Server cooling gas leakage detection method and system based on closed-loop control

Abstract

The invention discloses a server cooling gas leakage detection method and system based on closed-loop control. According to the invention, the gas sensor is matched with the temperature sensor to carry out server cooling gas leakage detection, and the method can achieve the following technical effects: based on the characteristics that the gas sensor is sensitive to common cooling gas, high in sensitivity and high in response speed, the method can be applied to detection scenes aiming at different cooling gases; according to the method, leakage detection with the minimum detectable quantity of 0.1 L / min can be achieved, the application scene is wide, and the detection precision and the response speed are remarkably improved; after the pressure sensor is removed, the hardware cost is greatly reduced; the gas sensor is small in size and is more suitable for compact scenes such as edge computing nodes; temperature compensation is carried out on the resistance value of the gas sensor according to the environment temperature change, temperature drift interference can be eliminated through temperature compensation, the anti-interference performance is high, the environment adaptability is high, and therefore the detection precision is effectively improved, and the false alarm rate is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of server cooling, and particularly relates to a server cooling gas leakage detection method and system based on closed-loop control. BACKGROUND

[0002] At present, the leakage detection of the server closed-loop cooling system mainly adopts the following technologies:

[0003] 1. Pressure monitoring method: the pressure change of the pipeline is detected through a pressure sensor, but the response speed is slow (usually more than 30 seconds are needed), and the tiny leakage cannot be detected.

[0004] 2. Ultrasonic detection method: the ultrasonic sensor is used to capture the leakage sound wave, but the method is easily disturbed by environmental noise, and the false positive rate is high.

[0005] 3. Infrared imaging method: the device cost is high, and the method is not suitable for large-scale deployment.

[0006] In summary, the existing server cooling system gas leakage detection technology has the problems of insufficient sensitivity, poor environmental adaptability and high false positive rate. SUMMARY

[0007] The technical problem to be solved by the present application is to provide a server cooling gas leakage detection method and system based on closed-loop control, so as to solve the problems of insufficient sensitivity, poor environmental adaptability and high false positive rate of the existing server cooling system gas leakage detection technology.

[0008] To solve the above technical problems, the present application adopts the following technical solutions:

[0009] The first aspect of the present application provides a server cooling gas leakage detection method based on closed-loop control, comprising the following steps: S10, collecting the resistance value of a gas sensor and the environmental temperature data in real time; S20, compensating the resistance value of the gas sensor according to the change of the environmental temperature by using different compensation strategies; S30, performing leakage characteristic analysis based on the collected resistance value of the gas sensor and the resistance value of the gas sensor after temperature compensation, and determining whether leakage occurs, if yes, executing step S40, if not, returning to step S10; S40, taking corresponding measures according to the leakage level, and jumping back to step S10.

[0010] The second aspect of the present application provides a closed-loop control based server cooling gas leakage detection method, comprising the following steps: S110, collecting the resistance value of a gas sensor and the ambient temperature data through a first sampling frequency; S120, temperature compensating the resistance value of the gas sensor according to the change of the ambient temperature by using different compensation strategies; S130, judging whether the resistance value of the gas sensor is abnormal, if yes, executing step S140, if not, returning to step S110; S140, continuously sampling the resistance value of the gas sensor and the ambient temperature data through a second sampling frequency; S150, judging whether the resistance value of the gas sensor obtained by continuous sampling is continuously changed, if yes, executing step S160, if not, returning to step S110; S160, temperature compensating the resistance value of the gas sensor obtained by continuous sampling according to the change of the ambient temperature by using different compensation strategies; S170, performing leakage characteristic analysis based on the collected resistance value of the gas sensor and the resistance value of the gas sensor obtained after temperature compensation, and judging whether leakage occurs, if yes, executing step S180, if not, returning to step S110; S180, taking corresponding measures according to the leakage level, and jumping back to step S140.

[0011] The third aspect of the present application provides a closed-loop control based server cooling gas leakage detection system, comprising: a gas sensor for detecting whether the server cooling gas leaks; a temperature sensor for the change of the ambient temperature around the gas sensor; a BMC chip, the BMC chip is connected with the gas sensor and the temperature sensor respectively; the BMC chip is used to realize each step of the closed-loop control based server cooling gas leakage detection method.

