A machine room cold aisle sealing management method, device and terminal equipment

Abstract

The application relates to the technical field of data center computer room energy-saving, and discloses a computer room cold aisle sealing management method, a device and terminal equipment, the method comprises the following steps: obtaining a cold aisle state through a deployment environment monitoring node, constructing a heat-flow coupling model and a stability function, designing a feedback controller to generate a damper and refrigeration control instruction, realizing linkage regulation of air flow rate and temperature, constructing a closed-loop control, and realizing efficient energy-saving management of the cold aisle; the device comprises an environment monitoring module, a state modeling module, a stability evaluation module, a state feedback control module and an execution linkage module. Through state fusion updating, component control structure, stability function driving strategy and linkage execution mechanism, the application realizes fine adjustment and efficient collaborative control of the cold aisle temperature and airflow, improves regulation stability and response efficiency, significantly reduces energy consumption, and solves the problems of poor control precision and weak linkage of the existing system.

Description

Technical Field

[0001] This invention relates to the field of energy-saving technology for data center computer rooms, specifically to a method, device, and terminal equipment for closed management of cold aisles in computer rooms. Background Technology

[0002] In the current context of green data center construction, energy efficiency has become a core consideration. The widespread deployment of cold aisle containment systems aims to reduce energy waste caused by the mixing of hot and cold airflows. However, traditional cold aisle systems mostly rely on fixed physical enclosures, such as simple separation using fences, doors, or covers. While these solutions achieve preliminary thermal isolation, they lack sensing capabilities and cannot respond to real-time changes in thermal conditions.

[0003] Currently, some systems attempt to introduce basic sensor monitoring methods, such as temperature probes and infrared detection. However, these devices often have limited deployment, provide only single feedback, and cannot form an effective feedback control closed loop. The common practice still relies on manual inspection or timed adjustments to dampers and air conditioning parameters. The entire system operates essentially in a "passive response" state, with coarse adjustment granularity and long adjustment cycles, making it difficult to adapt to rapidly changing data center load environments. Especially when there are slight changes in airtightness, the system often remains unaware, and cooling capacity is wasted unnoticed.

[0004] Furthermore, existing technologies generally lack global optimization capabilities based on the heat-flow coupling relationship in terms of control strategies. Control commands are mostly issued as point-to-point setpoints, lacking coordination logic between dampers and refrigeration equipment, resulting in unclear adjustment directions and conflicting execution sequences. Once heat accumulates in a certain area, the system may simply increase the cooling power while ignoring the reconstruction of the airflow guidance path, leading to frequent situations of "cooling without effect." The resulting decrease in cooling efficiency and increase in energy consumption are difficult to fundamentally solve using traditional methods.

[0005] Therefore, this invention proposes a method, device, and terminal equipment for the closed management of cold aisles in computer rooms to address the shortcomings of existing technologies. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method, device, and terminal equipment for the closed management of cold aisles in computer rooms, which solves the problems of incomplete status perception, delayed adjustment response, and poor equipment linkage in existing systems, making it difficult to achieve high-efficiency control.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for closed management of cold aisle passageways in computer rooms, comprising the following steps:

[0008] S1. Deploy environmental monitoring nodes at key locations within the cold aisle area to collect air temperature and airflow velocity at key locations in real time, and integrate the environmental parameters at each key location to construct a cold aisle state vector.

[0009] S2. Input the state vector into the thermal-fluid coupling state space modeling module to establish a state space model of the temperature and airflow velocity inside the cold aisle as a function of time. The model is used to dynamically describe the state evolution process of the system under multi-source disturbance conditions. S3. Based on the state space model, construct a stability function to evaluate the stability of the thermal-fluid field in the cold aisle. The stability function is represented by a quadratic function composed of the state vector and a symmetric positive definite matrix, and is used to monitor the stability deviation of the system under the influence of the current disturbance in real time.

[0010] S4. Based on the evolution trend of the stability function, design a state feedback controller and use the current state vector to calculate the control instructions of the cold aisle execution structure. The control instructions include the damper angle control signal and the refrigeration unit power control signal.

[0011] S5. The control command is synchronously output to the damper actuator and cooling equipment in the cold aisle to adjust the airflow and temperature distribution inside the cold aisle, and the state vector is updated to enter the next cycle, so as to form a closed-loop control process for continuous monitoring and adjustment.

[0012] Preferably, the key locations include the air ducts located at the cold aisle inlet, the front side of the server rack, the rear side of the server rack, and the cold aisle outlet.

[0013] Preferably, the environmental monitoring node includes:

[0014] Temperature sensors are used to detect air temperature at critical locations;

[0015] Wind speed sensor, used to detect airflow speed at key locations;

[0016] Humidity sensor, used to monitor air humidity;

[0017] The data acquisition module is used to integrate measurement data from various sensors and upload them to the host computer.

[0018] Preferably, the state-space model is established in the following form:

[0019] x(t+1)=Ax(t)+Bu(t)+w(t);

[0020] Where x(t) represents the state vector, u(t) represents the control command vector, w(t) represents the disturbance term, and A and B are system matrices.

[0021] Preferably, the stability function is a quadratic function of the following form:

[0022] V(x)=x T Px;

[0023] Where x is the current state vector, x T Let x be the transpose of x, and P be a symmetric positive definite matrix used to measure the deviation of the system's energy level and stability.

