Modular data center air-conditioning room air tightness monitoring system

Abstract

The invention relates to the technical field of airtightness monitoring, in particular to a modular data center air-conditioning room airtightness monitoring system which comprises a sensor network, data acquisition and processing, intelligent analysis and diagnosis and robot maintenance and repair. Wherein the sensor network comprises a pressure difference sensor used for monitoring the static pressure difference of each room, and real-time adjustment is performed by setting a dynamic threshold value so as to adapt to changes of different rooms and external environments; the temperature sensor is used for accurately measuring the temperature of the air conditioner room, multi-point temperature measurement is fused, and the influence of local abnormity on overall judgment is reduced; and the electric power sensor is used for monitoring instantaneous electric power, is linked with the pressure difference and temperature sensor and is used for analyzing specific reasons of overload of the air conditioner. According to the system, an innovative method is provided to enhance the air tightness management of the air-conditioning room of the data center, and the efficiency, reliability and cost effectiveness of the system are improved through intelligent monitoring, automatic fault processing and an adaptive optimization algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field of airtightness monitoring, and in particular to an airtightness monitoring system for an air-conditioning room in a modular data center. Background Art

[0002] As the core infrastructure of information technology, the energy consumption of the air conditioning system of the data center accounts for a significant proportion of the energy consumption of the entire data center, usually between 30% and 45%. Therefore, optimizing the energy consumption management of the air-conditioned room in the data center is of great significance for achieving energy conservation and consumption reduction. If the coldness of the air-conditioned room penetrates into the non-air-conditioned area through the gaps in the room, the energy consumption of the air-conditioning system will increase significantly. The modular data center consists of multiple modules. Although sealing measures are taken during the construction process, gaps may still occur due to factors such as construction technology. Therefore, it is very necessary to monitor the air tightness of the air-conditioned room before and during the use of the data center, which helps to find and repair gaps in time, improve the airtightness of the enclosure structure, and thus reduce air conditioning energy consumption. To this end, an air tightness monitoring system for the air-conditioned room of a modular data center is proposed. Summary of the invention

[0003] The object of the present invention is to provide a modular data center air conditioning room air tightness monitoring system to solve the problems raised in the above background technology.

[0004] To achieve the above-mentioned object, the present invention provides the following technical solutions: a modular data center air conditioning room air tightness monitoring system, including a sensor network, data acquisition and processing, intelligent analysis and diagnosis, and robot maintenance and repair;

[0005] The sensor network includes a pressure difference sensor for monitoring the static pressure difference in each room, which can be adjusted in real time to adapt to changes in different rooms and external environment by setting dynamic thresholds;

[0006] Temperature sensors used to accurately measure the temperature of air-conditioning rooms, integrating multi-point temperature measurements to reduce the impact of local anomalies on overall judgment;

[0007] Power sensors used to monitor instantaneous electrical power, working in conjunction with pressure difference and temperature sensors to analyze the specific causes of air conditioning overload;

[0008] The data collection and processing includes a data collector, which uses high-speed data collection, supports data packet redundancy, and uses timestamps to accurately align data from different sensors;

[0009] Intelligent analysis and diagnosis include intelligent algorithms, pattern recognition based on machine learning, comprehensive analysis of pressure difference, temperature, and electric power data, dynamic adjustment of model parameters to improve fault detection accuracy, threshold management, automatic adjustment of thresholds, adapting to different seasons and load changes. When long-term change trends are detected, the system will automatically update the baseline parameters;

[0010] Robot maintenance and repair include fault detection and scheduling, using the shortest path algorithm to reduce response time, quickly locating the source of airtightness problems, repair strategies, backing up repair records in advance, combining AI algorithms to predict possible problems, and optimizing robot inspection routes based on fault frequency to improve maintenance efficiency.

[0011] Preferably, the specific implementation steps of the modular data center air conditioning room airtightness monitoring system are as follows:

[0012] Step 1: Sensor installation and calibration;

[0013] Step 2: Data collection and transmission;

[0014] Step 3: Data processing and analysis;

[0015] Step 4: Exception handling and automatic scheduling;

[0016] Step 5: Optimize and learn.

[0017] Preferably, the step one: sensor installation and calibration, installs pressure difference, temperature, and power sensors at key locations in the air-conditioned room, and calibrates the sensors regularly, and then sets initial thresholds and baselines in combination with historical data.

[0018] Preferably, the step 2: data collection and transmission: the data collector collects all sensor data at a frequency of 10 times per minute, and the data packets are transmitted to the computer host via a wireless network, using an encrypted transmission protocol to ensure data security.