[0012] The present application adopts a gas sensor to cooperate with a temperature sensor to detect the leakage of server cooling gas, and can achieve the following beneficial technical effects:

[0013] 1. Based on the characteristics that the gas sensor is sensitive to common cooling gas, the present application can be applied to detection scenes for different cooling gases, and the application scenarios are wide;

[0014] 2. Based on the characteristics that the gas sensor has high sensitivity and fast response speed, the present application can realize leakage detection with a minimum detectable amount of 0.1 L / min, and the detection accuracy and response speed are significantly improved;

[0015] 3. The gas sensor is used to replace the pressure sensor for leakage detection, which greatly reduces the hardware cost after removing the pressure sensor; and the gas sensor has small size and is suitable for high-density deployment, and is more suitable for application in compact scenes such as edge computing nodes;

[0016] 4. The temperature sensor collects environmental temperature data, and the resistance of the gas sensor is temperature-compensated according to the change of the environmental temperature. The temperature compensation can eliminate the temperature drift interference, solve the false alarm problem caused by the baseline drift of the gas sensor in a high-temperature environment, has strong anti-interference and high environmental adaptability, and thus improves the detection precision and reduces the false alarm rate. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 FIG. 1 is a structural schematic diagram of a server cooling gas leakage detection system based on closed-loop control in an embodiment of the present application;

[0018] Figure 2 FIG. 2 is a flowchart of a server cooling gas leakage detection method based on closed-loop control in an embodiment of the present application;

[0019] Figure 3 FIG. 3 is a flowchart of a server cooling gas leakage detection method based on closed-loop control in another embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to make the ordinary skilled in the art more clearly understand the purpose, technical scheme and advantages of the present application, the present application is further described below in conjunction with the drawings and embodiments.

[0021] As shown in Figure 1 In an embodiment of the present application, the server cooling gas leakage detection system based on closed-loop control includes a gas sensor, a temperature sensor and a BMC (Baseboard Management Controller) chip. The gas sensor is installed on or near the pipeline of the server cooling system, and is used to detect whether the server cooling gas leaks. The temperature sensor is installed on or near the pipeline of the server cooling system, and is used to detect the change of the environmental temperature around the gas sensor. The BMC chip is connected with the gas sensor and the temperature sensor respectively, and receives the sensing signals of the gas sensor and the temperature sensor.

[0022] The gas sensor is connected with a resistance with a resistance value of 10KΩ in series, one end of which is connected with 5V direct current, and the other end is grounded, thereby forming a cooling gas leakage detection circuit, a junction between the gas sensor and the resistance is connected with an ADC (Analog-to-Digital Converter) chip, the gas sensor and the resistance in series form a voltage division circuit to divide the 5V direct current, and a voltage signal after voltage division is output to the ADC chip, the ADC chip converts the voltage signal after voltage division into a digital signal and transmits the digital signal to the BMC chip, and the BMC chip converts the digital voltage signal output by the ADC chip into a resistance value of the gas sensor, that is, obtains a cooling gas leakage detection value.

[0023] In the embodiment of the present application, the gas sensor adopts a gas sensor with a model of MP510C. The MP510C is a semiconductor gas sensor, which has the following characteristics: 1, sensitive to common cooling gas (such as R134a, R404a, etc.); 2, fast response speed: response time < 10 seconds, which is more than 3 times faster than the pressure monitoring method; 3, high sensitivity: the resistance value decreases significantly with the increase of the concentration of cooling gas (such as R134a, nitrogen, etc.); 4, small volume: suitable for high-density deployment. Of course, in other embodiments of the present application, other models of gas sensors such as MQ2 or TGS2600 can be used instead of the gas sensor with the model of MP510C to detect the leakage of cooling gas.

[0024] In the embodiment of the present application, the temperature sensor adopts a temperature sensor with a model of TPM112 to collect the ambient temperature, which is used to detect the change of ambient temperature around the gas sensor. Of course, in other embodiments of the present application, other models of temperature sensors such as DS18B20 or TM75BD can be used instead of the temperature sensor with the model of TPM112 to detect the change of ambient temperature.

[0025] In the embodiment of the present application, the BMC chip adopts a chip with a model of AST2600; of course, in other embodiments of the present application, other models of BMC chips can be used according to actual conditions. The BMC chip receives the sensor signal of the temperature sensor, and adopts different compensation strategies according to the change of ambient temperature to compensate the resistance value of the gas sensor. Specifically, the BMC chip judges whether the change of ambient temperature in adjacent sampling periods exceeds 2 degrees, if not, the temperature compensation is zero; if yes, a second-order polynomial compensation model Rcomp=Rraw×1 / [1+α(T-Tref)+β(T-Tref) 2Temperature compensation is performed on the resistance value of the gas sensor, wherein Rraw is the original resistance value of the gas sensor, Rcomp is the compensated original resistance value of the gas sensor, T is the current ambient temperature, Tref is the reference temperature, and a and b are temperature coefficients.