[0024] Preferably, the state feedback controller constructs the following optimization objective based on the real-time changing trend of the stability function by minimizing the system deviation:

[0025]

[0026] Where u(t) is the control input vector at time t, x(t+1) represents the system state vector at time t+1, and P is a symmetric positive definite matrix. T Px(t+1) represents the stability function, which is a quadratic function, and V(x(t+1)) represents the stability function value of the system at the next time step.

[0027] Based on the system model x(t+1)=Ax(t)+Bu(t), the optimal control input u(t) is calculated, wherein the control input includes:

[0028] The damper angle adjustment signal based on the disturbance response is used to correct the cold airflow path.

[0029] Cooling power adjustment signals based on heat load fluctuations are used to combat abnormal temperatures.

[0030] The feedback controller is implemented using a linear quadratic regulator method or its improved algorithm, and has the ability to adaptively adjust the gain parameter to meet the control accuracy requirements under different dynamic changes in server load.

[0031] Preferably, the damper angle control signal drives an adjustable guide vane installed at the top or bottom of the cold aisle, thereby adjusting the spatial distribution of airflow direction and flow rate through the linkage of multiple air-guiding actuators, wherein:

[0032] When the monitoring node detects that the temperature at the front end of a rack exceeds the set threshold, the controller will prioritize adjusting the opening angle of the air guide plate above that area to increase the supply of cold air to that area.

[0033] The opening adjustment of multiple air guide vanes is optimized in a coordinated manner based on the changing trend of the local temperature gradient in the state vector, thereby forming a directional cooling path for the hot spot area.

[0034] Preferably, the refrigeration unit power control signal is used to dynamically control the cooling capacity of the refrigeration module in the cooling system, and its control logic includes:

[0035] Based on the state-space model, the temperature trend of multiple future cycles is predicted, and the output power of the refrigeration unit is adjusted in advance to avoid hysteresis.

[0036] In non-uniform heat load scenarios, cooling priority levels are divided according to the distribution characteristics of the state vector, and a hierarchical control strategy is adopted to apply different target powers to each cooling module in order to achieve both local enhanced cooling and global energy efficiency optimization.

[0037] The control signals also include composite adjustment parameters for compressor frequency, cooling water flow rate, or air supply speed, used to improve the overall cooling response bandwidth.

[0038] A cold aisle airtight management device for computer rooms includes:

[0039] The environmental monitoring module is used to collect air temperature, airflow velocity and humidity information at multiple key locations within the cold aisle and generate state vectors.

[0040] The state modeling module is used to construct a state-space model of the cold aisle thermal-fluid field based on state vectors;

[0041] The stability assessment module is used to calculate the stability function and assess the degree of stability deviation in the cold aisle.

[0042] The state feedback control module is used to calculate the damper angle and the refrigeration unit power control signal based on the stability function;

[0043] An execution linkage module is used to output control signals to the damper actuator and cooling equipment to adjust the airflow and temperature distribution in the cold aisle. A terminal device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the aforementioned method for airtight management of a computer room cold aisle.

[0044] This invention provides a method, apparatus, and terminal equipment for the closed management of cold aisles in computer rooms. It has the following beneficial effects:

[0045] 1. This invention constructs a state update mechanism that integrates measured data and model predictions, achieving dynamic correction and filtering of state feedback information, effectively improving the stability and continuity of system control. Compared with existing technologies that rely on single-point real-time measurement for direct drive control, this invention reduces feedback jitter caused by environmental fluctuations and sampling errors, solving the problem of over-adjustment of the system caused by state acquisition errors in existing technologies.

[0046] 2. This invention employs a component control structure that distinguishes between damper angle and cooling power, enabling independent control of different physical areas and providing high adjustment accuracy and flexibility. Compared to the problems of uniform distribution of overall cooling capacity and coarse control granularity in existing technologies, this invention can finely adjust the airflow and cooling intensity within the cold aisle based on local temperature conditions, significantly reducing system energy consumption and solving the problems of low cooling efficiency and resource waste.

[0047] 3. This invention introduces a stability function-driven control strategy generation mechanism, enabling the system adjustment process to no longer rely solely on temperature threshold triggering, but rather to make global judgments and controls based on the dynamic stability state of the entire thermal-fluid field. Compared to existing technologies that simply cool down based on temperature anomalies, this invention possesses stronger foresight and proactive intervention capabilities, solving the problems of delayed response and unclear control direction in existing systems.

[0048] 4. This invention achieves coordinated and linked regulation between the damper actuator and the refrigeration equipment by setting a control signal mapping table and priority scheduling logic in the execution linkage module. Compared with the existing control methods where various devices operate independently and lack coordination mechanisms, this invention improves the overall system regulation efficiency and solves the problems of inconsistent linkage responses and low execution efficiency between devices. Attached Figure Description

[0049] Figure 1 This is a flowchart of the method of the present invention;

[0050] Figure 2 This is a diagram of the device architecture of the present invention;

[0051] Figure 3 This is a schematic diagram of the terminal device structure of the present invention. Detailed Implementation

[0052] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] Please see Figure 1 This invention provides a method for closed management of cold aisles in computer rooms, comprising the following steps:

[0054] S1. Deploy environmental monitoring nodes at key locations within the cold aisle area to collect air temperature and airflow velocity at key locations in real time, and integrate the environmental parameters at each key location to construct a cold aisle state vector.

