calibrate the collected information to obtain a corrected temperature data set. For the corrected temperature data set, the measurement results of multiple sensors are fused to construct a unified temperature distribution matrix. The temperature data of different sensors are weighted according to their positional importance and sensor accuracy to obtain an overall reference temperature, and an interpolation algorithm is used to estimate the temperature of uncovered areas to calculate a complete temperature gradient distribution map of the equipment surface.

[0027] For example, infrared thermal imaging sensors can be used to monitor the temperature distribution of various parts of the equipment in real time during operation. A high-resolution infrared thermal imaging camera, such as the FLIR T540, can be used. Its resolution is 464x348 pixels, and its temperature measurement range is -20°C to 1500°C with an accuracy of ±2°C. While the equipment is operating, the sensors collect data every one second, covering key areas of the equipment surface, such as the motor, bearings, and transmission components. This generates temperature matrix data for multiple regions. For example, the average temperature for the motor region is 85.3°C, the bearing region is 72.8°C, and the transmission component is 65.1°C. Temperature data fusion technology is then used to integrate the measurement results from multiple sensors. A weighted average algorithm can be used to weight the temperature data from different sensors based on their locational importance and sensor accuracy. For example, the weight for the motor region is set to 0.5, the bearing region to 0.3, and the transmission component to 0.2. The fused temperature distribution data is: 85.3x0.5+72.8x0.3+65.1x0.2=78.34°C, which serves as the overall reference temperature. Subsequently, an interpolation algorithm (such as universal kriging) is used to estimate the temperature of uncovered areas. This calculation generates a complete temperature gradient distribution map of the equipment surface, showing a gradual decrease in temperature from the motor area outward, reaching a minimum of 60.2°C at the edge. This process is integrated through an automated system, with sensor data collection and algorithm processing performed by a backend program to ensure real-time performance and accuracy.

[0028] S104. Analyze the heat load data, heat load density distribution data, and heat dissipation requirements of each area using a heat load calculation model based on the temperature gradient distribution map and thermal conductivity, and determine the priority of the flushing area based on the heat load distribution map and heat dissipation requirements. In this embodiment, see the attached Figure 2 , obtain the temperature distribution data of each area through the temperature gradient distribution map, and use the pre-established heat load calculation model to obtain the heat load distribution data and heat load density distribution data of each area; according to the heat load distribution results, combined with the thermal conductivity in the material property database, obtain the heat dissipation demand of each area; if the regional temperature in the heat dissipation demand exceeds the preset threshold, the area is judged as a high heat load area, and its location information is recorded and a heat load distribution map is generated; according to the heat load distribution map and the heat dissipation demand, the priority of the flushing area is determined.

[0029] For example, heat load refers to the amount of heat experienced per unit area per unit time. It reflects the intensity of heat transfer in an area or device and is often used to assess the concentration of cooling or heating needs. Temperature values ​​for each area are extracted from a temperature gradient distribution map (such as an infrared thermograph or simulation output). Using a pre-established heat load calculation model (based on thermodynamic formulas such as the heat conduction equation), temperature data is converted into a heat load (in W), representing the heat generation rate of each area. Integrating this model with a database of material properties (such as thermal conductivity), each area's heat dissipation demand (in W) is calculated, representing the amount of heat that needs to be actively removed. If a zone's temperature exceeds a preset threshold, it is marked as a high heat load area and its location is recorded. Based on the recorded location, a visual heat map is created to highlight high heat load areas. Based on the heat load distribution map, flushing (such as coolant flow) is prioritized, prioritizing high heat load areas. A temperature gradient distribution map (such as an image or data table) is obtained using a thermal imager. Preset parameters include a temperature threshold of 80°C (exceeding this value is considered an overheating risk). Material property database: The device material is silicon, with a thermal conductivity of 150W / (m·K); this database is used to calculate heat dissipation requirements. Heat load calculation model: A pre-established simplified model: Heat load (W) = temperature (°C) × thermal conductivity factor (assuming a thermal conductivity factor of 2W / °C, based on experimental calibration). The heat load density calculation formula (heat load density = heat load / area) uses the motor area's heat load of 20 kilowatts and an area of ​​0.2 square meters as an example. The heat load density is 100 kilowatts / square meter, reflecting the concentration of the heat load. A larger value indicates more concentrated heat. The specific process for determining priorities includes the following steps: Step 1: For example, assume a device has four zones. Obtain temperature distribution data for each zone using a temperature gradient distribution map. The temperature gradient distribution map is generated by scanning the device surface with an infrared thermal imager (e.g., a pseudo-color image showing temperature changes). Extract temperature data: Read the temperature values ​​for each zone from the temperature gradient distribution map: Zone 1: 75°C, Zone 2: 82°C, Zone 3: 88°C, Zone 4: 70°C. Step 2: Use a pre-established heat load calculation model to obtain the heat load distribution results for each zone. This example uses a simplified model: heat load = temperature × thermal conductivity (2W / °C). This model is calibrated based on historical data and takes into account device power consumption and heat dissipation characteristics. The calculated heat loads are: Zone 1: 75°C × 2W / °C = 150W; Zone 2: 82°C × 2W / °C = 164W; Zone 3: 88°C × 2W / °C = 176W; Zone 4: 70°C × 2W / °C = 140W. The heat load distribution results are output as follows: Zone 1: 150 W, Zone 2: 164 W, Zone 3: 176 W, Zone 4: 140 W.