[0026] After the temperature compensation of the resistance value of the gas sensor is completed, the BMC chip performs leakage feature analysis based on the resistance value of the gas sensor to determine whether leakage occurs, and if so, corresponding measures are taken according to the leakage level, and if not, the resistance value of the gas sensor and the ambient temperature data are continuously collected to continuously monitor the server cooling gas leakage state.

[0027] The server cooling gas leakage detection system based on closed-loop control in the embodiment of the application uses the gas sensor MP510C in cooperation with the temperature sensor TPM112 to perform server cooling gas leakage detection, and can achieve the following beneficial technical effects:

[0028] 1. Based on the characteristics that the gas sensor MP510C is sensitive to common cooling gases, the application can be applied to detection scenes for different cooling gases, and the application scenarios are wide.

[0029] 2. Based on the characteristics that the gas sensor MP510C has high sensitivity and fast response speed, the application can realize leakage detection with a minimum detectable amount of 0.1 L / min, which is higher than the minimum detectable amount of 0.5 L / min of the traditional method, improves the detection rate of small leakage, and significantly improves the detection accuracy; and the response time for 0.1 L / min leakage is shortened from 30s to 8s, the sensitivity is improved by 375%, and the response speed is significantly improved.

[0030] 3. The gas sensor MP510C is used to replace the pressure sensor for leakage detection, and the hardware cost is reduced by 40% after removing the pressure sensor, which greatly reduces the cost; and the gas sensor has small size and is suitable for high-density deployment, and is more suitable for application in compact scenes such as edge computing nodes.

[0031] 4. The temperature sensor TPM112 is used to collect ambient temperature data, the resistance value of the gas sensor is compensated according to the change of the ambient temperature, the temperature compensation can eliminate the temperature drift interference, can solve the false alarm problem caused by the baseline drift of the gas sensor MP510C in a high-temperature environment, has strong anti-interference performance and high environmental adaptability, thereby improving the detection accuracy and reducing the false alarm rate.

[0032] Based on the above-mentioned server cooling gas leakage detection system based on closed-loop control, the application further provides a server cooling gas leakage detection method based on closed-loop control.

[0033] As Figure 2As shown, in one embodiment of the present application, the server cooling gas leakage detection method based on closed-loop control comprises steps S10 to S40:

[0034] S10, real-time acquisition of gas sensor resistance and ambient temperature data.

[0035] The present application adopts a gas sensor for cooling gas leakage detection, which is connected in series with a resistance of 10KΩ to form a voltage dividing circuit to divide 5V DC voltage, and the voltage signal after voltage division is output to the ADC chip. The voltage signal after voltage division of the ADC chip is transmitted to the BMC chip after analog-to-digital conversion. The BMC chip converts the digital voltage signal output by the ADC chip into the resistance of the gas sensor. In an embodiment of the present application, the gas sensor adopts a gas sensor of model MP510C, which is a semiconductor gas sensor with the following characteristics: 1. sensitive to common cooling gases (such as R134a, R404a, etc.); 2. fast response speed: response time <10 seconds, which is 3 times faster than the pressure monitoring method; 3. high sensitivity: resistance decreases significantly with increasing concentration of cooling gas (such as R134a, nitrogen, etc.); 4. small volume: suitable for high-density deployment. Of course, in other embodiments of the present application, other models of gas sensors such as MQ2 or TGS2600 can be used instead of the gas sensor of model MP510C for gas leakage detection.

[0036] In an embodiment of the present application, the temperature sensor adopts a temperature sensor of model TPM112; of course, in other embodiments of the present application, other models of temperature sensors such as DS18B20 or TM75BD can be used instead of the temperature sensor of model TPM112 for ambient temperature change detection.

[0037] S20, temperature compensation of the resistance of the gas sensor according to the change of ambient temperature.

[0038] The step S20 specifically comprises steps S21 to S22:

[0039] S21, judging whether the change of ambient temperature in the adjacent sampling period exceeds 2 degrees, if yes, executing step S22, if not, the temperature compensation is zero;

[0040] S22, temperature compensation of the resistance of the gas sensor using a second-order polynomial compensation model Rcomp = Rraw x 1 / [1+α(T-Tref)+β(T-Tref) 2 ]。

[0041] The temperature compensation calculation process is as follows: 22.1 establish a temperature-resistance relationship model

[0042] The temperature characteristic data of MP510C without leakage is shown in the following table:

[0043] Temperature (°C) Reference resistance value (kΩ) Temperature coefficient α 25 12 -0.0032 / ℃ 35 10.8 -0.0035 / ℃ 45 9.6 -0.0038 / ℃

[0044] 22.2 Establish compensation formula

[0045] The application adopts a second-order polynomial compensation model to compensate the resistance value of the gas sensor:

[0046] Rcomp = Rraw x 1 / [1+α(T-Tref)+β(T-Tref) 2 ]

[0047] Wherein:

[0048] Rraw is the original resistance value of the gas sensor;

[0049] Rcomp is the compensated resistance value of the gas sensor;

[0050] T is the current environmental temperature;

[0051] Tref is the reference temperature (25℃)

[0052] α, β are temperature coefficients (calibrated by experiment).

