A data center gas fire extinguishing protection area linkage unlocking and environment recovery control method

Abstract

This invention relates to the field of fire protection linkage and environmental control technology, and discloses a method for linkage unlocking and environmental restoration control of a data center gas extinguishing protection zone. The method acquires multi-dimensional physical environment monitoring data of the protection zone, performs spatiotemporal alignment after a preset immersion time is met, maps the data to a three-dimensional spatial coordinate system to construct a three-dimensional concentration field matrix, extracts the concentration difference between spatial nodes as data gradient features, and fuses them to generate a comprehensive environmental safety feature vector. Based on this, a real-time environmental index is calculated; if the safety baseline is not reached, the access control remains locked, and the exhaust fan speed is dynamically adjusted based on the difference. After the baseline is reached, the first derivative of the target data over time within a preset time window is calculated. When the absolute value of the first derivative is lower than a preset fluctuation threshold, the concentration extreme value is extracted, bound to the current timestamp, and an access control unlock signal is output. This invention eliminates monitoring blind spots, prevents secondary escaping of exhaust gas, and filters false safety signals, ensuring personnel access safety.

Description

Technical Field

[0001] This invention relates to the field of fire protection linkage and environmental control technology, and in particular to a method for linkage unlocking and environmental restoration control of a data center gas fire suppression protection zone. Background Technology

[0002] In data center fire protection management, after the gas extinguishing system has been discharged and undergone the impregnation period, the protected area needs to be vented and the environment restored. Traditional linkage control methods usually rely on single-plane sensor monitoring and exhaust fans running at constant speed at industrial frequency, which has significant engineering drawbacks.

[0003] First, due to the raised floors, densely packed server racks, and enclosed hot and cold aisle structures within data centers, airflow organization is extremely complex. This easily creates localized high-concentration dead zones at the bottom of server racks or under anti-static floors, making traditional single-point or single-plane monitoring methods unable to reflect the true physical environment and resulting in serious monitoring blind spots. Second, traditional exhaust control often uses constant speed operation, which can easily lead to the re-entrainment of heavily concentrated exhaust gases settled at the bottom during the later stages of exhaust, causing secondary fluctuations in environmental concentration. Finally, during environmental recovery, strong airflow disturbances within the data center can easily cause sensors to collect brief low-concentration values, generating "false safety" signals. If these signals prematurely trigger access control unlocking, it poses a significant safety hazard to maintenance personnel. Summary of the Invention

[0004] This invention provides a method for linkage unlocking and environmental restoration control of data center gas fire suppression protection zones. By constructing a three-dimensional concentration field to eliminate monitoring blind spots, using nonlinear frequency conversion ventilation to prevent secondary re-entrainment of exhaust gas, and combining environmental convergence verification based on the first derivative to filter false security signals, it achieves efficient and stable data center environmental restoration and secure linkage unlocking of access control.

[0005] This invention provides a method for linkage unlocking and environmental restoration control of a data center gas fire suppression protection zone, wherein the protection zone is equipped with an exhaust fan and an access control controller, and the method includes the following steps:

[0006] S1. Acquire multi-dimensional physical environment monitoring data and gas release status signals from the fire control panel within the data center protection zone. When the gas release status signal indicates that the gas release duration has reached the preset immersion time, perform spatiotemporal alignment on the multi-dimensional physical environment monitoring data to obtain target environment monitoring data.

[0007] S2. Map the target environmental monitoring data to a three-dimensional spatial coordinate system containing multiple spatial grid nodes to construct a three-dimensional concentration field matrix, and extract the concentration difference between adjacent spatial grid nodes in the three-dimensional concentration field matrix as data gradient features, and fuse it with the target environmental monitoring data to generate a comprehensive environmental safety feature vector.

[0008] S3. Calculate the real-time environmental index based on the comprehensive environmental safety feature vector. When the real-time environmental index does not reach the preset safety baseline, output the access control lock maintenance signal to the access control controller, and dynamically adjust the speed of the exhaust fan based on the difference between the real-time environmental index and the preset safety baseline.

[0009] S4. When the real-time environmental index reaches the preset safety baseline, the first derivative of the target environmental monitoring data over time within the preset time window is calculated, starting from the moment the preset safety baseline is reached.

[0010] S5. When the absolute value of the first derivative is lower than the preset fluctuation threshold within the preset time window, the concentration extreme value is extracted from the three-dimensional concentration field matrix at this time, the concentration extreme value is bound and stored with the current timestamp, and the access control unlock signal is output to the access control controller.

[0011] Furthermore, S1 specifically includes:

[0012] Real-time acquisition of environmental physical parameters distributed at different physical nodes within the data center protection zone as multi-dimensional physical environment monitoring data, and acquisition of the linkage status of the fire control panel as a gas release status signal; wherein, the multi-dimensional physical environment monitoring data includes oxygen concentration values, extinguishing agent concentration values, and temperature and humidity values ​​distributed at different spatial heights within the protection zone.

[0013] The gas release status signal is analyzed to extract the start timestamp of the gas release and start timing to determine whether the gas release duration has reached the preset immersion time.

[0014] When the gas release duration reaches the preset immersion time, the environmental recovery process is triggered to synchronize the time dimension and align the spatial domain of the multi-dimensional physical environment monitoring data collected at different frequencies, thereby generating the target environmental monitoring data.

[0015] Furthermore, S2 specifically includes:

[0016] Establish a three-dimensional spatial coordinate system that matches the physical boundary of the data center protection zone and discretize it into a set of grids containing multiple spatial grid nodes;

[0017] The target environmental monitoring data is assigned to the spatial grid nodes corresponding to its coordinates, and the physical parameter estimates of unknown spatial grid nodes without sensor coverage are calculated using a spatial interpolation algorithm to construct a three-dimensional concentration field matrix covering the entire globe.