[0030] Step 3: Based on the heat load distribution results and the thermal conductivity in the material property database, obtain the heat dissipation requirements for each zone. Identify and record areas with high heat loads. Calculate heat dissipation requirements: Heat dissipation requirements represent the amount of heat that needs to be removed through active cooling (such as liquid flushing) to prevent further temperature increases. Formula: Heat dissipation requirement (W) = heat load × heat dissipation coefficient. The heat dissipation coefficient is calculated based on the thermal conductivity (for simplicity in this example, the heat dissipation coefficient = 1.2). Zone 1: 150 W × 1.2 = 180 W, Zone 2: 164 W × 1.2 = 196.8 W ≈ 197 W, Zone 3: 176 W × 1.2 = 211.2 W ≈ 211 W, Zone 4: 140 W × 1.2 = 168 W. Determine high heat load areas: Compare temperature with preset threshold (80°C): Zone 1: 75℃<80℃ → not exceeded, not a high heat load area; Area 2: 82℃ > 80℃ → exceeds the threshold and is marked as a high heat load area; Area 3: 88℃ > 80℃ → exceeds the threshold and is marked as a high heat load area; Zone 4: 70℃<80℃→not exceeded, non-high heat load area; Record location information: Area 2 (location coordinates: x=10mm, y=20mm) and Area 3 (location coordinates: x=30mm, y=20mm) are recorded as high heat load areas. Cooling demand data and a list of high heat load areas are generated; a heat load distribution map is generated based on the cooling demand.

[0031] Step 4: Generate a heat load distribution map. Based on the cooling demand and high heat load area information, create a heat load distribution map (e.g., using software such as MATLAB or Python matplotlib). The map uses a color gradient to represent cooling demand (e.g., blue = low demand, red = high demand). High heat load areas (Zones 2 and 3) are marked with flashing or border highlighting. Example map depicts: Zone 3 (211W) is the hottest, followed by Zone 2 (197W), and Zones 1 (180W) and 4 (168W) are cooler. Output a visual heat map for intuitive decision making.

[0032] Step 5: Prioritize flushing areas based on the heat load distribution map. Flushing (e.g., coolant flow) is prioritized based on the heat load distribution map. The rule is: areas with high heat loads are prioritized (due to the greater risk of temperature exceedance). Next, areas with high cooling requirements are prioritized. If multiple areas exceed thresholds, prioritize them based on the degree of temperature exceedance or cooling requirements.

[0033] Priority sorting in this example: Zone 3: Highest priority (temperature 88°C > 80°C, highest heat dissipation demand of 211W, greatest risk). Zone 2: Second highest priority (temperature 82°C > 80°C, heat dissipation demand of 197W). Zone 1: Medium priority (heat dissipation demand of 180W, but temperature within the specified range). Zone 4: Lowest priority (lowest heat dissipation demand of 168W, safe temperature). Flushing operation: In the cooling system, increase the coolant flow rate to zone 3 first, then adjust the flow rate to zone 2, and finally process other zones. Output: Flushing priority list (for example, in JSON format or as a control system instruction) to guide real-time cooling scheduling.

[0034] In this example, the process begins with a temperature gradient distribution map. The model calculates heat load and cooling requirements, identifying Zones 2 and 3 as high heat load areas (due to temperatures exceeding 80°C) and generating a heat load distribution map. Finally, based on the heat load distribution map, the flushing priority is determined: Zone 3 > Zone 2 > Zone 1 > Zone 4. This ensures that cooling resources are allocated preferentially to the hottest areas, preventing equipment damage and improving energy efficiency.

