Equipment cooling control system and method based on visual temperature measurement

Abstract

The invention relates to the technical field of intelligent temperature control of industrial equipment, in particular to an equipment cooling control system and method based on visual temperature measurement, and the system comprises an infrared thermal imaging module, a hot spot tracking module, a temperature modeling module, a control decision module, a cooling execution module and a safety protection module. The method comprises the following steps: acquiring a full-visual-angle thermal imaging sequence of the surface of equipment through an infrared camera, identifying a high-temperature area and constructing a three-dimensional temperature field model, dynamically calculating a power regulation quantity according to a difference value between Tmax and Tset, starting directional cooling based on temperature gradient distribution, and triggering predictive control when detecting that the temperature change rate is abnormal. According to the equipment cooling control system and method based on visual temperature measurement, accurate sensing of a temperature field and intelligent temperature control decision are achieved through visual temperature measurement, and the system and method are suitable for scenes needing accurate heat dissipation guarantee such as server rooms, power equipment and numerical control machine tools.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent temperature control of industrial equipment, in particular to a device cooling control system and method based on visual temperature measurement. BACKGROUND

[0002] Current device cooling mainly relies on contact temperature sensors (such as thermocouples and thermal resistors), which have the following defects: limited coverage: single-point measurement cannot reflect the overall temperature distribution of the device, and local overheating areas are easily missed; response delay: sensor contact leads to temperature feedback lag, affecting the real-time of temperature control. Existing infrared temperature measurement schemes are mainly used for temperature monitoring and are not deeply coupled with the control system, lacking a dynamic temperature control closed-loop mechanism. Based on this, the present application provides a device cooling control system and method based on visual temperature measurement. SUMMARY

[0003] The present application overcomes the shortcomings of the prior art and provides a device cooling control system and method based on visual temperature measurement, which realizes accurate temperature field perception and intelligent temperature control decision-making through visual temperature measurement, and is suitable for scenarios such as server rooms, power equipment, and numerical control machine tools that require accurate heat dissipation protection.

[0004] The device cooling control system based on visual temperature measurement in the present application specifically includes:

[0005] An infrared thermal imaging module configured with an infrared camera with a wavelength range of 8-14μm to collect full-view infrared images of the device surface at a frequency of ≥5Hz;

[0006] A hotspot tracking module that uses a YOLOv3 model to identify high-temperature areas in the image where the temperature exceeds a first threshold value;

[0007] A temperature modeling module that generates a two-dimensional / three-dimensional temperature field thermal map of the device surface based on the infrared image and calculates the temperature gradient;

[0008] A control decision module that receives a target temperature threshold T_set, real-time temperature field data, and device load rate, and executes the following strategies: when the maximum temperature point T_max>T_set, proportionally increase the switching power; when the temperature gradient>10℃ / cm, start directional cooling;

[0009] A cooling execution module that supports multiple types of actuators such as PWM adjustment grouping, variable frequency water pumps, and semiconductor refrigeration plates; and

[0010] A safety protection module that performs device load reduction or emergency shutdown when T_max exceeds a safety threshold.

[0011] Further, the control decision module further includes a predictive control unit: when a temperature change rate>5℃ / s is detected, pre-start the cooling execution module.

[0012] Further, the temperature modeling module uses Laplacian operator for temperature gradient analysis, which can identify local overheating areas with an area <1mm 2 , with a positioning accuracy of ±0.5mm.

[0013] Further, the safety protection module sets a two-level fuse mechanism, with a first response: performing proportional power adjustment when T_max>T_set; and a second response: triggering device load reduction or emergency shutdown when T_max> safety threshold.

[0014] Further, the control decision module further includes a predictive control unit: when detecting that the temperature change rate >5°C / s, pre-starting the cooling execution module.

[0015] In a second aspect, the present application further provides a control method suitable for the device cooling control system according to any one of the first aspect, specifically including the following steps:

[0016] S1. Obtaining a full-view thermal imaging sequence of the device surface through an infrared camera;

[0017] S2. Identifying high-temperature areas and constructing a three-dimensional temperature field model;

[0018] S3. Dynamically calculating the power adjustment amount according to the difference between T_max and T_set;

[0019] S4. Starting directional cooling based on temperature gradient distribution;

[0020] S5. Triggering predictive control when detecting abnormal temperature change rate.

[0021] Further, the implementation of directional cooling in step S4 includes: starting semiconductor refrigeration sheet precise cooling on local hot spot areas, and / or starting variable frequency water pump to enhance liquid cooling flow on linear high-temperature zones.

[0022] The present application has the following advantages:

[0023] The present application realizes 100% dead angle-free temperature scanning of the device surface through infrared thermal imaging technology, completely solves the problem of blind area of single-point monitoring of traditional contact sensors, and can identify micron-level local hot spots with a positioning accuracy of ±0.5mm through temperature gradient analysis algorithm; realizes precise monitoring of the temperature field of the device surface with full-view and no blind area, with a response time ≤0.5s, dynamically adjusts the local cooling power through temperature distribution, reduces invalid energy consumption by more than 20%, and improves system safety through a double-threshold protection mechanism; the system has a predictive maintenance function and is suitable for various industrial and data center cooling scenarios. DETAILED DESCRIPTION

[0024] The application is based on a device cooling control system of visual temperature measurement, comprising:

[0025] An infrared thermal imaging module is configured with an infrared camera with a wavelength range of 8-14pm to collect full-view infrared images of the device surface at a frequency of ≥5Hz;