[0018] The three-dimensional concentration field matrix is ​​traversed, and the concentration difference between adjacent spatial grid nodes is calculated along the three coordinate axes of the three-dimensional spatial coordinate system to extract data gradient features. The target environmental monitoring data and the data gradient features are then dimensionally concatenated and multidimensional tensor fused to generate a comprehensive environmental safety feature vector.

[0019] Furthermore, a three-dimensional spatial coordinate system matching the physical boundary of the data center protection zone is established and discretized into a set of meshes containing multiple spatial mesh nodes, specifically including:

[0020] A Cartesian coordinate system is established as a three-dimensional spatial coordinate system based on the physical boundary of the data center protection zone, and the three-dimensional size parameters of solid obstacles inside the protection zone are input into the three-dimensional spatial coordinate system for elimination marking.

[0021] According to the preset physical resolution, the three-dimensional free space after removing solid obstacles is divided into multiple uniformly arranged cubic grids, and the center point of each cubic grid is defined as a spatial grid node.

[0022] The physical parameters of unknown spatial grid nodes without sensor coverage are estimated using spatial interpolation algorithms, specifically including:

[0023] Centered on the unknown spatial grid node to be searched, multiple known nodes within a preset search radius are selected. Weights are set according to the three-dimensional straight-line distance between the known nodes and the unknown spatial grid node, and the concentration estimate of the unknown spatial grid node is obtained by weighted averaging.

[0024] Furthermore, S3 specifically includes:

[0025] The real-time environmental index is obtained by performing an inner product operation on the comprehensive environmental safety feature vector using a preset safety weight matrix, and then the real-time environmental index is compared with a preset safety baseline.

[0026] When the real-time environmental index does not reach the preset safety baseline, an access control lock maintenance signal is output to the access control controller, and the difference between the real-time environmental index and the preset safety baseline is calculated.

[0027] The difference is substituted into a preset nonlinear decreasing function to calculate the target speed parameter, and a dynamic exhaust control signal containing the target speed parameter is output to the exhaust fan to drive the exhaust fan to perform variable frequency exhaust action.

[0028] Furthermore, the formula for calculating the real-time environmental index is as follows:

[0029]

[0030] in, This represents the real-time environmental index, where n is the total number of dimensions in the comprehensive environmental safety feature vector. For the feature value of the i-th dimension in the comprehensive environmental safety feature vector, The i-th weight coefficient in the safety weight matrix is ​​set as a penalty coefficient, and the weight coefficients in the safety weight matrix corresponding to the data gradient features are set as penalty coefficients.

[0031] The nonlinear decreasing function is:

[0032]

[0033] in, The target rotational speed parameters are calculated at the current moment. This is the rated maximum speed of the exhaust fan. To maintain the minimum speed threshold for preventing leakage under negative pressure in the protected area, e is a natural constant. This is an environmental damping coefficient pre-calibrated based on the rack layout density and airflow resistance within the data center. This is the difference between the real-time environmental index and the preset safety baseline.

[0034] Furthermore, S4 specifically includes:

[0035] Real-time environmental index is monitored, and the trigger time is recorded when the real-time environmental index first reaches the preset safety baseline. The target environmental monitoring data within the preset time window is extracted from the trigger time.

[0036] The least squares method is used to perform k-order polynomial fitting on discrete target environment monitoring data within a preset time window to generate the continuous time function. Its formula is:

[0037]

[0038] Where t is the time variable within the preset time window. The coefficients are the polynomial fitting coefficients calculated using the least squares method, where k is the fitting order.

[0039] For the time function The first derivative is obtained by analytically differentiating the derivative with respect to the time variable t. Its formula is: .

[0040] Furthermore, S5 specifically includes:

[0041] The absolute value of the first derivative within a preset time window is compared with a preset fluctuation threshold in real time. When it is determined that the absolute value of all derivatives is lower than the preset fluctuation threshold, the environmental convergence condition is met.

[0042] A global traversal search is performed on all spatial grid nodes within the three-dimensional concentration field matrix at this time. The global minimum value of oxygen concentration and the global maximum value of fire extinguishing agent concentration in the three-dimensional concentration field matrix are extracted respectively and combined as the worst environmental index in the world to form the concentration extreme value.

[0043] The extreme concentration value is hashed and bound to the current timestamp when the environmental convergence condition is met to generate an environmental safety confirmation record. The environmental safety confirmation record is stored in the database, and at the same time, an access control unlock signal that triggers the relay is output to the access control controller.

[0044] This invention also provides a data center gas fire suppression protection zone linkage unlocking and environmental restoration control device, based on the data center gas fire suppression protection zone linkage unlocking and environmental restoration control method described above. The protection zone is equipped with an exhaust fan and an access control controller. The device includes:

[0045] The data processing module is used to acquire multi-dimensional physical environment monitoring data and gas release status signals from the fire control panel within the data center protection zone. When the gas release status signal indicates that the gas release duration has reached the preset immersion time, the module performs spatiotemporal alignment on the multi-dimensional physical environment monitoring data to obtain target environmental monitoring data.

[0046] The spatial mapping module is used to map the target environmental monitoring data to a three-dimensional spatial coordinate system containing multiple spatial grid nodes to construct a three-dimensional concentration field matrix, and extract the concentration difference between adjacent spatial grid nodes in the three-dimensional concentration field matrix as data gradient features, which are then fused with the target environmental monitoring data to generate a comprehensive environmental safety feature vector.