[0035] S105 , calculating the shortest path and flushing sequence between flushing points based on the priority using a dynamic programming method to obtain a flushing route.

[0036] In this embodiment, based on the priority, the area with the highest heat load is determined as the starting flushing point to obtain initial flushing position data; based on the initial flushing position data, the heat load distribution information of the adjacent area is obtained, and if the heat load of the adjacent area is lower than that of the current area and meets the preset heat load threshold, it is marked as the next flushing candidate point, and a set of candidate flushing points is determined; for the set of candidate flushing points, a dynamic programming method is used to calculate the path distance between each point, all possible paths from the current flushing point to the candidate point are obtained, and the shortest path is determined as the next flushing route; based on the shortest path data, the specific location information of the next flushing point is obtained to determine whether all areas have been included in the flushing route; if there are areas that are not included in the flushing route, the new area with the highest heat load is obtained as a new flushing point, and a new flushing order is determined; based on the new flushing order, the path optimization calculation is repeated to obtain the shortest path connection method for the remaining areas to obtain a complete flushing route.

[0037] For example, assume there is a system consisting of 5 areas: A, B, C, D, and E. The heat load values ​​of the areas (the higher the value, the greater the heat load) and the adjacent relationships are as follows: Heat load distribution: A: 100 (highest), B: 90, C: 85, D: 80, E: 75; preset heat load threshold: 50 (that is, only areas with a heat load ≥ 50 may be considered as candidate points). Area adjacent relationships (based on the graph structure, areas can be directly connected): A is adjacent to B and C, B is adjacent to A, C, and D, C is adjacent to A, B, and E, D is adjacent to B and E, and E is adjacent to C and D; path distance (used for dynamic programming calculations): The distance from A to B is 2, the distance from A to C is 3, the distance from B to C is 1, the distance from B to D is 4, the distance from C to E is 2, and the distance from D to E is 3; Optimization goal: Starting from the area with the highest heat load, generate a complete flushing route, ensure that the area with higher heat load is flushed first at each step, and minimize the path distance through dynamic programming. The optimization process is as follows:

[0038] Step 1: Determine the starting flushing point. Based on the priority, the area with the highest heat load is A (heat load 100). Initial flushing position data: A (starting point).

[0039] Step 2: Obtain neighboring area information and mark candidate points. Starting from the current point A, obtain neighboring areas B and C. Check the heat load of the adjacent areas: B: If the heat load is 90 < 100 of the current point A, and 90 ≥ the threshold of 50, mark it as a candidate point. C: If the heat load is 85 < 100 of the current point A, and 85 ≥ the threshold of 50, mark it as a candidate point. The candidate point set is: {B, C}.

[0040] Step 3: Use dynamic programming to calculate the shortest path to the candidate points. For the set of candidate points {B, C}, calculate all possible paths from the current point A to each candidate point: Path to B: A→B, distance 2 (unique path). Path to C: A→C, distance 3 (unique path). Of all possible paths, the shortest path is A→B (distance 2). Therefore, the next flushing point is B. Update the flushing route: A→B. Current location: B.

[0041] Step 4: Determine whether all areas have been included in the flushing route. Included areas: A and B. Unincluded areas: C, D, and E. There are still unincluded areas, so execute: Obtain the remaining unincluded area with the highest heat load as the new flushing point. The remaining area heat loads are: C (85), D (80), and E (75), with C (85) being the highest. New flushing point: C. New flushing sequence: C serves as the new starting point, and the optimization process will be repeated based on C.

[0042] Step 5: Move to the new flushing point and repeat the path optimization. First, you need to move from the current location B to the new flushing point C (because the new starting point C is an independent point and needs to be connected to the existing route). The shortest path from B to C is: B→C, distance 1. Update the flushing route: A→B→C (add the B→C segment). Current location: C. Now, starting from C, repeat steps 2-3: Obtain neighboring area information and mark candidate points. Starting from the current point C, obtain the neighboring areas: A (already included), B (already included), and E (not included). Check that E has a heat load of 75 < the current C's 85, and 75 ≥ the threshold of 50 → mark it as a candidate. The set of candidate points is: {E} (A and B are already included and are no longer considered). Dynamic programming is used to calculate the shortest path to the candidate points. For the candidate point set {E}, the path from the current point C to E is calculated. The path to E is: C→E, distance 2 (unique path). The shortest path is C→E (distance 2). Therefore, the next flushing point is E. Update flushing route: A→B→C→E. Current location: E.