[0055] In implementing the cold aisle airtight management method of this invention, in order to accurately grasp the typical airflow paths and heat distribution within the cold aisle, thus providing a foundation for subsequent modeling and feedback control, it is necessary to prioritize the deployment of environmental monitoring nodes at key locations within the cold aisle area. This step S1 not only constitutes the sensing input layer of this system but also serves as the data foundation for subsequent state-space model construction S2, stability function calculation S3, and controller execution S4. The quality of data acquisition directly affects the convergence efficiency and dynamic response capability of the entire system's adjustment strategy.

[0056] In this embodiment, monitoring nodes need to be deployed at the following key locations inside the cold aisle in order to construct a high spatiotemporal resolution data sensing network:

[0057] Cold aisle air inlet: This is usually the initial section of the air conditioning supply path, where the air temperature is the lowest and the airflow is the fastest;

[0058] The front of each server rack: the inlet surface before the cold air enters the equipment, used to monitor the effectiveness of air supply;

[0059] The rear of each server rack is the equipment exhaust vent, where the temperature is highest and airflow disturbance is significant.

[0060] Cold aisle outlet: Connects to the hot air return channel, used to determine local return air short-circuiting and airflow distribution balance.

[0061] Generally, at least one monitoring node should be deployed in each of the above areas. As an option, for areas with a rack density of more than 20 units, the deployment density can be increased by configuring one node for every two racks to strengthen the deployment.

[0062] In one possible implementation, the environmental monitoring node includes the following core modules:

[0063] Temperature sensor: Employs digital or analog sensors to measure ambient temperature T in degrees Celsius (°C). Typical measurement range is -10°C to 85°C, with a resolution of 0.1°C. Used to monitor heat accumulation.

[0064] Wind speed sensor: measures the local airflow linear velocity V, in meters per second (m / s), with a typical measurement range of 0 to 10 m / s, used to determine whether the airflow effectively penetrates the rack area;

[0065] Humidity sensor: Used to collect relative humidity (H) of the air, expressed as a percentage (%RH), with a typical range of 0–100%; Data acquisition module: Completes the synchronous acquisition, time-series packaging, encoding, and uploading of temperature, wind speed, and humidity measurement data. Its communication method can be wired (RS485, CAN) or wireless (Wi-Fi, ZigBee, LoRa, etc.) communication protocols.

[0066] In some embodiments, the data acquisition module integrates a real-time clock (RTC) and a cache chip to support breakpoint resume and time series alignment, avoiding sampling misalignment.

[0067] The monitoring node data will serve as the input variable for the system's state-space model. Within each sampling period, the system constructs a set of state vectors. Specifically:

[0068] x(t)=[T1(t),V1(t),H1(t),T2(t),V2(t),H2(t),…,T n (t),V n (t),H n (t)] T ;

[0069] Where x(t) represents the system's state vector at time t, with a dimension of 3n×1; T i (t) represents the temperature value collected by the i-th node, in °C; V i (t) represents the wind speed value collected by the i-th node, in m / s; H i (t) represents the relative humidity value collected by the i-th node, in %RH; n is the number of monitoring nodes; T represents the vector transpose, and the result is a column vector;

[0070] In practical systems, if humidity changes have little impact on modeling, the humidity dimension can be ignored. However, in the complete solution of this invention, it is recommended to retain all dimensions to maintain the uniformity of the input vector structure.

[0071] Before the data is input into the modeling system, it is necessary to normalize each dimension to improve the stability of the modeling values. The normalization process is as follows:

[0072]

[0073] in: Let x be the i-th state variable after normalization; i (t) represents the original state variables; μ i σ is the sample mean of the i-th variable; i Let be the sample standard deviation of the i-th variable;

[0074] This normalization method (Z-score) is robust and suitable for non-mean stable dynamic systems.

[0075] Taking a typical medium-sized data center as an example, it includes 4 cold aisle rows, with 12 server racks in each row. According to the recommended strategy in this embodiment, each row should deploy no fewer than:

[0076] One air inlet node;

[0077] 12 rack front nodes;

[0078] 12 rack back nodes;

[0079] One exit node.

[0080] Each cold aisle has a total of 26 nodes, and the four columns have a total of 104 nodes. The system can automatically construct a state vector position mapping matrix based on the node numbers, which facilitates the subsequent calculation of state space matrices A and B.

[0081] In one possible extended scenario, environmental monitoring nodes can also incorporate infrared thermal imaging or TOF-based spatial ranging capabilities to assist in identifying turbulent airflow regions and reconstructing streamline fields, thereby supporting higher-order state estimation processes.

[0082] S2. Input the state vector into the thermal-fluid coupled state space modeling module to establish a state space model of the temperature and airflow velocity inside the cold aisle changing with time. The model is used to dynamically describe the state evolution process of the system under multi-source disturbance conditions.

[0083] After collecting real-time temperature and wind speed data for key areas of the cold aisle and constructing state vectors, the state vector x(t) constructed in step S1 needs to be input into the thermal-fluid coupling state modeling module to further describe the dynamic evolution of the internal environment of the cold aisle over time. The core function of this module is to establish a mathematical model describing the evolution of the system state based on state-space theory, thereby supporting subsequent stability assessment and feedback control calculations. The model should effectively reflect the temperature-airflow coupling effect, as well as the relationship between control variables and disturbances in the system evolution.