[0053] 22.3 Actual calculation example

[0054] 22.3.1 Known conditions:

[0055] T = 45℃;

[0056] Rraw = 8.7kΩ;

[0057] Calibration parameters: α = -0.0038 / ℃; β = 0.000015 / ℃ 2 .

[0058] 22.3.2 Calculation steps:

[0059] (1) Calculate the temperature difference:

[0060] ΔT = 45-25 = 20℃

[0061] (2) Calculate the compensation factor:

[0062] comp_factor = 1+(-0.0038)×20+0.000015×20 2 = 1-0.076+0.006 = 0.93.

[0063] (3) Get the compensated resistance value:

[0064] Rcomp = 8.7kΩ / 0.93 = 9.35kΩ.

[0065] By temperature compensation of the gas sensor resistance through the step S20, the temperature drift influence can be eliminated, for example, the original resistance of the gas sensor before compensation is 8.7kΩ, which is quite different from the 45℃ reference value 9.6kΩ, seemingly a leakage occurs, but the resistance of the gas sensor after compensation is 9.35kΩ, which is close to the 45℃ reference value 9.6kΩ, and can be determined as no leakage occurs. Therefore, by temperature compensation, the temperature drift interference can be eliminated, the false alarm problem caused by baseline drift of the gas sensor MP510C in high temperature environment can be solved, the anti-interference performance is high, the environmental adaptability is high, so as to improve the detection precision and reduce the false alarm rate.

[0066] S30, based on the collected gas sensor resistance and the gas sensor resistance after temperature compensation, leakage characteristic analysis is carried out, and it is judged whether leakage occurs or not, if yes, step S40 is executed, if not, step S10 is returned.

[0067] 3.1 Feature extraction:

[0068] Based on the collected gas sensor resistance and the gas sensor resistance after temperature compensation, key feature quantities are extracted:

[0069] Characteristic quantity Calculation formula Typical value of leakage Typical value of disturbance Resistance drop rate (ΔR) (R_comp-R_raw) / R_raw >15％ <5％ Change rate (dR / dt) ΔR / Δt >5％ / s <1％ / s Duration (t) Duration of ΔR>5% >3s <1s

[0070] 3.2 Leakage judgment:

[0071] Based on the comparison and analysis of the extracted key feature quantities and the typical values of leakage and interference, it is judged whether leakage occurs or not. For example:

[0072] (1) When the extracted key feature quantity is ΔR = 18%, dR / dt = 7% / s, t = 10s, each feature is consistent with the true leakage feature, therefore, it is determined that leakage occurs.

[0073] (2) When the extracted key feature quantity is ΔR = 8%, dR / dt = 30% / s, t = 0.2s, it can be found that dR / dt is obviously too high, and in the fan interference, the change rate is rapidly pulled up within 0.1s, which is caused by the fan stop leading to the formation of local high temperature air mass around the gas sensor, therefore, it is determined as environmental interference (fan interference) and actual leakage does not occur.

[0074] S40, corresponding measures are taken according to the leakage level, and jump back to step S10.

[0075] When it is determined that leakage occurs, corresponding measures are taken according to the leakage level. The leakage level classification and response measures table is as follows:

[0076]

[0077] After taking corresponding measures according to the leakage level, jump back to step S10, continue to collect the resistance value of the gas sensor and the environmental temperature data, and continuously monitor the server cooling gas leakage state.

[0078] The server cooling gas leakage detection method based on closed-loop control in the embodiment of the application adopts the gas sensor MP510C to cooperate with the temperature sensor TPM112 to detect the server cooling gas leakage, and can realize the following beneficial technical effects:

[0079] 1. Based on the characteristics that the gas sensor MP510C is sensitive to common cooling gas, the application can be applied to detection scenes for different cooling gases, and the application scenarios are wide.

[0080] 2. Based on the characteristics that the gas sensor MP510C has high sensitivity and fast response speed, the application can realize leakage detection with a minimum detectable amount of 0.1 L / min, which is higher than the minimum detectable amount of 0.5 L / min of the traditional method. The micro-leakage detection rate is improved, and the detection accuracy is significantly improved. Moreover, the response time for 0.1 L / min leakage is shortened from 30 s to 8 s, the sensitivity is improved by 375%, and the response speed is significantly improved.