[0043] Step 6: Determine whether all areas have been included. Included areas: A, B, C, E. Unincluded area: D. There are unincluded areas, so execute: Obtain the remaining unincluded area with the highest heat load as the new flushing point. The only remaining area is D (80). New flushing point: D. New flushing sequence: D is used as the new starting point.

[0044] Step 7: Move to the new flushing point and repeat the path optimization, moving from the current position E to the new flushing point D. The shortest path from E to D is: E→D, with a distance of 3 (direct connection). Update the flushing route: A→B→C→E→D (add the E→D segment). Current position: D. Now, repeat steps 2-3 starting from D: Step 2 (repeat): Obtain neighboring area information and mark candidate points. Starting from the current point D, obtain the neighboring areas: B (already included), E (already included). There are no unincluded neighboring areas, so the candidate point set is empty. Since the candidate point set is empty and all areas have been included (D is the last one), the process ends.

[0045] Final flushing route and results, complete flushing route: A→B→C→E→D, total path distance: 2(A→B)+1(B→C)+2(C→E)+3(E→D)=8, the order of inclusion of all areas: based on heat load priority and path optimization, the route covers all areas, and the heat load gradually decreases from high to low (100→90→85→75→80), but each step ensures that the heat load of the current point is higher than the next flushing point and meets the threshold.

[0046] Effect of heat load threshold: All areas have heat loads above the threshold of 50, so the threshold does not filter any points. If there is an area with a heat load of less than 50 (for example, suppose there is an area F with a heat load of 40), it will not be marked as a candidate point until its heat load is increased in some way or the threshold is adjusted. Application of dynamic programming: In step 3, when the candidate point set has multiple points (such as the initial {B, C}), dynamic programming calculates all possible paths and selects the shortest. When the set has only one point (such as {E}), the only path is directly selected. New starting point processing: When there are areas that are not included, the new starting point is the one with the highest remaining heat load and moves to that point (the path is calculated by direct connection or dynamic programming). This ensures that the route always prioritizes high-load areas while optimizing connecting paths. Route continuity: The final route is connected, but the introduction of new starting points may result in "jumps" (such as going directly from B to C), but path optimization ensures that the overall distance is minimized.

[0047] This example demonstrates how the process balances heat load priority and path efficiency. In real-world applications, the number of regions, heat load distribution, and network structure will affect the optimization details, but the core logic remains the same.

[0048] S106. Based on the flushing route and the heat load density distribution data of each area, a parameter adaptive adjustment algorithm is used to determine the flushing pressure, flow rate, and duration of each flushing point, and obtain a personalized flushing parameter configuration table; In this embodiment, based on a preset flushing scheme and combined with heat load density distribution data, if the heat load density in a certain area exceeds a preset regional threshold, the flushing pressure and flushing flow rate are adjusted through an adaptive algorithm to obtain optimized flushing parameters. Based on the optimized flushing parameters, the duration requirements are analyzed. If the heat load density continues to exceed the regional threshold, the flushing time is extended to determine the final flushing duration configuration. Using the final flushing time configuration and combining it with the flushing intensity requirements, the intensity adjustment value of each flushing point is obtained to generate intensity correction data. Through the intensity correction data, combined with the configuration parameters, a unified formatting processing tool is used to obtain a personalized flushing parameter configuration table.