[0084] In this embodiment, a discrete-time state-space model is used to model the cold aisle system. The specific model form is as follows:

[0085] x(t+1)=Ax(t)+Bu(t)+w(t);

[0086] Where: x(t) represents the state vector of the system at time t, with dimension n. x ×1, where n x = 3n, where n is the number of monitoring nodes, and each node corresponds to a temperature T. i Wind speed V i Humidity H i The system has three components: x(t+1) is the predicted state vector of the system at time t+1, with the same meaning as x(t); u(t) is the control command input vector, with dimension n. u ×1, containing the control quantities of each execution module (such as damper opening, refrigeration unit power, etc.) at time t; w(t) is the external disturbance vector with dimension n. x×1 is used to describe uncontrollable factors such as random load changes and nonlinear heat source fluctuations; A is the system state transition matrix with dimension n. x ×n x B describes the influence of the current state on the state at the next time step; B is the control input influence matrix with dimension n. x ×n u This describes the regulatory effect of each control input on different state dimensions.

[0087] In one possible implementation, the construction of matrix A needs to consider the spatial adjacency and physical interactions between state variables, and should have a sparse structure to improve computational efficiency. For example:

[0088] The temperature of a node in the next moment is mainly affected by its own current state and the wind speed and temperature of surrounding nodes.

[0089] The airflow speed is affected by the upstream damper regulation and the nearby heat source.

[0090] Therefore, the non-zero elements of matrix A are only distributed along the diagonal and its nearest neighboring banded regions, forming a banded sparse matrix structure.

[0091] In general, matrix B is also a sparse matrix. Its non-zero elements are located at the mapping positions between a specific state and its controlled input. For example, damper adjustment only affects the wind speed component of a specific region node, while cooling power adjustment mainly affects the temperature dimension of the corresponding region.

[0092] In some embodiments, matrix A can be identified using the following data-driven approach:

[0093]

[0094] Where: T represents the length of the training sequence; λ represents the sparse constraint strength coefficient; ||A||1 represents the L1 norm of the matrix elements, used to control overfitting and enhance interpretability.

[0095] In this embodiment, the perturbation vector w(t) can be modeled as a time-dependent Gaussian process to reflect the persistence and regional coupling of uncontrollable fluctuations in the real system. The definition is as follows:

[0096]

[0097] Where: μ w Σ is the mean vector of the disturbance, usually taken as the zero vector; w Let n be the perturbation covariance matrix, with dimension n. x ×n e Its non-zero structure is similar to that of matrix A, in order to characterize the propagation effect of disturbances among neighboring state variables.

[0098] Specifically, in some instances, the covariance matrix Σ w It can be configured as a block-based diagonal structure, with each block corresponding to a different cold aisle sub-region; or it can be configured as an exponentially decaying spatial correlation model.

[0099]

[0100] Where: σ 2 d represents the disturbance intensity. i,j Let θ be the physical distance between node i and node j; θ is the correlation attenuation factor, which controls the spatial propagation rate of the disturbance.

[0101] To improve the model's generalization ability in dynamic scenarios, the system supports online updating of model parameters. Taking state matrix A as an example, the update method is as follows:

[0102]

[0103] Where: η is the update step size (learning rate), which is generally set to 10. -3 Up to 10 -2 Within the range; This represents the current predicted value; the update strategy is fine-tuned based on the backpropagation of the prediction error.

[0104] In practice, the above update process is triggered once every few sampling periods to avoid model oscillation.

[0105] During the modeling process, to ensure that the predicted state variables meet physical reasonableness, boundary values ​​need to be set for each component of x(t). For example:

[0106] Temperature: T i (t)∈[15,45](unit:℃);

[0107] Wind speed: V i (t)∈[0.2,7.5](unit: m / s);

[0108] Humidity: H i (t)∈[20,90](unit: %RH);

[0109] If the prediction results exceed the boundaries, they can be corrected by truncation or reset strategies to avoid generating abnormal control instructions.

[0110] S3. Based on the state-space model, construct a stability function to evaluate the stability of the cold aisle thermal-fluid field. The stability function is represented by a quadratic function composed of a state vector and a symmetric positive definite matrix, and is used to monitor the stability deviation of the system under the influence of the current disturbance in real time.

[0111] Having completed the thermal-fluid coupled state-space modeling, the system requires a mechanism to monitor in real time whether the current state deviates from the thermodynamic steady state, thereby providing a basis for the controller's decision-making. Therefore, in step S3, a class of mathematical functions, namely "stability functions," is constructed based on the state vector to measure the degree to which the overall system deviates from the steady-state center.

[0112] The core idea of ​​the stability function is to perform a weighted evaluation of the current system state x(t), extract the global features of the energy of state change, and thus compress the high-dimensional multi-source state into a scalar index, simplifying monitoring and judgment.

[0113] In this embodiment, the stability function is constructed using the following standard quadratic form function:

[0114] V(x(t))=x(t) T Px(t);

[0115] Where: V(x(t)) represents the system stability function value at time t, in dimensionless scalar form, used to measure the degree to which the system deviates from steady state; x(t) represents the system's state vector at time t, with dimension n. x ×1, where n x = 3n, where n is the total number of deployed monitoring nodes; x(t) T Let x(t) be the transpose of x(t), with dimension 1×n. x P is the weight matrix in the stability function, with dimension n. x ×n x Satisfying P = P T Furthermore, P is a symmetric positive definite matrix used to assign corresponding weights to different state variables, while ensuring the non-negativity and monotonicity of the function values.