[0081] 3. The gas sensor MP510C is used to replace the pressure sensor for leakage detection. After removing the pressure sensor, the hardware cost is reduced by 40%, which greatly reduces the cost. Moreover, the gas sensor has small size and is suitable for high-density deployment, and is more suitable for application in compact scenes such as edge computing nodes.

[0082] 4. The temperature sensor TPM112 is used to collect environmental temperature data, and the resistance value of the gas sensor is temperature-compensated according to the environmental temperature change. The temperature drift interference can be eliminated through temperature compensation, the false alarm problem caused by baseline drift of the gas sensor MP510C in a high-temperature environment can be solved, the anti-interference performance is high, the environmental adaptability is high, and thus the detection accuracy is improved and the false alarm rate is reduced.

[0083] Again referring to Figure 2 In some preferred embodiments, the step S10 further includes a step S01 of initializing and calibrating the gas sensor before the step S10.

[0084] After the BMC chip starts up, the MP510C gas sensor is first initialized and calibrated. A multiple averaging algorithm is used to calibrate the voltage value converted from the resistance of the MP510C gas sensor. Specifically, after the BMC chip starts up, the resistance of the MP510C gas sensor is sampled multiple times consecutively. The voltage values ​​obtained from these multiple samples are then accumulated and averaged. This average value is used as a benchmark for initializing and calibrating the gas sensor. By initializing and calibrating the gas sensor before initiating subsequent leak detection, the accuracy of subsequent leak detection can be improved.

[0085] like Figure 3 As shown, in another embodiment of the present invention, the server cooling gas leakage detection method based on closed-loop control includes steps S110 to S180:

[0086] S110: Collect gas sensor resistance and ambient temperature data through the first sampling frequency.

[0087] This invention employs a gas sensor for detecting cooling gas leaks. The gas sensor is connected in series with a 10KΩ resistor to form a voltage divider circuit, which divides the 5V DC voltage. The voltage signal after voltage division is output to the ADC chip. The voltage signal after voltage division by the ADC chip is converted from analog to digital and then transmitted to the BMC chip. The BMC chip converts the digital voltage signal output by the ADC chip into the gas sensor resistance value. In this embodiment, the gas sensor is an MP510C gas sensor, a semiconductor gas sensor with the following characteristics: 1. Sensitive to common cooling gases (such as R134a, R404a, etc.); 2. Fast response speed: response time <10 seconds, more than 3 times faster than pressure monitoring methods; 3. High sensitivity: resistance value decreases significantly with increasing concentration of cooling gas (such as R134a, nitrogen, etc.); 4. Small size: suitable for high-density deployment. Of course, in other embodiments of the present invention, other gas sensors such as MQ2 or TGS2600 can be used to replace the gas sensor of MP510C for gas leak detection.

[0088] In this embodiment of the invention, the temperature sensor is a TPM112 temperature sensor; of course, in other embodiments of the invention, other temperature sensors such as DS18B20 or TM75BD can be used instead of the TPM112 temperature sensor for detecting changes in ambient temperature.

[0089] After the system starts testing, the BMC chip collects the gas sensor resistance value and ambient temperature data at a first sampling frequency (e.g., 1Hz). By collecting the gas sensor resistance value and ambient temperature data at a lower first sampling frequency, the server cooling gas leakage status can be continuously monitored, which can reduce the computational pressure on the BMC chip and thus improve system performance.

[0090] S120. Different compensation strategies are used to compensate the resistance of the gas sensor according to changes in ambient temperature.

[0091] Step S120 specifically includes steps S121 to S122:

[0092] S121. Determine whether the change in ambient temperature within adjacent sampling periods exceeds 2 degrees. If yes, proceed to step S122. If no, the temperature compensation is zero.

[0093] S122. A second-order polynomial compensation model is adopted: Rcomp = Rraw × 1 / [1 + α(T - Tref) + β(T - Tref)]. 2 Temperature compensation is applied to the resistance of the gas sensor.

[0094] The temperature compensation calculation process is as follows: 122.1 Establishing a temperature-resistance relationship model

[0095] The temperature characteristics of MP510C under leak-free conditions, as measured experimentally, are shown in the table below:

[0096]

[0097]

[0098] 122.2 Establishing the Compensation Formula

[0099] This invention employs a second-order polynomial compensation model to perform temperature compensation on the resistance of the gas sensor:

[0100] Rcomp=Rraw×1 / [1+α(T-Tref)+β(T-Tref) 2 ]

[0101] in:

[0102] Rraw is the original resistance value of the gas sensor;

[0103] Rcomp is the compensated resistance value of the gas sensor;

[0104] T represents the current ambient temperature;

[0105] Tref is the reference temperature (taken as 25℃).

[0106] α and β are temperature coefficients (calibrated experimentally).