[0047] The output adjustment module is used to calculate the real-time environmental index based on the comprehensive environmental safety feature vector. When the real-time environmental index does not reach the preset safety baseline, it outputs an access control lock maintenance signal to the access controller and dynamically adjusts the speed of the exhaust fan based on the difference between the real-time environmental index and the preset safety baseline.

[0048] The derivative calculation module is used to calculate the first derivative of the target environmental monitoring data over time within a preset time window, starting from the moment the real-time environmental index reaches the preset safety baseline.

[0049] The storage unlock module is used to extract the concentration extreme value from the three-dimensional concentration field matrix when the absolute value of the first derivative is lower than the preset fluctuation threshold within the preset time window, bind the concentration extreme value with the current timestamp and store it, and output the access control unlock signal to the access control controller.

[0050] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0051] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described method.

[0052] The beneficial effects of this invention are as follows:

[0053] This invention aligns multidimensional physical environment monitoring data in the spatiotemporal domain and constructs a three-dimensional concentration field matrix covering the entire environment. It extracts data gradient features to identify local high-concentration dead zones caused by complex airflow organization, eliminating the blind spot risks brought about by traditional single-point monitoring. At the same time, it dynamically and non-linearly adjusts the exhaust fan speed based on the difference between the real-time environmental index and the preset safety baseline, smoothly transitioning the air volume and preventing secondary entrapment and mixing of high-density residual gas at the bottom. In addition, this invention extracts a time window and verifies the environmental convergence condition by calculating the first derivative of the monitoring data over time. This filters out transient "pseudo-safety" phenomena caused by airflow disturbances or sensor drift, ensuring that the global extreme value is extracted and access control is triggered only after the environment has truly reached a physically stable state, improving the efficiency of environmental recovery and the safety of personnel access. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of a method flow according to an embodiment of the present invention.

[0055] Figure 2 This is a schematic diagram of the device structure according to an embodiment of the present invention.

[0056] Figure 3 This is a schematic diagram of the internal structure of a computer device according to an embodiment of the present invention.

[0057] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0058] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0059] like Figure 1As shown, this invention provides a method for linkage unlocking and environmental restoration control of a data center gas fire suppression protection zone. The protection zone is equipped with an exhaust fan and an access control controller. The method includes the following steps:

[0060] S1. Acquire multi-dimensional physical environment monitoring data and gas release status signals from the fire control panel within the data center protection zone. When the gas release status signal indicates that the gas release duration has reached the preset immersion time, perform spatiotemporal alignment on the multi-dimensional physical environment monitoring data to obtain target environment monitoring data.

[0061] In a specific embodiment of the present invention, step S1 specifically includes the following sub-steps:

[0062] S101. Real-time acquisition of environmental physical parameters distributed at different physical nodes within the protected area as multi-dimensional physical environment monitoring data, and acquisition of the linkage status of the fire control panel as a gas release status signal; parsing the gas release status signal, extracting the start timestamp of the gas release and starting the timing, and determining whether the gas release duration has reached the preset immersion time.

[0063] Specifically, due to the presence of raised floors, server racks, and enclosed hot and cold aisle structures within data centers, monitoring from a single plane cannot reflect the true physical environment. The multi-dimensional physical environment monitoring data includes real-time values ​​collected by oxygen concentration sensors, fire extinguishing agent concentration sensors, and temperature and humidity sensors distributed at different spatial heights within the protected area (e.g., below the raised floor, in the vertical middle of the server rack, and at the ceiling return air vent).

[0064] Simultaneously, the fire control panel is monitored in real time via underlying communication protocols (such as RS485 or dry contact signals) to obtain gas release status signals characterizing the progress of the fire extinguishing system. After the gas extinguishing system (such as IG541 or heptafluoropropane system) is activated, a sealed impregnation period is necessary to ensure that the extinguishing agent can fully flood and effectively extinguish deep-seated fires. The difference between the current time and the start timestamp is calculated in real time to obtain the gas release duration. When the gas release duration reaches the preset impregnation time set by the standard or on-site (e.g., 10 to 20 minutes), it indicates that the physical extinguishing phase has ended, and the subsequent environmental recovery process is automatically triggered. If the preset impregnation time is not reached, the sealed state of the protected area is maintained and the timer continues.

[0065] S102. When the gas release status signal indicates that the gas release duration has reached the preset immersion time and the environmental recovery process is triggered, the time dimension synchronization is performed on the multi-dimensional physical environment monitoring data with different acquisition frequencies, and the spatial domain alignment is further performed on the multi-dimensional physical environment monitoring data that has been synchronized in the time dimension, and finally the target environmental monitoring data is generated.

[0066] Specifically, due to the physical differences in the hardware sampling periods of temperature and humidity sensors, oxygen concentration sensors, and fire extinguishing agent concentration sensors (e.g., temperature and humidity are sampled once per second, while gas concentration is sampled once every three seconds), direct fusion would lead to data misalignment. Therefore, the timestamps of the data from each sensor are first extracted to establish a unified reference timestamp sequence. Then, for monitoring data that deviates from this reference timestamp, a cubic spline interpolation algorithm is used to smoothly fit the data points, thereby strictly aligning the monitoring data of all dimensions to the reference timestamp on the time axis and eliminating time-domain asynchrony caused by inconsistent hardware sampling rates.

[0067] After completing the above time-dimensional synchronization, the sensor deployment ledger pre-stored in the underlying database is read to obtain the physical installation coordinates of each sensor. The monitoring values ​​for each dimension, after being synchronized over time, are then bound and associated with their corresponding physical installation coordinate labels.