[0049] For example, for the personalized parameter configuration of the flushing route plan and the heat load density of each area, the heat load density data of each area is first collected through the system. For example, the heat load density of area A is The threshold is set at 120 kW / m2. The system automatically determines that the area exceeds the threshold and needs to increase the flushing intensity. Then, the parameter adaptive adjustment algorithm is used to adjust the initial flushing pressure. Set to 5 bar, flow rate is 10 liters / minute, duration is 5 minutes, because the heat load density of area A exceeds the standard, the algorithm is based on the formula The new pressure is calculated as 5*(1+(120-100) / 100*0.2)= 5.2 bar, and the flow rate is proportionally adjusted to 10.4 liters / minute. The duration is extended by 20% to 6 minutes. The system records these parameters as the personalized configuration for area A. Further analysis shows that if the heat load density in area B is 90 kilowatts / square meter, which is lower than the threshold, the basic parameters will remain unchanged: pressure 5 bar, flow rate 10 liters / minute, and duration 5 minutes to ensure reasonable resource allocation. The system then integrates all area data and generates a flushing parameter configuration table, which contains fields such as area number, heat load density, adjusted pressure, flow rate, and time. For example, the record for area A is {area A, 120, 5.2 bar, 10.4 liters / minute, 6 minutes}, and the record for area B is {area B, 90, 5 bar, 10 liters / minute, 5 minutes}. To ensure strict logic, the system also combines a flushing route optimization algorithm to give priority to areas with high heat loads, ensuring that the flushing equipment covers area A in sequence before entering area B, reducing energy consumption from frequent parameter adjustments of the equipment, and ultimately verifying the effectiveness of the adjusted parameters in reducing heat loads through data analysis. For example, after flushing area A, the heat load density dropped to 95 kilowatts / square meter, proving that the parameter adjustment was reasonable. The system automatically updates the database to provide a reference basis for subsequent flushing.

[0050] S107, based on the flushing route and flushing parameter configuration table to perform automated flushing operations, real-time monitoring of temperature changes in each area, if the temperature drop rate is lower than the expected value, the flushing parameters are dynamically adjusted; In this embodiment, see the attached Figure 3 , obtain the flushing parameter data from the flushing parameter configuration table through the intelligent control system, initialize the automated operation process according to the obtained parameters, and obtain a preliminary operation execution plan; according to the preliminary operation execution plan, start the automated flushing operation, and at the same time monitor the temperature change data of each area in real time through the sensor network to obtain real-time records of temperature changes; calculate the temperature drop rate of each area based on the obtained real-time records of temperature changes, and if the calculated drop rate is lower than the preset expected value, trigger the parameter adjustment mechanism; through the parameter adjustment mechanism, dynamically update the flushing parameter data, and use the updated parameters to reconfigure the automated operation process to obtain the adjusted operation execution plan; according to the adjusted operation execution plan, continue to execute the flushing operation, and at the same time obtain the updated temperature change data through the sensor network to determine whether the temperature drop rate reaches the expected value; if the updated temperature drop rate still does not reach the expected value, the historical temperature change data and flushing parameter data are processed through the regression analysis model to determine the optimal adjustment parameters.

[0051] For example, an intelligent control system is used to automate flushing operations. Initial parameters are first set according to the flushing parameter configuration table. For example, the flushing water flow rate is set to 5.0 liters / second, the flushing frequency is set to 2 times per minute, and the flushing duration is 10 minutes. The system automatically drives the flushing equipment to start according to these parameters to ensure full coverage of the target area. Next, the temperature changes in each area are monitored in real time, and data is collected every 5 seconds using temperature sensors deployed in the area. Assuming that the initial temperature of a certain area is 50.0 degrees Celsius and the target temperature is reduced to 30.0 degrees Celsius, the expected cooling rate is set to 2.0 degrees Celsius per minute. The system calculates the actual cooling rate through an algorithm. For example, if the temperature drops from 50.0 to 42.0 degrees Celsius within 5 minutes, the actual rate is calculated to be only 1.6 degrees Celsius / minute, which is lower than the expected value. In response to this situation, the system automatically triggers the dynamic adjustment mechanism and increases the water flow rate to 6.5 liters / second based on the preset algorithm. At the same time, the flushing frequency is increased to 3 times per minute. Through simulation analysis, it is predicted that the cooling rate after adjustment can reach 2.2 degrees Celsius / minute, which is close to or exceeds the expected value. Subsequently, the system continuously monitors the effect of the adjustment. If the temperature drops to 31.0 degrees Celsius after 10 minutes, which is close to the target value, the final cooling effect is recorded as a drop of 19.0 degrees Celsius, and a heat map is generated through the temperature distribution analysis algorithm to show the temperature distribution status of each area. For example, the regional temperature is 30.5 degrees Celsius and the edge area is 32.0 degrees Celsius, with a deviation of less than 2.0 degrees Celsius, indicating that the equipment temperature is evenly distributed. The system automatically saves the data and generates optimization suggestions, such as fine-tuning the flushing angle for the edge area to further balance the temperature distribution and ensure that subsequent operations are more efficient. If the updated temperature drop rate still does not reach the expected value, the historical temperature change data and flushing parameter data are processed through the regression analysis model to determine the optimal adjustment parameters. Through the automated processing of the above entire process, a complete closed-loop logic is formed from parameter configuration to effect evaluation.