[0107] 122.3 Practical Calculation Example

[0108] 122.3.1 Given conditions:

[0109] T = 45℃;

[0110] Rraw = 8.7kΩ;

[0111] Calibration parameters: α = -0.0038 / ℃; β = 0.000015 / ℃ 2 .

[0112] 122.3.2 Calculation steps:

[0113] (1) Calculate the temperature difference:

[0114] ΔT = 45 - 25 = 20℃

[0115] (2) Calculate the compensation factor:

[0116] comp_factor=1+(-0.0038)×20+0.000015×20 2 =1-0.076+0.006=0.93.

[0117] (3) Obtain the compensated resistance value:

[0118] Rcomp=8.7kΩ / 0.93=9.35kΩ.

[0119] By performing temperature compensation on the gas sensor resistance in step S120, the influence of temperature drift can be eliminated. For example, if the original resistance of the gas sensor before compensation is 8.7kΩ, which differs significantly from the 45℃ reference value of 9.6kΩ, it may appear as if a leak has occurred. However, the gas sensor resistance after compensation is 9.35kΩ, which is close to the 45℃ reference value of 9.6kΩ, indicating that no leak has occurred. Therefore, temperature compensation can eliminate temperature drift interference, solving the false alarm problem caused by baseline drift of the MP510C gas sensor in high-temperature environments. It exhibits strong anti-interference capabilities and high environmental adaptability, thereby improving detection accuracy and reducing the false alarm rate.

[0120] S130. Determine whether the resistance value of the gas sensor is abnormal. If yes, proceed to step S140; otherwise, return to step S110.

[0121] Specifically, it is determined whether the difference between the resistance value of the gas sensor and the reference resistance value exceeds a preset range (the preset range is set according to the actual situation, for example [0, 0.5]). If yes, the resistance value of the gas sensor is determined to be abnormal, and step S140 is executed; if no, the resistance value of the gas sensor is determined to be normal, and step S110 is returned to continue to collect the gas sensor resistance value and ambient temperature data at the first sampling frequency to continuously monitor the server cooling gas leakage status.

[0122] Before performing leak characteristic analysis, it is first determined whether the gas sensor resistance is abnormal. When the gas sensor resistance is normal, it can be determined that no leak has occurred, and the process can directly return to step S110 to continue collecting gas sensor resistance and ambient temperature data to continuously monitor the server cooling gas leak status. This eliminates the need for subsequent leak characteristic analysis steps, reduces the computational load of the BMC chip, and thus improves system performance.

[0123] S140. The gas sensor resistance and ambient temperature data are sampled multiple times continuously using the second sampling frequency.

[0124] After determining that the gas sensor resistance is abnormal, the BMC chip adjusts the sampling frequency and samples the gas sensor resistance and ambient temperature data multiple times (e.g., 3 times) at a second sampling frequency (e.g., 100Hz) for further determination of whether a leak has occurred.

[0125] After determining that the gas sensor resistance is abnormal, the BMC chip collects gas sensor resistance and ambient temperature data at a higher second sampling frequency, which can improve the response speed of leak detection.

[0126] S150. Determine whether the resistance value of the gas sensor obtained from multiple consecutive samplings changes continuously. If yes, proceed to step S160; otherwise, return to step S110.

[0127] This step determines whether the gas sensor resistance value obtained from multiple consecutive samplings is continuously changing. If the gas sensor resistance value is continuously changing, it indicates that the abnormal gas sensor resistance value may be caused by a cooling gas leak, and the subsequent process can continue, i.e., step S160 is executed. If the gas sensor resistance value is not continuously changing, it indicates that the abnormal resistance value is not caused by a cooling gas leak, and the gas sensor may be malfunctioning. It is necessary to return to step S110 and re-collect the gas sensor resistance value and ambient temperature data for verification.

[0128] This step, by determining whether the resistance value of the gas sensor obtained from multiple consecutive samples changes continuously, can eliminate false alarms caused by abnormal conditions such as gas sensor malfunction.

[0129] S160. Different compensation strategies are used to perform temperature compensation on the gas sensor resistance value obtained from multiple consecutive samplings based on changes in ambient temperature.

[0130] This step uses the same method (strategy) as step S120 to perform temperature compensation on the gas sensor resistance values ​​obtained from multiple consecutive samplings, which will not be described in detail here.

[0131] By performing temperature compensation on the gas sensor resistance value obtained from multiple consecutive samplings in step S160, temperature drift interference can be eliminated, and the false alarm problem caused by baseline drift of the MP510C gas sensor in high-temperature environments can be solved. It has strong anti-interference ability and high environmental adaptability, thereby improving detection accuracy and reducing the false alarm rate.