[0116] The positive definiteness of a symmetric positive definite matrix P satisfies the following condition:

[0117]

[0118] This condition ensures that any non-zero state shift will cause a positive increase in V(x), thus possessing the physical property of an "energy function".

[0119] In one possible implementation, matrix P can be constructed in one of three ways:

[0120] Method 1: Manually set diagonal weights, with off-diagonal elements set to 0, which only reflects the independence of each state dimension;

[0121] Method 2: Construct a symmetric banded sparse matrix based on the spatial adjacency relationship between state variables;

[0122] Method 3: Based on the system closed-loop performance indicators, obtain them from the following optimization problem:

[0123]

[0124] Where: T is the sampling sequence length; λ is the regularization parameter, controlling the sparsity of the weights; ||P|| F The Frobenius norm is used to constrain the size of a matrix. P is a positive definite matrix.

[0125] Generally, the diagonal elements of matrix P are used to adjust the relative importance of different state dimensions (temperature, wind speed, humidity); the off-diagonal elements can be used to reflect the state linkage relationship between nodes.

[0126] In some embodiments, the state variables can be divided into multiple regions, such as the inlet air zone, the server front zone, the server rear zone, and the return air vent zone, and a submatrix P can be defined for each region. k The structure is as follows:

[0127]

[0128] Where: K is the total number of regions; x k P is the state sub-vector of the k-th region; k R is the local positive definite weighting matrix for the corresponding region; ij The weight matrix represents the coupling effects between different regions.

[0129] Such partitioned stability functions can enhance the system's ability to respond to "local anomalies" or "thermal islands".

[0130] Within each sampling period, the system calculates the stability function V(x(t)) in real time based on the acquired state vector x(t) and compares it with the reference steady-state function value V. ref For comparison, the following deviation is defined:

[0131] ΔV(t)=V(x(t))-V ref ;

[0132] Where: V ref The reference steady-state function value can be the historical average value during the system training period or obtained based on steady-state simulation calculation; ΔV(t) is the stability offset value of the current state, reflecting the degree of offset of the current state from the steady-state center.

[0133] If ΔV(t) > ε, the system will be judged as "deviation from steady state", where ε is a manually set margin threshold.

[0134] In some embodiments, the system employs a sliding window approach for multi-period detection to avoid misjudgments due to transient disturbances. The sliding average is defined as follows:

[0135]

[0136] Where: N is the length of the sliding window, which is usually set to 3-10; This represents the average stability function value over the past N periods;

[0137] like consistently higher than V ref +ε will trigger the controller to start executing the compensation command.

[0138] To further enhance the responsiveness to strong disturbances, this embodiment also introduces a disturbance margin optimization function, defined as follows:

[0139]

[0140] Where: M w (t) is the disturbance response margin index; ||w(t)||2 is the L2 norm of the current disturbance vector, representing the magnitude of the disturbance energy; if M w A significant increase in (t) indicates that the system is abnormally sensitive to disturbances, which may indicate model mismatch or insufficient adjustment. The modeling parameters or weight matrix should be adjusted in a timely manner.

[0141] S4. Based on the evolution trend of the stability function, a state feedback controller is designed. The current state vector is used to calculate the control commands for the cold aisle execution structure. These control commands include damper angle control signals and refrigeration unit power control signals. After completing system state modeling S2 and stability function construction S3, to ensure the cold aisle environment maintains stable operation under the combined effects of dynamic loads and external disturbances, it is necessary to design a controller structure with state feedback capabilities to achieve an effective closed-loop response from system state to execution commands. The controller's task is to output a set of control inputs u(t) based on the current state vector x(t) and the system state-space model, used to adjust the damper angle and refrigeration equipment power, thereby guiding the system towards a low-stability deviation state.

[0142] In this example, the controller design is based on a strategy of minimizing the future evolution trend of the stability function. Specifically, with the system state model as a constraint, the goal is to find the control input u(t) that minimizes the stability function V(x(t+1)) at the next time step. The optimization objective function is defined as follows:

[0143]

[0144] Combining the aforementioned state evolution model:

[0145] x(t+1)=Ax(t)+Bu(t);

[0146] Substituting the values, we obtain the expansion of the optimization objective:

[0147]

[0148] Expanding on this further:

[0149]

[0150] Where: x(t) is the current system state vector with dimension n. x ×1; u(t) is the control input vector with dimension n. u ×1; P is a symmetric positive definite matrix in the stability function, with dimension n. x ×n x A is the system state transition matrix, with dimension n. x ×n x B is the control input influence matrix, with dimension n. x ×n u P is a positive definite weight matrix (or Lyapunov matrix), with dimensions n×n, used to measure the cost of state deviation; A T B is the transpose of matrix A; T Let x(t) be the transpose of matrix B. T Let x(t) be the transpose of vector x(t), with dimension 1×n; u(t) T Let u(t) be the transpose of vector u(t), with dimension 1×m.

[0151] This optimization problem is a standard convex quadratic programming problem, and the objective function is a quadratic function with respect to u(t), which is easy to solve.