[0052] The above description is merely a preferred embodiment of the present application and an illustration of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to the technical solutions formed by the specific combination of the above-mentioned technical features, but also encompasses other technical solutions formed by any combination of the above-mentioned technical features or their equivalents without departing from the concept of this application. For example, a technical solution formed by replacing the above-mentioned features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A flushing and cooling method with automatic route identification, characterized in that: The method comprises the following steps: S101. Use 3D laser scanning technology to scan the surface of the device in all directions to obtain a 3D model of the device; S102. Based on the three-dimensional model, use a material identification method to analyze the material property information of the device surface, and obtain the thermal conductivity of the device surface based on the material property information; S103. Monitor the temperature distribution of each part of the equipment in real time during operation, and use temperature data fusion technology to integrate the measurement results of multiple sensors to obtain a temperature gradient distribution map of the equipment surface; S104. According to the temperature gradient distribution map and thermal conductivity, use the heat load calculation model to analyze the heat load data, heat load density distribution data and heat dissipation requirements of each area, and determine the flushing priority based on the heat load distribution map and heat dissipation requirements; S105. Based on the priority, calculate the shortest path and flushing order between each flushing point through a dynamic programming method to obtain a flushing route; S106. According to the flushing route and the heat load density distribution data of each area, use a parameter adaptive adjustment algorithm to determine the flushing pressure, flow rate and duration of each flushing point, and obtain a personalized flushing parameter configuration table; S107. Perform automated flushing operations based on the flushing route and the flushing parameter configuration table, monitor the temperature changes of each area in real time, and dynamically adjust the flushing parameters if the temperature drop rate is lower than the expected value.

2. The method according to claim 1, characterized in that The step S101 includes: Use 3D laser technology to fully detect the surface of the equipment, collect original geometric shape information and spatial coordinate data, and obtain the initial scanning data set; The point cloud data processing algorithm is used to denoise the initial scan data set, remove noise interference, and obtain a cleaned point cloud data set; Perform registration operations on the cleaned point cloud dataset, adjust the data consistency of different scanning angles, and determine a unified point cloud registration result; Based on the unified point cloud registration results, a 3D model of the equipment is constructed, key structural feature information is extracted, and preliminary 3D model data is obtained; By optimizing the preliminary 3D model data and correcting the geometric deviation in the 3D model, a refined 3D model data set is obtained; If the structural feature information of some areas in the refined 3D model data set is incomplete, the missing parts are supplemented by an interpolation algorithm to obtain the final 3D model data.

3. The method according to claim 1, characterized in that The step S102 includes: Based on the three-dimensional model data of the device, an initial model is generated by a modeling tool, and a mesh is performed on the surface of the initial model to obtain surface geometric feature data; Based on the surface geometric feature data, a material identification method is used to analyze the material properties of each grid area on the surface to obtain material property information of each area, and a material distribution map is output based on the material property information; According to the material distribution map and combined with the pre-established material thermal conductivity database, the thermal conductivity of each area is obtained.

4. The method according to claim 1, wherein The step S103 includes: Use infrared thermal imaging sensors to collect real-time data from various parts of the equipment during operation to obtain a set of original temperature data; Based on the original temperature data set, data preprocessing technology is used to denoise and calibrate the collected information to obtain a corrected temperature data set; For the corrected temperature data set, the measurement results of multiple sensors are integrated to construct a unified temperature distribution matrix; Based on the temperature distribution matrix, the temperature data of different sensors are weighted according to their location importance and sensor accuracy to obtain the overall reference temperature, and the temperature of the uncovered area is estimated using an interpolation algorithm to calculate the complete temperature gradient distribution map of the device surface.

5. The method according to claim 1, wherein The step S104 includes: The temperature distribution data of each area is obtained through the temperature gradient distribution diagram, and the heat load distribution data and heat load density distribution data of each area are obtained using the pre-established heat load calculation model; Based on the heat load distribution results and the thermal conductivity in the material property database, the heat dissipation requirements of each area are obtained; If the temperature of a region in the heat dissipation demand exceeds a preset threshold, the region is determined to be a high heat load region, and its location information is recorded and a heat load distribution map is generated; Prioritize washdown areas based on heat load profiles and cooling requirements.