[0132] S170. Based on the collected gas sensor resistance value and the gas sensor resistance value obtained after temperature compensation, perform leakage characteristic analysis and determine whether a leak has occurred. If yes, proceed to step S180; otherwise, return to step S110.

[0133] 17.1 Feature Extraction:

[0134] This step extracts the key feature quantities: the gas sensor resistance values ​​obtained from multiple consecutive samplings, and the gas sensor resistance values ​​obtained after temperature compensation of the multiple consecutive samplings. The extracted key feature quantities are shown in the table below:

[0135] Characteristic quantity Calculation formula Typical value of leakage Typical value of disturbance Resistance drop rate (ΔR) (R_comp-R_raw) / R_raw >15％ <5％ Change rate (dR / dt) ΔR / Δt >5％ / s <1％ / s Duration (t) Duration of ΔR>5% >3s <1s

[0136] 17.2 Leakage Assessment:

[0137] By comparing and analyzing the extracted key features with typical leakage and interference values, it is possible to determine whether a leak has occurred. For example:

[0138] (1) When the extracted key feature quantities are ΔR = 18%, dR / dt = 7% / s, and t = 10s, all features are consistent with the actual leakage features. Therefore, it is determined that a leakage has occurred, and step S180 is executed.

[0139] (2) When the extracted key feature quantities are ΔR = 8%, dR / dt = 30% / s, and t = 0.2s, it can be found that dR / dt is significantly higher. In the case of fan interference, the rate of change increases rapidly within 0.1s. This is caused by the fan stopping, which leads to the formation of a local high-temperature air mass around the gas sensor. Therefore, it is determined to be environmental interference (fan interference). In fact, no leakage has occurred. Return to step S110 and continue to collect the gas sensor resistance value and ambient temperature data at a lower first sampling frequency to reduce the computational pressure when continuously monitoring the server cooling gas leakage status.

[0140] S180. Take appropriate measures according to the leakage level, and jump back to step S140.

[0141] When a leak is detected, appropriate measures should be taken according to the leak level. The leak level classification and response measures are shown in the table below:

[0142]

[0143] After taking appropriate measures according to the leakage level, the process jumps back to step S140 to continue collecting gas sensor resistance and ambient temperature data at a higher second sampling frequency to continuously monitor the server cooling gas leakage status. Here, in the case of an existing leak, continuously monitoring the server cooling gas leakage status at a higher second sampling frequency can improve the detection response speed.

[0144] The server cooling gas leak detection method based on closed-loop control in this embodiment of the invention uses an MP510C gas sensor in conjunction with a TPM112 temperature sensor to detect server cooling gas leaks, and can achieve the following beneficial technical effects:

[0145] 1. Based on the MP510C gas sensor's sensitivity to common cooling gases, this invention can be applied to detection scenarios for different cooling gases, making it applicable to a wide range of scenarios.

[0146] 2. Based on the high sensitivity and fast response speed of the MP510C gas sensor, this invention can detect leaks with a minimum detectable amount of 0.1L / min, which is significantly improved compared to the traditional method with a minimum detectable amount of 0.5L / min. This improves the detection rate of minute leaks and significantly enhances the detection accuracy. Moreover, the response time for leaks at the 0.1L / min level is shortened from 30s to 8s, and the sensitivity is increased by 375%, significantly improving the response speed.

[0147] 3. The MP510C gas sensor is used to replace the pressure sensor for leak detection. Removing the pressure sensor reduces hardware costs by 40%, which greatly reduces costs. Moreover, the gas sensor is small in size, suitable for high-density deployment, and is more suitable for compact scenarios such as edge computing nodes.

[0148] 4. The TPM112 temperature sensor is used to collect ambient temperature data. The resistance of the gas sensor is compensated for temperature changes based on the ambient temperature. Temperature compensation can eliminate temperature drift interference and solve the false alarm problem caused by baseline drift of the MP510C gas sensor in high-temperature environments. It has strong anti-interference ability and high environmental adaptability, thereby improving detection accuracy and reducing false alarm rate.

[0149] 5. By using different sampling frequencies to collect gas sensor resistance and ambient temperature data under different conditions, the system's computational load and detection response speed can be balanced.

[0150] Refer again Figure 2 In some preferred embodiments, step S101 is included before step S110: initializing and calibrating the gas sensor.

[0151] After the BMC chip starts up, the MP510C gas sensor is first initialized and calibrated. A multiple averaging algorithm is used to calibrate the voltage value converted from the resistance of the MP510C gas sensor. Specifically, after the BMC chip starts up, the resistance of the MP510C gas sensor is sampled multiple times consecutively. The voltage values ​​obtained from these multiple samples are then accumulated and averaged. This average value is used as a benchmark for initializing and calibrating the gas sensor. By initializing and calibrating the gas sensor before initiating subsequent leak detection, the accuracy of subsequent leak detection can be improved.