[0068] After the combined processing of timestamp interpolation fitting and spatial coordinate binding described above, spatiotemporal alignment of the multidimensional physical environment monitoring data is achieved, and the output dataset is the target environment monitoring data. This target environment monitoring data is not only on the same cross-section in the time dimension, but each data point also carries accurate spatial location attributes, providing the underlying data foundation for constructing a three-dimensional concentration field matrix.

[0069] S2. The target environment monitoring data is mapped to a three-dimensional spatial coordinate system containing multiple spatial grid nodes to construct a three-dimensional concentration field matrix, and the concentration difference between adjacent spatial grid nodes in the three-dimensional concentration field matrix is ​​extracted as a data gradient feature, which is then fused with the target environment monitoring data to generate a comprehensive environmental safety feature vector.

[0070] In a specific embodiment of the present invention, step S2 specifically includes the following sub-steps:

[0071] S201. Establish a three-dimensional spatial coordinate system that matches the physical boundary of the data center protection zone and discretize it into a set of grids containing multiple spatial grid nodes. Then, assign the target environmental monitoring data to the spatial grid nodes with corresponding coordinates and use a spatial interpolation algorithm to calculate the estimated physical parameters of unknown spatial grid nodes without sensor coverage in order to construct a three-dimensional concentration field matrix covering the entire area.

[0072] Specifically, firstly, a Cartesian coordinate system is established as the three-dimensional spatial coordinate system based on the physical boundaries (length, width, and height) of the data center protection zone. Then, the three-dimensional dimensions of solid obstructions within the protection zone (such as rack layout boundaries, hot and cold aisle enclosures, raised floor partitions, etc.) are input into the three-dimensional spatial coordinate system for removal marking. Subsequently, according to a preset physical resolution (e.g., 0.5m × 0.5m × 0.5m), the effective three-dimensional free space is divided into multiple uniformly arranged cubic grids, with the center point of each cubic grid defined as a spatial grid node.

[0073] After defining the grid nodes, the target environment monitoring data generated in step S1 already carries clear physical installation coordinates. At this point, these actually collected monitoring values ​​are directly assigned to the spatial grid nodes (i.e., known nodes) that coincide with or are closest to each other in the three-dimensional spatial coordinate system.

[0074] However, since it is impossible to deploy sensors on all grids in practical engineering, an inverse distance weighted interpolation (IDW) algorithm is used for numerical extrapolation of the remaining spatial grid nodes without sensor coverage. Specifically, taking the unknown spatial grid node to be determined as the center, multiple known nodes within a preset search radius are selected. Weights are assigned based on the three-dimensional straight-line distance between the known nodes and the unknown node (the closer the distance, the greater the weight). The estimated values ​​of oxygen concentration and extinguishing agent concentration for the unknown node are calculated by weighted averaging. After the aforementioned comprehensive interpolation calculation, all spatial grid nodes are successfully assigned concentration values, thus ultimately generating a complete three-dimensional concentration field matrix covering the entire area. .

[0075] S202. Traverse the three-dimensional concentration field matrix and calculate the concentration difference between adjacent spatial grid nodes along the three coordinate axes of the three-dimensional spatial coordinate system to extract data gradient features. Then, perform dimensional splicing and multi-dimensional tensor fusion of the target environmental monitoring data and data gradient features to generate a comprehensive environmental safety feature vector.

[0076] Specifically, the complex airflow organization in data centers easily creates "localized high-concentration dead zones" (such as the bottom of server racks or under anti-static floors). To accurately identify these dead zones, we first traverse each spatial grid node in the constructed three-dimensional concentration field matrix, and use a discrete difference algorithm to calculate the absolute value of the concentration difference between it and its neighboring nodes in the three orthogonal directions of the X-axis, Y-axis, and Z-axis (e.g., These sets of differences, which characterize the severity of spatial concentration changes, constitute the data gradient feature, which can capture abnormal boundaries where gas cannot be effectively dispersed due to airflow obstruction.

[0077] After extracting the aforementioned features, considering that relying solely on the absolute concentration values ​​of a single node can easily lead to misjudgments, and that combining spatial variation trends is necessary to reflect the true safety status, the target environmental monitoring data containing actual absolute concentration values ​​was further concatenated with the previously calculated data gradient features (i.e., the relative variation features of spatial distribution) using a multi-dimensional array. The concatenated tensor was then normalized to generate a comprehensive environmental safety feature vector characterizing the current absolute state and spatial variation trend of the protected area. This feature vector contains both the specific environmental parameters at each point within the protected area and the step-like variation law of gas spatial distribution, providing data support for the subsequent nonlinear adjustment of exhaust fan speed.

[0078] S3. Calculate the real-time environmental index based on the comprehensive environmental safety feature vector. When the real-time environmental index does not reach the preset safety baseline, output the access control lock maintenance signal to the access control controller, and dynamically adjust the speed of the exhaust fan based on the difference between the real-time environmental index and the preset safety baseline.

[0079] In a specific embodiment of the present invention, step S3 specifically includes the following sub-steps:

[0080] S301. Calculate the real-time environmental index by performing an inner product operation on the comprehensive environmental safety feature vector using a preset safety weight matrix, and compare it with a preset safety baseline. When the real-time environmental index does not reach the preset safety baseline, output a power-off / power-on access control signal to the access controller, and calculate the difference between the real-time environmental index and the preset safety baseline.