[0152] Taking the gradient of the objective function with respect to u(t) and setting the derivative to zero, we have:

[0153]

[0154] The optimal control law is obtained as follows:

[0155] u * (t)=-(B T PB) -1 B T PAx(t);

[0156] Where: u * (t) represents the optimal control input solution at the current time; (B) T PB) -1 For matrix B T The inverse of PB must be guaranteed to be a positive definite matrix (i.e., B). T PB>0) to ensure the existence of the control solution;

[0157] The overall control law can be expressed as a linear feedback structure, that is:

[0158] u(t) = Kx(t), where K = -(B T PB) -1 B T PA;

[0159] Matrix K is the feedback gain matrix with dimension n. u ×n x It directly maps the current state to control commands.

[0160] In this invention, the control input u(t) includes the following two types of physical signals:

[0161] Damper adjustment signal: Used to control the electric damper actuator at each critical air duct position. The signal form is PWM duty cycle command or 0-10V analog quantity. The typical control frequency is 1Hz, and the adjustment angle range is 0°-90°.

[0162] Cooling power control signal: used to adjust the air conditioner compressor frequency, fan speed or chilled water valve opening. The signal can be transmitted via RS485, Modbus and other protocols, with a response time of less than 2 seconds.

[0163] In some embodiments, the two control quantities mentioned above can be further divided into multiple sub-channels to achieve regional independent adjustment.

[0164] In one possible implementation, to enhance the system's tolerance to external disturbances w(t), the disturbance term can be directly introduced into the controller optimization structure to construct a robust control strategy. The system state model is extended as follows:

[0165] x(t+1)=Ax(t)+Bu(t)+w(t);

[0166] The control objective becomes:

[0167]

[0168] That is, the expected stability function value is minimized considering the influence of disturbances, where the disturbances satisfy... This method introduces a perturbation covariance matrix Σ while keeping the control solution structure unchanged. w It is used to adjust the amplitude of the control gain and enhance the adaptability to thermal burst interference.

[0169] To ensure the system operates within the physical constraints of the actuators, the control output must also meet the following constraints: u min ≤u(t)≤u max ;

[0170] Where: u minTo control the lower limit of input, including minimum damper opening and minimum cooling power; u max To control the upper limit of input, it is limited according to the physical specifications of the device;

[0171] If the optimal solution u * (t) If the value exceeds the above range, it can be corrected to the nearest feasible value through projection. The specific method is as follows: u clip (t)=min(max(u * (t),u min ),u max );

[0172] Among them, u clip (t) represents the actual control command value after the amplitude limit is issued to the actuator, which has been constrained to the allowable range.

[0173] In addition, in certain special cases (such as multiple devices sharing control bandwidth), soft constraints or penalty functions can be added to extend the problem into a constrained optimization problem.

[0174] S5. The control command is synchronously output to the damper actuator and cooling equipment in the cold aisle to adjust the airflow and temperature distribution inside the cold aisle and update the state vector to enter the next cycle, so as to form a closed-loop control process for continuous monitoring and adjustment.

[0175] Based on the aforementioned state modeling (S2), stability function construction (S3), and feedback controller design (S4) processes, the state assessment and control command calculation are completed. In step S5, the control commands generated by the controller are applied to the actual execution object to achieve physical adjustment of the cold aisle heat-flow state. To construct a continuous, high-response, and low-latency control closed loop, the system state needs to be re-acquired after execution to update the state vector and enter the next control cycle, thus forming a complete closed-loop control process.

[0176] In this embodiment, the control input vector u(t) output by the controller includes two types: damper execution signal and refrigeration control signal, and its structure is defined as follows:

[0177]

[0178] Where u(t) is the total control input vector with dimension n. u ×1; u f (t) is the damper angle adjustment signal vector, with dimension n. f ×1, corresponding to each deflector; u c (t) is the cooling power adjustment signal vector with dimension n. c ×1, corresponding to each cooling module; n u =n f +n c This represents the sum of the dimensions of the control inputs.

[0179] Generally, control signals are sent to the execution port number via serial communication (such as RS485), industrial Ethernet, or wireless protocols; the instruction structure includes the device address, adjustment target value, execution limit parameters, acknowledgment bit, etc.

[0180] In one possible implementation, the controller internally maintains a set of actuator mapping tables:

[0181]

[0182] Where: addr i The physical address of the i-th executor; type i ∈{damper, refrigeration} indicates the actuator type; This indicates the scope of its control.

[0183] This mapping structure is used to quickly route and distribute control signals, improving the real-time performance of the distribution.

[0184] The damper control signal u of the present invention f (t) corresponds to the baffle structure acting on the top or bottom of the cold aisle in the computer room, and its function is to adjust the airflow direction and local air volume distribution.

[0185] Specifically, in this embodiment:

[0186] When the temperature T at monitoring node i i (t) Exceeds the preset cooling target At that time, the controller determined that the location was a localized thermal anomaly;

[0187] The controller will identify the air guide plate j above the corresponding area and adjust the damper angle as follows:

[0188]

[0189] Where: θ j (t) represents the opening angle of the j-th guide vane at time t; δ θ To control the increment step size; sign(·) represents the sign function, used to determine the adjustment direction.

[0190] To avoid disturbance transfer caused by single-point adjustment, the system supports adjustment based on local temperature gradient vectors. Achieve coordinated damper regulation:

[0191]

[0192] The controller constructs a guiding path based on the local high temperature gradient area and links multiple sets of damper angles to form a directional cooling air corridor.