6. The method according to claim 1, characterized in that The step S105 includes: Based on the priority, determine the area with the highest heat load as the starting flushing point to obtain initial flushing position data; Based on the initial flushing position data, the heat load distribution information of the adjacent area is obtained. If the heat load of the adjacent area is lower than that of the current area and meets the preset heat load threshold, it is marked as the next flushing candidate point, and the candidate flushing point set is determined; For the candidate flushing point set, the dynamic programming method is used to calculate the path distance between each point, obtain all possible paths from the current flushing point to the candidate point, and determine the shortest path as the next flushing route; Based on the shortest path data, obtain the specific location information of the next flushing point and determine whether all areas have been included in the flushing route; If there are areas that are not included in the flushing route, the new area with the highest heat load is obtained as the new flushing point and a new flushing order is determined; According to the new flushing order, the path optimization calculation is repeated to obtain the shortest path connection method of the remaining areas and obtain the complete flushing route.

7. The method according to claim 1, characterized in that The step S106 includes: Based on the preset flushing scheme and combined with the heat load density distribution data, if the heat load density of a certain area exceeds the preset regional threshold, the flushing pressure and flushing flow rate are adjusted through an adaptive algorithm to obtain the optimized flushing parameters; Analyze the duration requirements for the optimized flushing parameters. If the heat load density continues to be higher than the regional threshold, extend the flushing time to determine the final flushing duration configuration. Using the final flushing time configuration and combining it with the flushing intensity requirements, the intensity adjustment value of each flushing point is obtained to generate intensity correction data; By combining intensity correction data with configuration parameters and using a unified formatting tool, a personalized flushing parameter configuration table can be obtained.

8. The method according to claim 1, characterized in that The step S107 includes: The intelligent control system obtains flushing parameter data from the flushing parameter configuration table, initializes the automated operation process based on the obtained parameters, and obtains a preliminary operation execution plan; According to the preliminary operation execution plan, the automated flushing operation is started. At the same time, the temperature change data of each area is monitored in real time through the sensor network to obtain real-time records of temperature changes; Based on the real-time records of temperature changes, the temperature drop rate of each area is calculated. If the calculated drop rate is lower than the preset expected value, the parameter adjustment mechanism is triggered; Through the parameter adjustment mechanism, the flushing parameter data is dynamically updated, and the updated parameters are used to reconfigure the automated operation process to obtain the adjusted operation execution plan; According to the adjusted operation execution plan, the flushing operation is continuously performed, and the updated temperature change data is obtained through the sensor network to determine whether the temperature drop rate reaches the expected value; If the updated temperature drop rate still does not reach the expected value, the historical temperature change data and flushing parameter data are processed through a regression analysis model to determine the optimal adjustment parameters.

9. A flushing and cooling system with automatic route identification, characterized in that: The system comprises: The scanning module uses 3D laser scanning technology to perform a full-scale scan of the equipment surface to obtain a 3D model of the equipment; a material analysis module, which analyzes material property information of the device surface using a material identification method based on the three-dimensional model, and obtains a thermal conductivity coefficient of the device surface based on the material property information; The real-time monitoring module monitors the temperature distribution of various parts of the equipment in real time during operation, and uses temperature data fusion technology to integrate the measurement results of multiple sensors to obtain a temperature gradient distribution map of the equipment surface; the priority determination module uses a heat load calculation model to analyze the heat load data, heat load density distribution data and heat dissipation requirements of each area according to the temperature gradient distribution map and thermal conductivity, and determines the priority of flushing based on the heat load distribution map and heat dissipation requirements; the flushing route acquisition module calculates the shortest path and flushing order between each flushing point through a dynamic programming method based on the priority to obtain the flushing route; the flushing parameter acquisition module uses a parameter adaptive adjustment algorithm to determine the flushing pressure, flow rate and duration of each flushing point according to the flushing route and the heat load density distribution data of each area, and obtains a personalized flushing parameter configuration table; the dynamic adjustment module performs automated flushing operations based on the flushing route and the flushing parameter configuration table, monitors the temperature changes of each area in real time, and dynamically adjusts the flushing parameters if the temperature drop rate is lower than the expected value.

10. A flushing and cooling device with automatic route identification, characterized in that: The device includes: a memory, a processor, and a flushing and cooling program with automatic route identification stored in the memory and executable on the processor, wherein the flushing and cooling program with automatic route identification is configured to implement the flushing and cooling method with automatic route identification as described in any one of claims 1 to 8.

Images