[0081] Specifically, the comprehensive environmental safety feature vector output in step S2 includes the absolute concentration values ​​at multiple points in the protected area and the data gradient features characterizing abrupt changes in the spatial distribution of gas. Let the comprehensive environmental safety feature vector be a row vector containing n dimensions. A safety weight matrix is ​​pre-defined and stored. This matrix is ​​represented in data structure as a column vector containing n weight coefficients. This matrix assigns higher weight penalty coefficients to the data gradient features representing local high-concentration blind spots and the maximum values ​​of the underlying fire extinguishing agent concentration. The inner product operation is performed between the comprehensive environmental safety feature vector F and the safety weight matrix W, with the following formula:

[0082]

[0083] Ultimately, the multidimensional tensor is reduced to a scalar value, namely the real-time environment index. As a specific example, suppose the comprehensive environmental safety feature vector, after normalization, contains three core dimensions: the average extinguishing agent concentration in the protected area. Maximum global extinguishing agent concentration and spatial maximum concentration gradient characteristics To prevent blind spots, the preset safety weight matrix can be configured as follows: When performing inner product operations, the data gradient characteristics representing local high-concentration clusters are considered. It was given the highest weighting penalty coefficient (0.5), which made the calculated It can reflect risks in hidden blind spots with extreme sensitivity.

[0084] Continuously calculate the results Compared with the preset safety baseline representing the human body safety access standard ( Perform numerical comparison. When Not achieved When the oxygen concentration in the environment is insufficient or the concentration of toxic exhaust gas exceeds the standard, the underlying control logic continuously outputs high / low level access control lock maintenance signals to the access controller at the physical boundary of the protected area through the I / O interface to maintain the locking state of the electromagnetic lock or electric bolt lock and prevent personnel from entering by mistake.

[0085] Simultaneously, the absolute value of the difference between the real-time environmental index and the preset safety baseline is calculated and extracted in real time. This serves as the reference input parameter for subsequent frequency conversion control.

[0086] S302. Substitute the difference into the preset nonlinear decreasing function to calculate the target speed parameter, and output a dynamic exhaust control signal containing the target speed parameter to the exhaust fan to drive the exhaust fan to perform variable frequency exhaust action according to the nonlinear decay curve.

[0087] Specifically, traditional exhaust control often uses constant speed operation at industrial frequency. This can easily lead to the re-entrainment of heavily polluted exhaust gases (such as heptafluoropropane fire extinguishing agent, which is denser than air) that have settled below the raised floor due to excessive airflow in the later stages of exhaust, causing secondary fluctuations in environmental concentration. To overcome this deficiency, the static output logic is modified to change the difference calculated in step S301. Substituting the parameters into the preset nonlinear decreasing function, the target rotational speed parameter V(t) of the exhaust fan at time t is dynamically calculated. The specific formula of the nonlinear decreasing function is as follows:

[0088]

[0089] in, This refers to the rated maximum speed of the exhaust fan; The minimum rotational speed threshold required to maintain negative pressure and prevent leakage in the protected area; e is a natural constant; This is the environmental damping factor, which is pre-calibrated based on the rack density and airflow resistance within the data center.

[0090] Based on calculations The corresponding analog adjustment signal (such as 4-20mA or 0-10V signal) or bus control command is generated as the dynamic exhaust control signal and sent to the frequency converter driver of the exhaust fan. Based on the mathematical characteristics of the nonlinear decreasing function, when... When the concentration of exhaust gas is relatively high (i.e., in the early stages of environmental recovery, when the concentration of exhaust gas is extremely high), the exhaust fan operates at near full load speed. Operation enables rapid exhaust of large air volumes; as exhaust proceeds, As the ambient temperature gradually decreases, the exhaust fan speed exhibits an exponential, smooth decline; when the environment approaches a preset safety baseline, the speed smoothly transitions to a low-wind state. This effectively prevents secondary winch and mixing of high-density residual gas at the bottom at the physical level, thereby improving the stability and recovery efficiency of environmental convergence.

[0091] S4. When the real-time environmental index reaches the preset safety baseline, the first derivative of the target environmental monitoring data over time within the preset time window is calculated, starting from the moment the preset safety baseline is reached.

[0092] In a specific embodiment of the present invention, step S4 specifically includes the following sub-steps:

[0093] S401. Monitor the real-time environmental index in real time, and record the trigger time when the real-time environmental index first reaches the preset safety baseline. Use the trigger time as the starting point to extract the target environmental monitoring data within the preset time window.

[0094] Specifically, as the exhaust fan continues to operate, the concentration of harmful gases in the protected area gradually decreases. Real-time environmental indices are continuously compared within a global clock cycle. With preset safety baseline .when First drop to When the concentration is 100 or below, it indicates that the protected area has reached the safety access standard in terms of transient values. However, since the exhaust fan is still running at this time, the strong airflow disturbance inside the machine room can easily cause the sensor to collect a brief "false safe" low concentration value.

[0095] To verify the true convergence of the environmental state, the unlocking action is not executed immediately; instead, the timestamp at that moment is marked as the trigger point. Then with Starting from [the point], extend forward to open a [width] [value]. Preset time window (e.g., set) (30 seconds or 60 seconds). Within this time window During the continuous period, high-frequency extraction of target environmental monitoring data generated within that time period that has been aligned in the spatiotemporal domain forms a discrete data set containing time series.

[0096] S402. Perform polynomial fitting on the discrete target environmental monitoring data intercepted within the preset time window to generate a continuous time-concentration function, and calculate the derivative of the function to obtain the first derivative of the target environmental monitoring data as a function of time.