[0193] Cooling power command u output by the controller c (t) can be further decomposed into the following three control channels:

[0194]

[0195] Where: f c (t) represents the compressor frequency adjustment command, in Hz; q c (t) represents the cooling water flow rate adjustment, in L / min; v c (t) represents the fan's airflow speed adjustment value, in m / s.

[0196] As an alternative, to improve the feedforward response, the system employs a predicted temperature sequence. Adjust the cooling power curve in advance:

[0197]

[0198] Among them, P i (t) represents the control priority cycle index value of the i-th monitoring area at the current time t; This indicates the variable k corresponding to the maximum value, that is, the k value at which the objective function achieves its maximum value; This represents the air temperature value of the i-th region predicted at time t+k; N is the length of the prediction time window (number of prediction steps).

[0199] And set the target value for the current cooling power as follows:

[0200]

[0201] in: Let be the target output power of the i-th refrigeration unit at time t; η be the refrigeration regulation coefficient; k * The period corresponding to the maximum predicted temperature rise; P base This is the static maintenance power for the equipment.

[0202] The power of multiple cooling modules will be sorted according to the magnitude of regional temperature deviation, and graded targets will be applied to achieve both energy saving and improved cooling capacity in key areas.

[0203] Once completed, the system will immediately re-collect temperature, wind speed, and humidity data from all key locations to construct a new state vector:

[0204] x(t+1)=[T1(t+1),V1(t+1),H1(t+1),…,T n (t+1),V n (t+1),H n (t+1)] T ;

[0205] Where x(t+1) represents the state vector of the cold aisle system at time t+1, with a dimension of 3n×1, where n represents the total number of monitoring nodes; T i (t+1) represents the air temperature at the i-th monitoring location at time t+1, in degrees Celsius (°C); V i (t+1) represents the airflow velocity at the i-th monitoring location at time t+1, in meters per second (m / s); H i (t+1) represents the relative humidity at the i-th monitoring location at time t+1, in percentage (%); i∈{1,2,…,n} represents the monitoring location number, which is usually the index of a physical or virtual measuring point deployed in the cold aisle.

[0206] To avoid amplifying state noise due to execution delays or data residuals, the system introduces a fusion-based state update mechanism, defined as follows:

[0207]

[0208] in: This serves as the input for the merged state, used for calculation in the next cycle. This represents the predicted state of the system; α is the weighting coefficient between real-time observation and prediction, typically ranging from 0.6 to 0.9. This represents the predicted state vector.

[0209] This fusion approach can balance the practicality of execution feedback with the consistency of modeling structure, thereby improving the system's fault tolerance to sudden errors and data delays.

[0210] The cold aisle airtight management device described below and the cold aisle airtight management method described above can be referred to in correspondence.

[0211] Please see Figure 2 A cold aisle airtight management device for computer rooms, comprising:

[0212] The environmental monitoring module is used to deploy monitoring nodes at multiple key locations inside the cold aisle to collect real-time information on air temperature, airflow velocity, and humidity at each location. The collected data is standardized and structured through a unified data acquisition mechanism, thereby constructing a state vector covering the entire cold aisle area, providing a comprehensive data foundation for subsequent state modeling and control.

[0213] The state modeling module receives the state vectors generated by the environmental monitoring module and establishes a state-space model of the dynamic evolution of the heat-flow field based on them. This model can describe the heat propagation and airflow coupling laws between monitoring locations. It supports modeling and expressing the system response trend, providing a basis for stability assessment and control strategy reasoning.

[0214] The stability assessment module is used to calculate the stability index of the current cold aisle heat-flow system in real time based on the current system state output by the state modeling module; by comparing the state change trend with the set steady-state benchmark, it determines whether there is deviation, fluctuation or local instability in the system; the assessment result serves as the criterion signal for triggering the control strategy and is fed back to the control module for control signal generation.

[0215] The state feedback control module is used to generate corresponding control strategies based on the stability index output by the stability assessment module and the current state vector. The control strategies include angle adjustment signals for multiple dampers and power output adjustment signals for multiple refrigeration devices. It has linkage logic and priority mechanism to support adaptive adjustment for local thermal anomalies, changes in overall airflow distribution, and other situations.

[0216] The execution linkage module is used to output the damper and cooling control signals generated by the control module to each actuator. The actuators include electric guide vanes, compressors, chilled water pumps, and blower fans. By executing adjustment commands, the air volume distribution and cooling intensity in local areas within the cold aisle are changed, thereby achieving linkage optimization control of air velocity and temperature field. At the same time, status updates and data feedback are completed, forming a closed-loop regulation mechanism.

[0217] The system in this embodiment can be used to execute the above method embodiments, and its principle and technical effect are similar, so they will not be described again here.

[0218] Please see the appendix Figure 3 The present invention also provides a terminal device, including: a processor and a memory, wherein the memory stores a computer program executable by the processor, and the computer program performs the method described above when executed by the processor.

[0219] The present invention also provides a storage medium storing a computer program, which is executed by a processor to perform the method described above.