[0019] Preferably, the step three: data processing and analysis: the computer host uses the following formula to perform data analysis:

[0020] Pressure difference calculation: Among them, Δp i is the pressure difference value at each moment in the current time window;

[0021] Temperature mean calculation: Among them, T j Take temperature readings at different locations;

[0022] Power consumption calculation: P0 = V·I·cos(φ), where V is voltage, I is current, and cos(φ) is power factor;

[0023] Anomaly Detection:

[0024] If E>E threshold , it is judged as abnormal, and E is the abnormal score of the data point.

[0026] Preferably, the step four: exception handling and automatic scheduling: once the computer host detects an airtightness abnormality, the robot is immediately dispatched to conduct an on-site inspection. The robot scans the room through LIDAR or a camera to confirm the location of the gap and repair it.

[0027] Preferably, the step five: optimization and learning specifically includes data storage and management, adaptive optimization algorithm, model update and iteration, feedback mechanism and optimization strategy, and fault prediction and prevention.

[0028] Preferably, for the data storage and management, after each maintenance, the system should collect and store all relevant data, including the specific location, type, severity, repair method, time and resources used for the fault, and use a unified data format to store information, specifically using a structured database or an unstructured database for storage, and classify and archive historical data by time, type or other characteristics to facilitate long-term trend analysis.

[0029] Preferably, the adaptive optimization algorithm adopts a genetic algorithm, a particle swarm optimization algorithm or a Bayesian optimization algorithm to adapt to different types of fault data analysis, uses a machine learning method to extract features from fault data, identifies key factors such as the impact of temperature, pressure, power consumption, etc. on the fault, uses fault and repair data to train the model, continuously optimizes the prediction ability of the algorithm, adopts supervised learning or reinforcement learning methods, updates the model weights according to historical fault data, so that the model can accurately predict the probability of fault occurrence;

[0030] The model is updated and iterated, and model parameters such as learning rate, regularization parameters, and weights are dynamically adjusted according to the characteristics of new fault data. The model performance is evaluated regularly, and the prediction ability of the model is measured using cross-validation and indicators such as accuracy, precision, and recall. The model's online learning function is implemented, so that the model can be updated and adapted to new data in real time, and the prediction and repair strategies are continuously optimized.

[0031] Preferably, the feedback mechanism and optimization strategy collects feedback information in the repair process in real time, records the effectiveness and efficiency of the repair steps, uses the feedback information to continuously optimize the model and repair strategy, analyzes the repair results, finds the optimal repair solution, and applies Q-learning or policy gradient methods in reinforcement learning to gradually improve the success rate of repair;

[0032] The fault prediction and prevention described herein performs fault prediction based on the optimized model, warns of possible problems in advance, provides preventive maintenance suggestions, reduces the occurrence of potential faults, and performs system inspection and optimization on a regular basis.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] 1. The system provides an innovative approach to enhance the air tightness management of data center air conditioning rooms, improving the efficiency, reliability and cost-effectiveness of the system through intelligent monitoring, automatic fault handling and adaptive optimization algorithms;

[0035] 2. Through adaptive optimization algorithms, feedback mechanisms, fault prediction and other measures, the efficiency of fault detection and repair in data centers can be significantly improved. This not only improves the stability and reliability of the system, but also reduces operation and maintenance costs and achieves more efficient management and maintenance. The solution combines advanced data analysis technology and intelligent optimization algorithms to provide strong support for the intelligent management of data centers. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] Figure 1 This is a processing flow chart of the modular data center air conditioning room airtightness monitoring system of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] See also Figure 1 , the present invention provides a technical solution:

[0040] A modular data center air conditioning room air tightness monitoring system, including sensor network, data acquisition and processing, intelligent analysis and diagnosis, and robot maintenance and repair;

[0041] The sensor network includes a pressure difference sensor for monitoring the static pressure difference in each room, which can be adjusted in real time to adapt to changes in different rooms and external environment by setting dynamic thresholds;

[0042] Temperature sensors used to accurately measure the temperature of air-conditioning rooms, integrating multi-point temperature measurements to reduce the impact of local anomalies on overall judgment;

[0043] Power sensors used to monitor instantaneous electrical power, linked with pressure difference and temperature sensors, to analyze the specific causes of air conditioning overload;

[0044] The data collection and processing includes a data collector, which uses high-speed data collection, supports data packet redundancy, and uses timestamps to accurately align data from different sensors;

[0045] Intelligent analysis and diagnosis include intelligent algorithms, pattern recognition based on machine learning, comprehensive analysis of pressure difference, temperature, and electric power data, dynamic adjustment of model parameters to improve fault detection accuracy, threshold management, automatic adjustment of thresholds, adapting to different seasons and load changes. When long-term change trends are detected, the system will automatically update the baseline parameters;

[0046] Robot maintenance and repair include fault detection and scheduling, using the shortest path algorithm to reduce response time, quickly locating the source of airtightness problems, repair strategies, backing up repair records in advance, combining AI algorithms to predict possible problems, and optimizing robot inspection routes based on fault frequency to improve maintenance efficiency.