[0097] Specifically, since the raw target environment monitoring data collected by the sensor under airflow disturbance is discrete and accompanied by high-frequency noise jitter, if the rate of change is calculated by directly subtracting two adjacent discrete data points and dividing by the time difference (i.e., forward difference method), the noise will be amplified sharply, causing the derivative curve to oscillate violently and failing to reflect the real physical environment change trend.

[0098] To address this technical challenge, the least squares method is employed to perform k-order polynomial fitting on discrete target environmental monitoring data (such as a set of time-extinguishing agent concentration data points at a specific coordinate node) within a preset time window. Let the fitted continuous function be... Its formula is:

[0099]

[0100] Where t is the time variable within the time window. These are the polynomial fitting coefficients calculated using the least squares method.

[0101] In obtaining smooth continuous functions Then, the first derivative of the function is obtained by analytically differentiating the function with respect to time t. Its expression is:

[0102]

[0103] The first derivative By eliminating transient electrical signal noise from the sensor, it accurately and continuously represents the true rate of change of gas concentration in the protected area at the physical level within a preset time window, providing a basis for judging whether the environment has reached true "stable convergence".

[0104] S5. When the absolute value of the first derivative is lower than the preset fluctuation threshold within the preset time window, the concentration extreme value is extracted from the three-dimensional concentration field matrix at this time, the concentration extreme value is bound and stored with the current timestamp, and the access control unlock signal is output to the access control controller.

[0105] In a specific embodiment of the present invention, step S5 specifically includes the following sub-steps:

[0106] S501. In real time, compare the absolute value of the first derivative of each dimension with the preset fluctuation threshold within the preset time window. When it is determined that all the absolute values ​​of the derivatives are lower than the preset fluctuation threshold and the environmental convergence condition is met, traverse the three-dimensional concentration field matrix at this time to extract the worst global environmental index as the concentration extreme value.

[0107] Specifically, the first derivative calculated in step S4 Physically, it represents the instantaneous rate of change of gas concentration within the computer room. Over the entire time window... Continue to check whether the condition is met. (in This is a pre-calibrated small constant, i.e., a preset fluctuation threshold. If the rate of change of concentration in all dimensions (including oxygen and extinguishing agent) does not exceed this threshold throughout the entire time window, it indicates that not only are the values ​​within the protected area up to standard, but the airflow diffusion and replacement process has also stabilized, eliminating the "false safety" phenomenon caused by local vortices or sensor drift.

[0108] After determining that the environment has reached a true physical convergence state, the three-dimensional concentration field matrix at that moment is frozen to ensure the safety of the extreme dead zone. The system performs a global traversal search of all spatial grid nodes within the matrix. It extracts the global minimum value of oxygen concentration (i.e., the most oxygen-deficient point) and the global maximum value of fire extinguishing agent concentration (i.e., the point of highest exhaust gas concentration) from the matrix, and combines these two data points representing the worst environmental conditions inside the current computer room as the concentration extreme values. This safety fallback mechanism based on spatial global extreme values ​​eliminates the blind spot risks associated with traditional single-point monitoring.

[0109] S502. The extracted concentration extreme value is bound to the current timestamp when the environmental convergence condition is met using a hash algorithm to generate an environmental safety confirmation record. This record is written into the underlying database for storage. At the same time, the access control controller is output with a relay-triggered access unlocking signal through the underlying hardware interface.

[0110] Specifically, because fire-fighting coordination involves strict safety audits and accident tracing mechanisms, on-site data must be solidified before the physical lock is released. The precise system time at the moment the convergence condition is met is obtained as the current timestamp. The extreme concentration values ​​(minimum oxygen value, maximum extinguishing agent value) extracted in step S501 are encrypted and concatenated with the current timestamp using a hash algorithm (such as SHA-256) to generate an environmental safety confirmation record with a unique data signature. Subsequently, the environmental safety confirmation record is directly and persistently written to the underlying relational database or time-series database for storage via a database operation interface (such as SQL statements). This storage process ensures that the historical state of environmental restoration meeting standards is tamper-proof and traceable.

[0111] In the parallel threads executing data ingestion and storage, the underlying I / O control logic outputs specific commands or high / low level signals as access control unlock signals to the access controller at the physical boundary via hardwiring or an industrial communication bus (such as the RS485 protocol). Upon receiving this signal, the access controller drives an internal relay to cut off the power supply circuit of the electromagnetic lock or electric bolt lock, thereby unlocking the protected area and allowing secure access for data center maintenance personnel.

[0112] like Figure 2 As shown, the present invention also provides a data center gas fire suppression protection zone linkage unlocking and environmental restoration control device, based on the data center gas fire suppression protection zone linkage unlocking and environmental restoration control method described above. The protection zone is equipped with an exhaust fan and an access control controller. The device includes:

[0113] Data processing module 1 is used to acquire multi-dimensional physical environment monitoring data and gas release status signals from the fire control panel within the data center protection zone. When the gas release status signal indicates that the gas release duration has reached the preset immersion time, the multi-dimensional physical environment monitoring data is spatiotemporally aligned to obtain target environmental monitoring data.

[0114] The spatial mapping module 2 is used to map the target environmental monitoring data to a three-dimensional spatial coordinate system containing multiple spatial grid nodes to construct a three-dimensional concentration field matrix, and extract the concentration difference between adjacent spatial grid nodes in the three-dimensional concentration field matrix as data gradient features, and fuse it with the target environmental monitoring data to generate a comprehensive environmental safety feature vector.