[0026] A hotspot tracking module uses a YOLOv3 model to identify high-temperature areas in the image with temperatures exceeding a first threshold value (85℃);

[0027] A temperature modeling module generates a two-dimensional / three-dimensional temperature field thermal map of the device surface based on the infrared images and calculates the temperature gradient. The temperature modeling module uses a Laplacian operator for temperature gradient analysis, which can identify local overheating areas with an area <1mm2, with a positioning accuracy of ±0.5mm;

[0028] A control decision module receives a target temperature threshold T_set, real-time temperature field data, and device load rate, and executes the following strategies: when the maximum temperature point T_max>T_set, the switching power is proportionally increased; when the temperature gradient >10℃ / cm, directional cooling is started; the control decision module also includes a predictive control unit: when a temperature change rate >5℃ / s is detected, the cooling execution module is pre-started; the control decision module also includes a predictive control unit: when a temperature change rate >5℃ / s is detected, the cooling execution module is pre-started;

[0029] A cooling execution module supports multiple types of actuators such as PWM adjustment grouping, variable frequency water pumps, and semiconductor refrigeration plates; and

[0030] A safety protection module executes device load reduction or emergency shutdown when T_max exceeds a safety threshold, and specifically sets a two-level fuse mechanism: first response: when T_max>T_set, proportionally adjust the power; second response: when T_max> safety threshold, trigger device load reduction or emergency shutdown.

[0031] The device cooling control method based on visual temperature measurement described in the application comprises the following steps:

[0032] S1. Obtain a full-view thermal imaging sequence of the device surface through an infrared camera;

[0033] S2. Identify high-temperature areas and build a three-dimensional temperature field model;

[0034] S3. Dynamically calculate the power adjustment amount according to the difference between T_max and T_set;

[0035] S4. Start directional cooling based on temperature gradient distribution, start semiconductor refrigeration plate precise cooling for local hotspot areas, and / or start variable frequency water pump to enhance liquid cooling flow for linear high-temperature zones;

[0036] S5. Triggering predictive control when temperature rate of change anomaly is detected.

[0037] The present application realizes 100% dead angle-free temperature scanning of the surface of the equipment by infrared thermal imaging technology, and completely solves the problem of the blind area of the single-point monitoring of the traditional contact sensor; combined with a temperature gradient analysis algorithm (such as a Laplacian operator), a micrometer-level local hotspot (such as a <1mm 2 overheated area) can be identified, and the positioning accuracy is ±0.5mm.

[0038] Secondly, the present application has significant energy saving and consumption reduction, and the cooling strategy is strongly coupled with the temperature distribution: only the high-temperature area is directionally reinforced cooling (such as fan focused air supply), and the low-temperature area is automatically operated at low power (such as turning off the redundant fan); the measured data shows that in the data center scenario, the system reduces the invalid cooling energy consumption by 23% to 28% (the energy consumption of the traditional temperature control system accounts for as high as 40% of the total energy consumption).

[0039] In addition to the above embodiments, the present application can also have other implementation manners. Any technical solution formed by equivalent replacement or equivalent transformation falls within the protection scope required by the present application.

Claims

1. A device cooling control system based on visual temperature measurement, characterized by, Comprise: An infrared thermal imaging module configured with an infrared camera with a wavelength range of 8-14 pm to collect full-view infrared images of the device surface at a frequency of ≥5 Hz; A hotspot tracking module that uses a YOLOv3 model to identify high-temperature areas in the image where the temperature exceeds a first threshold value; A temperature modeling module that generates a two-dimensional / three-dimensional temperature field thermal map of the device surface based on the infrared images and calculates the temperature gradient; A control decision module that receives the target temperature threshold T_set, real-time temperature field data, and device load rate, and executes the following strategies: when the maximum temperature point T_max > T_set, proportionally increase the switching power; when the temperature gradient > 10℃ / cm, start directional cooling; A cooling execution module that supports multiple types of actuators such as PWM adjustment grouping, variable frequency water pumps, and semiconductor refrigeration plates; And A safety protection module that executes device load reduction or emergency shutdown when T_max exceeds a safety threshold.

2. The device cooling control system of claim 1, wherein, The control decision module also includes a predictive control unit: when a temperature change rate > 5℃ / s is detected, pre-start the cooling execution module.

3. The device cooling control system of claim 1, wherein, The temperature modeling module adopts Laplacian operator for temperature gradient analysis, can identify local overheating area with area <1mm 2 , positioning accuracy up to ±0.5mm.

4. The device cooling control system of claim 1, wherein, The safety protection module sets a two-level fuse mechanism: the first response: when T_max > T_set, execute proportional power adjustment; the second response: when T_max > the safety threshold, trigger device load reduction or emergency shutdown.

5. The device cooling control system of claim 1, wherein, The control decision module also includes a predictive control unit: when a temperature change rate > 5℃ / s is detected, pre-start the cooling execution module.

6. A control method of a cooling control system of an apparatus according to any one of claims 1 to 5, characterized by, Comprise the following steps: S1. Obtain a full-view thermal imaging sequence of the device surface through an infrared camera; S2. Identify high-temperature areas and build a three-dimensional temperature field model; S3. Dynamically calculate the power adjustment amount according to the difference between T_max and T_set; S4. Start directional cooling based on the temperature gradient distribution; S5. Trigger predictive control when an abnormal temperature change rate is detected.

7. The control method according to claim 6, characterized by The implementation of directional cooling in step S4 includes: starting semiconductor refrigeration plates for precise cooling of local hotspot areas, and / or starting variable frequency water pumps to enhance liquid cooling flow for linear high-temperature zones.