[0047] Specifically, the specific implementation steps of the modular data center air conditioning room airtightness monitoring system are as follows:

[0048] Step 1: Sensor installation and calibration;

[0049] Step 2: Data collection and transmission;

[0050] Step 3: Data processing and analysis;

[0051] Step 4: Exception handling and automatic scheduling;

[0052] Step 5: Optimize and learn.

[0053] Specifically, the step one: sensor installation and calibration, installs pressure difference, temperature, and power sensors at key locations in the air-conditioned room, calibrates the sensors regularly, and then sets initial thresholds and baselines based on historical data.

[0054] Specifically, the step 2: data collection and transmission: the data collector collects all sensor data at a frequency of 10 times per minute, and the data packets are transmitted to the computer host via a wireless network, using an encrypted transmission protocol to ensure data security.

[0055] Specifically, the step three: data processing and analysis: the computer host uses the following formula to perform data analysis:

[0056] Pressure difference calculation: Among them, Δp i is the pressure difference value at each moment in the current time window;

[0057] Temperature mean calculation: Among them, T j Take temperature readings at different locations;

[0058] Power consumption calculation: P0 = V·I·cos(φ), where V is voltage, I is current, and cos(φ) is power factor;

[0059] Anomaly Detection:

[0060] If E>E threshold , it is judged as abnormal, and E is the abnormal score of the data point.

[0062] Specifically, the step four: exception handling and automatic scheduling: once the computer host detects an airtightness abnormality, the robot is immediately dispatched to conduct an on-site inspection. The robot scans the room through LIDAR or a camera to confirm the location of the gap and repair it.

[0063] Specifically, the step five: optimization and learning specifically includes data storage and management, adaptive optimization algorithms, model updating and iteration, feedback mechanisms and optimization strategies, and fault prediction and prevention.

[0064] Specifically, as for data storage and management, after each maintenance, the system should collect and store all relevant data, including the specific location, type, severity, repair method, time and resources used of the fault, and use a unified data format to store information, specifically using a structured database or an unstructured database for storage, and classify and archive historical data by time, type or other characteristics to facilitate long-term trend analysis.

[0065] Specifically, the adaptive optimization algorithm adopts a genetic algorithm, a particle swarm optimization algorithm or a Bayesian optimization algorithm to adapt to different types of fault data analysis, uses a machine learning method to extract features from fault data, identifies key factors such as the impact of temperature, pressure, power consumption, etc. on the fault, uses fault and repair data to train the model, continuously optimizes the prediction ability of the algorithm, adopts supervised learning or reinforcement learning methods, updates the model weights according to historical fault data, and enables the model to accurately predict the probability of fault occurrence;

[0066] The model is updated and iterated, and model parameters such as learning rate, regularization parameters, and weights are dynamically adjusted according to the characteristics of new fault data. The model performance is evaluated regularly, and the prediction ability of the model is measured using cross-validation and indicators such as accuracy, precision, and recall. The model's online learning function is implemented, so that the model can be updated and adapted to new data in real time, and the prediction and repair strategies are continuously optimized.

[0067] Specifically, the feedback mechanism and optimization strategy collects feedback information in the repair process in real time, records the effectiveness and efficiency of the repair steps, uses the feedback information to continuously optimize the model and repair strategy, analyzes the repair results, finds the optimal repair solution, and applies Q-learning or policy gradient methods in reinforcement learning to gradually improve the success rate of repair;

[0068] The fault prediction and prevention described herein performs fault prediction based on the optimized model, warns of possible problems in advance, provides preventive maintenance suggestions, reduces the occurrence of potential faults, and performs system inspection and optimization on a regular basis.

[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A modular data center air conditioning room air tightness monitoring system, characterized in that: Including sensor networks, data collection and processing, intelligent analysis and diagnosis, and robot maintenance and repair; The sensor network includes a pressure difference sensor for monitoring the static pressure difference in each room, which can be adjusted in real time to adapt to changes in different rooms and external environment by setting dynamic thresholds; Temperature sensors used to accurately measure the temperature of air-conditioning rooms, integrating multi-point temperature measurements to reduce the impact of local anomalies on overall judgment; Power sensors used to monitor instantaneous electrical power, linked with pressure difference and temperature sensors, to analyze the specific causes of air conditioning overload; The data collection and processing includes a data collector, which uses high-speed data collection, supports data packet redundancy, and uses timestamps to accurately align data from different sensors; Intelligent analysis and diagnosis include intelligent algorithms, pattern recognition based on machine learning, comprehensive analysis of pressure difference, temperature, and electric power data, dynamic adjustment of model parameters to improve fault detection accuracy, threshold management, automatic adjustment of thresholds, adapting to different seasons and load changes. When long-term change trends are detected, the system will automatically update the baseline parameters; Robot maintenance and repair include fault detection and scheduling, using the shortest path algorithm to reduce response time, quickly locating the source of airtightness problems, repair strategies, backing up repair records in advance, combining AI algorithms to predict possible problems, and optimizing robot inspection routes based on fault frequency to improve maintenance efficiency.