[0115] Output adjustment module 3 is used to calculate the real-time environmental index based on the comprehensive environmental safety feature vector. When the real-time environmental index does not reach the preset safety baseline, it outputs an access control lock maintenance signal to the access controller and dynamically adjusts the speed of the exhaust fan based on the difference between the real-time environmental index and the preset safety baseline.

[0116] The derivative calculation module 4 is used to calculate the first derivative of the target environmental monitoring data over time within a preset time window, starting from the moment the real-time environmental index reaches the preset safety baseline.

[0117] The storage unlocking module 5 is used to extract the concentration extreme value from the three-dimensional concentration field matrix when the absolute value of the first derivative is lower than the preset fluctuation threshold within the preset time window, bind the concentration extreme value with the current timestamp and store it, and output the access control unlocking signal to the access control controller.

[0118] Each of the above modules is used to perform the corresponding steps in the above-mentioned data center gas fire suppression protection zone linkage unlocking and environmental restoration control method. The specific implementation method is as described in the above-mentioned method embodiment, and will not be repeated here.

[0119] like Figure 3 As shown, the present invention also provides a computer device, which may be a server, and its internal structure may be as follows: Figure 3 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores all data required for the data center gas fire suppression zone linkage unlocking and environmental restoration control method. The network interface communicates with external terminals via a network connection. The computer program is executed by the processor to implement the data center gas fire suppression zone linkage unlocking and environmental restoration control method.

[0120] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer equipment on which the present application is applied.

[0121] An embodiment of this application also provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the above-described method for linkage unlocking and environmental restoration control of any one of the data center gas fire suppression protection zones.

[0122] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by hardware related to computer program instructions. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the above method embodiments. Any references to memory, storage, databases, or other media provided in this application and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM), such as dynamic RAM (used as main storage) or static RAM (commonly used as cache memory). By way of illustration and not limitation, RAM has various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), and Rambus DRAM (RDRAM).

[0123] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0124] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for linkage unlocking and environmental restoration control of a data center gas fire suppression protection zone, characterized in that, The protected area is equipped with an exhaust fan and an access control system. The method includes the following steps: S1. Acquire multi-dimensional physical environment monitoring data and gas release status signals from the fire control panel within the data center protection zone. When the gas release status signal indicates that the gas release duration has reached the preset immersion time, perform spatiotemporal alignment on the multi-dimensional physical environment monitoring data to obtain target environment monitoring data. S2. Map the target environmental monitoring data to a three-dimensional spatial coordinate system containing multiple spatial grid nodes to construct a three-dimensional concentration field matrix, and extract the concentration difference between adjacent spatial grid nodes in the three-dimensional concentration field matrix as data gradient features, and fuse it with the target environmental monitoring data to generate a comprehensive environmental safety feature vector. S3. Calculate the real-time environmental index based on the comprehensive environmental safety feature vector. When the real-time environmental index does not reach the preset safety baseline, output the access control lock maintenance signal to the access control controller, and dynamically adjust the speed of the exhaust fan based on the difference between the real-time environmental index and the preset safety baseline. S4. When the real-time environmental index reaches the preset safety baseline, the first derivative of the target environmental monitoring data over time within the preset time window is calculated, starting from the moment the preset safety baseline is reached. S5. When the absolute value of the first derivative is lower than the preset fluctuation threshold within the preset time window, the concentration extreme value is extracted from the three-dimensional concentration field matrix at this time, the concentration extreme value is bound and stored with the current timestamp, and the access control unlock signal is output to the access control controller.

2. The data center gas fire suppression protection zone linkage unlocking and environmental restoration control method according to claim 1, characterized in that, S1 specifically includes: Real-time acquisition of environmental physical parameters distributed at different physical nodes within the data center protection zone as multi-dimensional physical environment monitoring data, and acquisition of the linkage status of the fire control panel as a gas release status signal; wherein, the multi-dimensional physical environment monitoring data includes oxygen concentration values, extinguishing agent concentration values, and temperature and humidity values ​​distributed at different spatial heights within the protection zone. The gas release status signal is analyzed to extract the start timestamp of the gas release and start timing to determine whether the gas release duration has reached the preset immersion time. When the gas release duration reaches the preset immersion time, the environmental recovery process is triggered to synchronize the time dimension and align the spatial domain of the multi-dimensional physical environment monitoring data collected at different frequencies, thereby generating the target environmental monitoring data.

3. The data center gas fire suppression protection zone linkage unlocking and environmental restoration control method according to claim 1, characterized in that, S2 specifically includes: Establish a three-dimensional spatial coordinate system that matches the physical boundary of the data center protection zone and discretize it into a set of grids containing multiple spatial grid nodes; The target environmental monitoring data is assigned to the spatial grid nodes corresponding to its coordinates, and the physical parameter estimates of unknown spatial grid nodes without sensor coverage are calculated using a spatial interpolation algorithm to construct a three-dimensional concentration field matrix covering the entire globe. The three-dimensional concentration field matrix is ​​traversed, and the concentration difference between adjacent spatial grid nodes is calculated along the three coordinate axes of the three-dimensional spatial coordinate system to extract data gradient features. The target environmental monitoring data and the data gradient features are then dimensionally concatenated and multidimensional tensor fused to generate a comprehensive environmental safety feature vector.