[0152] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention in any way. Those skilled in the art can make various equivalent changes and improvements based on the above embodiments, and all equivalent variations or modifications made within the scope of the claims should fall within the protection scope of the present invention.

Claims

1. A method for detecting server cooling gas leakage based on closed-loop control, characterized in that, Includes the following steps: S10. Real-time acquisition of gas sensor resistance and ambient temperature data; S20. Different compensation strategies are used to compensate the resistance of the gas sensor according to changes in ambient temperature. S30. Based on the collected gas sensor resistance value and the gas sensor resistance value obtained after temperature compensation, perform leakage characteristic analysis and determine whether a leak has occurred. If yes, proceed to step S40; otherwise, return to step S10. S40. Take appropriate measures according to the leakage level, and jump back to step S10.

2. The server cooling gas leakage detection method based on closed-loop control as described in claim 1, characterized in that, Before step S10, the method further includes: initializing and calibrating the gas sensor.

3. A method for detecting server cooling gas leakage based on closed-loop control, characterized in that, Includes the following steps: S110. Collect gas sensor resistance and ambient temperature data using the first sampling frequency; S120. Different compensation strategies are used to compensate the resistance of the gas sensor according to changes in ambient temperature. S130. Determine whether the resistance value of the gas sensor is abnormal. If yes, proceed to step S140; otherwise, return to step S110. S140. The gas sensor resistance and ambient temperature data are continuously sampled multiple times using the second sampling frequency. S150. Determine whether the resistance value of the gas sensor obtained from multiple consecutive samplings changes continuously. If yes, proceed to step S160; otherwise, return to step S110. S160. Different compensation strategies are used to compensate for the temperature of the gas sensor resistance value obtained from multiple consecutive samplings based on changes in ambient temperature. S170. Based on the collected gas sensor resistance value and the gas sensor resistance value obtained after temperature compensation, perform leakage characteristic analysis and determine whether a leak has occurred. If yes, proceed to step S180; otherwise, return to step S110. S180. Take appropriate measures according to the leakage level, and jump back to step S140.

4. The server cooling gas leakage detection method based on closed-loop control as described in claim 3, characterized in that, Step S120 includes: S121. Determine whether the change in ambient temperature within adjacent sampling periods exceeds 2 degrees. If yes, proceed to step S122. If no, the temperature compensation is zero. S122. A second-order polynomial compensation model is adopted: Rcomp = Rraw × 1 / [1 + α(T - Tref) + β(T - Tref)]. 2 Temperature compensation is applied to the resistance of the gas sensor, where Rraw is the original resistance of the gas sensor, Rcomp is the compensated resistance of the gas sensor, T is the current ambient temperature, Tref is the reference temperature, and α and β are temperature coefficients.

5. The server cooling gas leakage detection method based on closed-loop control as described in claim 3, characterized in that, In step S130, determining whether the resistance of the gas sensor is abnormal includes: determining whether the difference between the resistance of the gas sensor and the reference resistance exceeds a preset range; if so, the resistance of the gas sensor is determined to be abnormal; if not, the resistance of the gas sensor is determined to be normal.

6. The server cooling gas leakage detection method based on closed-loop control as described in any one of claims 3-5, characterized in that, Before step S110, the method further includes: initializing and calibrating the gas sensor.

7. A server cooling gas leak detection system based on closed-loop control, characterized in that, include: A gas sensor is used to detect whether there is a leak in the server's cooling gas; A temperature sensor is used to detect changes in the ambient temperature around the gas sensor. A BMC chip, wherein the BMC chip is connected to the gas sensor and the temperature sensor respectively; The BMC chip is used to implement each step of the server cooling gas leakage detection method based on closed-loop control as described in any one of claims 1-6.

8. The server cooling gas leak detection system based on closed-loop control as described in claim 7, characterized in that, The gas sensor is connected in series with a resistor, one end of which is connected to a 5V DC power supply and the other end is grounded. The connection point between the gas sensor and the resistor is connected to an ADC chip. The ADC chip is used to convert the voltage signal obtained by the voltage division between the gas sensor and the resistor into an analog-to-digital signal and then transmit it to the BMC chip.

9. The server cooling gas leak detection system based on closed-loop control as described in claim 8, characterized in that, The gas sensor is of model MP510C, MQ2, or TGS2600.

10. The server cooling gas leak detection system based on closed-loop control as described in claim 8, characterized in that, The temperature sensor is model DS18B20, TM75BD, or TPM112.

Images