[0220] The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0221] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for closed-loop management of cold aisles in computer rooms, characterized in that, Includes the following steps: S1. Deploy environmental monitoring nodes at key locations within the cold aisle area to collect air temperature and airflow velocity at key locations in real time, and integrate the environmental parameters at each key location to construct a cold aisle state vector. S2. Input the state vector into the thermal-fluid coupled state space modeling module to establish a state space model of the temperature and airflow velocity inside the cold aisle changing with time. The model is used to dynamically describe the state evolution process of the system under multi-source disturbance conditions. S3. Based on the state-space model, construct a stability function to evaluate the stability of the cold aisle thermal-fluid field. The stability function is represented by a quadratic function composed of a state vector and a symmetric positive definite matrix, and is used to monitor the stability deviation of the system under the influence of the current disturbance in real time. S4. Based on the evolution trend of the stability function, design a state feedback controller and use the current state vector to calculate the control instructions of the cold aisle execution structure. The control instructions include the damper angle control signal and the refrigeration unit power control signal. S5. The control command is synchronously output to the damper actuator and cooling equipment in the cold aisle to adjust the airflow and temperature distribution inside the cold aisle, and the state vector is updated to enter the next cycle, so as to form a closed-loop control process for continuous monitoring and adjustment.

2. The method for closed management of cold aisle in a computer room according to claim 1, characterized in that, The key locations include the air ducts at the cold aisle inlets, the front of the server racks, the rear of the server racks, and the cold aisle outlets.

3. The method for closed management of cold aisle in a computer room according to claim 1, characterized in that, The environmental monitoring nodes include: Temperature sensors are used to detect air temperature at critical locations; Wind speed sensor, used to detect airflow speed at key locations; Humidity sensor, used to monitor air humidity; The data acquisition module is used to integrate measurement data from various sensors and upload them to the host computer.

4. The method for closed management of cold aisle in a computer room according to claim 1, characterized in that, The state-space model is established in the following form: x(t+1)=Ax(t)+Bu(t)+w(t); Where x(t) represents the state vector, u(t) represents the control command vector, w(t) represents the disturbance term, and A and B are system matrices.

5. The method for closed management of cold aisle in a computer room according to claim 1, characterized in that, The stability function is a quadratic function of the following form: V(x)=x T Px; Where x is the current state vector, x T Let x be the transpose of x, and P be a symmetric positive definite matrix used to measure the deviation of the system's energy level and stability.

6. The method for closed management of cold aisle in a computer room according to claim 1, characterized in that, The state feedback controller, based on the real-time changing trend of the stability function, constructs the following optimization objective by minimizing the system deviation: Where u(t) is the control input vector at time t, x(t+1) represents the system state vector at time t+1, and P is a symmetric positive definite matrix. T Px(t+1) represents the stability function, which is a quadratic function, and V(x(t+1)) represents the stability function value of the system at the next time step. Based on the system model x(t+1)=Ax(t)+Bu(t), the optimal control input u(t) is calculated, wherein the control input includes: The damper angle adjustment signal based on the disturbance response is used to correct the cold airflow path. Cooling power adjustment signals based on heat load fluctuations are used to combat abnormal temperatures. The feedback controller is implemented using a linear quadratic regulator method or its improved algorithm, and has the ability to adaptively adjust the gain parameter to meet the control accuracy requirements under different dynamic changes in server load.

7. The method for closed management of cold aisle in a computer room according to claim 1, characterized in that, The damper angle control signal drives adjustable guide vanes installed at the top or bottom of the cold aisle, thereby adjusting the spatial distribution of airflow direction and flow rate through the linkage of multiple air-guiding actuators. When the monitoring node detects that the temperature at the front end of a rack exceeds the set threshold, the controller will prioritize adjusting the opening angle of the air guide plate above that area to increase the supply of cold air to that area. The opening adjustment of multiple air guide vanes is optimized in a coordinated manner based on the changing trend of the local temperature gradient in the state vector, thereby forming a directional cooling path for the hot spot area.

8. The method for closed management of cold aisle in a computer room according to claim 1, characterized in that, The power control signal of the refrigeration unit is used to dynamically control the cooling capacity of the refrigeration module in the cooling system, and its control logic includes: Based on the state-space model, the temperature trend of multiple future cycles is predicted, and the output power of the refrigeration unit is adjusted in advance to avoid hysteresis. In non-uniform heat load scenarios, cooling priority levels are divided according to the distribution characteristics of the state vector, and a hierarchical control strategy is adopted to apply different target powers to each cooling module in order to achieve both local enhanced cooling and global energy efficiency optimization. The control signals also include composite adjustment parameters for compressor frequency, cooling water flow rate, or air supply speed, used to improve the overall cooling response bandwidth.

9. A cold aisle airtight management device for a computer room, applied to the cold aisle airtight management method for a computer room as described in any one of claims 1-8, characterized in that, include: The environmental monitoring module is used to collect air temperature, airflow velocity and humidity information at multiple key locations within the cold aisle and generate state vectors. The state modeling module is used to construct a state-space model of the cold aisle thermal-fluid field based on state vectors; The stability assessment module is used to calculate the stability function and assess the degree of stability deviation in the cold aisle. The state feedback control module is used to calculate the damper angle and the refrigeration unit power control signal based on the stability function; The linkage module is used to output control signals to the damper actuator and cooling equipment to adjust the airflow and temperature distribution in the cold aisle.

10. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements a method for closed management of cold aisle in a computer room as described in any one of claims 1-8.

Images