4. The data center gas fire suppression protection zone linkage unlocking and environmental restoration control method according to claim 3, characterized in that, Establish a three-dimensional spatial coordinate system that matches the physical boundary of the data center protection zone and discretize it into a set of meshes containing multiple spatial mesh nodes, specifically including: A Cartesian coordinate system is established as a three-dimensional spatial coordinate system based on the physical boundary of the data center protection zone, and the three-dimensional size parameters of solid obstacles inside the protection zone are input into the three-dimensional spatial coordinate system for elimination marking. According to the preset physical resolution, the three-dimensional free space after removing solid obstacles is divided into multiple uniformly arranged cubic grids, and the center point of each cubic grid is defined as a spatial grid node. The physical parameters of unknown spatial grid nodes without sensor coverage are estimated using spatial interpolation algorithms, specifically including: Centered on the unknown spatial grid node to be searched, multiple known nodes within a preset search radius are selected. Weights are set according to the three-dimensional straight-line distance between the known nodes and the unknown spatial grid node, and the concentration estimate of the unknown spatial grid node is obtained by weighted averaging.

5. The data center gas fire suppression protection zone linkage unlocking and environmental restoration control method according to claim 1, characterized in that, S3 specifically includes: The real-time environmental index is obtained by performing an inner product operation on the comprehensive environmental safety feature vector using a preset safety weight matrix, and then the real-time environmental index is compared with a preset safety baseline. When the real-time environmental index does not reach the preset safety baseline, an access control lock maintenance signal is output to the access control controller, and the difference between the real-time environmental index and the preset safety baseline is calculated. The difference is substituted into a preset nonlinear decreasing function to calculate the target speed parameter, and a dynamic exhaust control signal containing the target speed parameter is output to the exhaust fan to drive the exhaust fan to perform variable frequency exhaust action.

6. The data center gas fire suppression protection zone linkage unlocking and environmental restoration control method according to claim 5, characterized in that, The formula for calculating the real-time environmental index is as follows: in, This represents the real-time environmental index, where n is the total number of dimensions in the comprehensive environmental safety feature vector. For the feature value of the i-th dimension in the comprehensive environmental safety feature vector, The i-th weight coefficient in the safety weight matrix is ​​set as a penalty coefficient, and the weight coefficients in the safety weight matrix corresponding to the data gradient features are set as penalty coefficients. The nonlinear decreasing function is: in, The target rotational speed parameters are calculated at the current moment. This is the rated maximum speed of the exhaust fan. To maintain the minimum speed threshold for preventing leakage under negative pressure in the protected area, e is a natural constant. This is an environmental damping coefficient pre-calibrated based on the rack layout density and airflow resistance within the data center. This is the difference between the real-time environmental index and the preset safety baseline.

7. The data center gas fire suppression protection zone linkage unlocking and environmental restoration control method according to claim 1, characterized in that, S4 specifically includes: Real-time environmental index is monitored, and the trigger time is recorded when the real-time environmental index first reaches the preset safety baseline. The target environmental monitoring data within the preset time window is extracted from the trigger time. The least squares method is used to perform k-order polynomial fitting on discrete target environment monitoring data within a preset time window to generate the continuous time function. Its formula is: Where t is the time variable within the preset time window. The coefficients are the polynomial fitting coefficients calculated using the least squares method, where k is the fitting order. For the time function The first derivative is obtained by analytically differentiating the derivative with respect to the time variable t. Its formula is: .

8. The data center gas fire suppression protection zone linkage unlocking and environmental restoration control method according to claim 1, characterized in that, S5 specifically includes: The absolute value of the first derivative within a preset time window is compared with a preset fluctuation threshold in real time. When it is determined that the absolute value of all derivatives is lower than the preset fluctuation threshold, the environmental convergence condition is met. A global traversal search is performed on all spatial grid nodes within the three-dimensional concentration field matrix at this time. The global minimum value of oxygen concentration and the global maximum value of fire extinguishing agent concentration in the three-dimensional concentration field matrix are extracted respectively and combined as the worst environmental index in the world to form the concentration extreme value. The extreme concentration value is hashed and bound to the current timestamp when the environmental convergence condition is met to generate an environmental safety confirmation record. The environmental safety confirmation record is stored in the database, and at the same time, an access control unlock signal that triggers the relay is output to the access control controller.

9. A data center gas fire suppression protection zone linkage unlocking and environmental restoration control device, based on the data center gas fire suppression protection zone linkage unlocking and environmental restoration control method according to any one of claims 1 to 8, characterized in that, The protected area is equipped with an exhaust fan and an access control system. The device includes: The data processing module is used to acquire multi-dimensional physical environment monitoring data and gas release status signals from the fire control panel within the data center protection zone. When the gas release status signal indicates that the gas release duration has reached the preset immersion time, the module performs spatiotemporal alignment on the multi-dimensional physical environment monitoring data to obtain target environmental monitoring data. The spatial mapping module is used to map the target environmental monitoring data to a three-dimensional spatial coordinate system containing multiple spatial grid nodes to construct a three-dimensional concentration field matrix, and extract the concentration difference between adjacent spatial grid nodes in the three-dimensional concentration field matrix as data gradient features, which are then fused with the target environmental monitoring data to generate a comprehensive environmental safety feature vector. The output adjustment module is used to calculate the real-time environmental index based on the comprehensive environmental safety feature vector. When the real-time environmental index does not reach the preset safety baseline, it outputs an access control lock maintenance signal to the access controller and dynamically adjusts the speed of the exhaust fan based on the difference between the real-time environmental index and the preset safety baseline. The derivative calculation module is used to calculate the first derivative of the target environmental monitoring data over time within a preset time window, starting from the moment the real-time environmental index reaches the preset safety baseline. The storage unlock module is used to extract the concentration extreme value from the three-dimensional concentration field matrix when the absolute value of the first derivative is lower than the preset fluctuation threshold within the preset time window, bind the concentration extreme value with the current timestamp and store it, and output the access control unlock signal to the access control controller.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

Images