2. A modular data center air conditioning room air tightness monitoring system according to claim 1, characterized in that: The specific implementation steps of the modular data center air conditioning room airtightness monitoring system are as follows: Step 1: Sensor installation and calibration; Step 2: Data collection and transmission; Step 3: Data processing and analysis; Step 4: Exception handling and automatic scheduling; Step 5: Optimize and learn.

3. A modular data center air conditioning room air tightness monitoring system according to claim 2, characterized in that: The step 1: sensor installation and calibration, installing differential pressure, temperature, and power sensors at key locations in the air-conditioned room, calibrating the sensors regularly, and then setting initial thresholds and baselines based on historical data.

4. A modular data center air conditioning room air tightness monitoring system according to claim 2, characterized in that: Step 2: Data collection and transmission: The data collector collects all sensor data at a frequency of 10 times per minute, and the data packets are transmitted to the computer host via a wireless network, using an encrypted transmission protocol to ensure data security.

5. The modular data center air conditioning room air tightness monitoring system according to claim 2, characterized in that: Step 3: Data processing and analysis: The computer host uses the following formula to perform data analysis: Pressure difference calculation: Among them, Δp i is the pressure difference value at each moment in the current time window; Temperature mean calculation: Among them, T j Take temperature readings at different locations; Power consumption calculation: P0 = V·I·cos(φ), where V is voltage, I is current, and cos(φ) is power factor; Anomaly Detection: If E>E threshold , it is judged as abnormal, and E is the abnormal score of the data point.

6. A modular data center air conditioning room air tightness monitoring system according to claim 2, characterized in that: Step 4: Exception handling and automatic scheduling: Once the computer host detects an airtightness abnormality, the robot is immediately dispatched to conduct an on-site inspection. The robot scans the room through LIDAR or a camera to confirm the location of the gap and repair it.

7. A modular data center air conditioning room air tightness monitoring system according to claim 2, characterized in that: The step five: optimization and learning specifically includes data storage and management, adaptive optimization algorithms, model updating and iteration, feedback mechanisms and optimization strategies, and fault prediction and prevention.

8. A modular data center air conditioning room air tightness monitoring system according to claim 7, characterized in that: As for data storage and management, after each maintenance, the system should collect and store all relevant data, including the specific location, type, severity, repair method, time and resources used for the fault, and use a unified data format to store information. Specifically, a structured database or an unstructured database is used for storage, and historical data is classified and archived by time, type or other characteristics to facilitate long-term trend analysis.

9. The modular data center air conditioning room air tightness monitoring system according to claim 7, characterized in that: The adaptive optimization algorithm adopts a genetic algorithm, a particle swarm optimization algorithm or a Bayesian optimization algorithm to adapt to different types of fault data analysis, uses a machine learning method to extract features from fault data, identifies key factors such as the impact of temperature, pressure, power consumption, etc. on faults, uses fault and repair data to train models, continuously optimizes the prediction ability of the algorithm, and uses supervised learning or reinforcement learning methods to update model weights based on historical fault data so that the model can accurately predict the probability of faults. The model is updated and iterated, and model parameters such as learning rate, regularization parameters, and weights are dynamically adjusted according to the characteristics of new fault data. The model performance is evaluated regularly, and the prediction ability of the model is measured using cross-validation and indicators such as accuracy, precision, and recall. The model's online learning function is implemented, so that the model can be updated and adapted to new data in real time, and the prediction and repair strategies are continuously optimized.

10. The modular data center air conditioning room air tightness monitoring system according to claim 7, characterized in that: The feedback mechanism and optimization strategy collects feedback information in the repair process in real time, records the effectiveness and efficiency of the repair steps, uses the feedback information to continuously optimize the model and repair strategy, analyzes the repair results, finds the best repair solution, and applies Q-learning or policy gradient methods in reinforcement learning to gradually improve the success rate of repair; The fault prediction and prevention described herein performs fault prediction based on the optimized model, warns of possible problems in advance, provides preventive maintenance suggestions, reduces the occurrence of potential faults, and performs system inspection and optimization on a regular basis.

Images