An intelligent management method and system for server power supply

Abstract

The present invention relates to the technical field of server power management. More specifically, it relates to an intelligent server power management method and system, including: obtaining the historical load data sequence and current load data of the server; performing topological dynamics load prediction based on the historical load data sequence to obtain a predicted load sequence; optimizing the predicted load sequence through group theory transformation to obtain an optimized predicted load sequence; generating a control strategy using a chaos control strategy based on the optimized predicted load sequence; applying quantum probability field optimization to the control strategy to obtain an optimized control strategy; fine-tuning the optimized control strategy through non-commutative geometry to obtain a final control strategy; outputting the final control strategy for regulating the server power. The significant improvement in prediction accuracy enables the system to optimize resource allocation in advance, effectively reducing energy waste and enhancing the stability and reliability of the system.

Description

Technical Field

[0001] The present invention relates to the technical field of server power management, and more specifically, to an intelligent server power management method and system thereof. Background Art

[0002] With the continuous expansion of the scale of data centers and the increasing growth of service demands, server power management has become a key issue that urgently needs to be solved. Efficient power management not only concerns the operating costs of data centers, but also directly affects their reliability, performance, and environmental friendliness. However, the server power management methods commonly used in the industry at present still have many deficiencies and are difficult to meet the complex requirements of modern data centers.

[0003] Traditional server power management methods mainly rely on simple threshold control or rule-based strategies. Although these methods are easy to implement, they lack the ability to predict the changes in server loads prospectively, often resulting in power management lagging behind actual demands. In addition, these methods usually adopt fixed control parameters and are difficult to adapt to the dynamically changing load environment, easily causing energy waste or a decline in service quality.

[0004] In recent years, some researchers have attempted to introduce machine learning technologies into the field of server power management. For example, using neural networks for load prediction or adopting reinforcement learning algorithms to generate control strategies. These methods have improved the level of power management intelligence to a certain extent, but there are still some inherent limitations. First of all, these methods usually require a large amount of training data and computing resources and may face challenges in terms of efficiency and real-time performance in actual deployment. Secondly, they are often difficult to handle complex non-linear load patterns and emergencies, resulting in insufficient prediction accuracy and control stability. Finally, these methods lack in-depth consideration of the overall dynamic characteristics of the server system and are difficult to achieve a truly optimized power management strategy.

[0005] In addition, the existing server power management methods generally have problems of slow response speed and poor adaptability. In the face of sudden load changes or abnormal situations, these methods often cannot make effective adjustments in a timely manner, which may lead to service interruptions or energy waste. At the same time, the existing methods also seem powerless when dealing with the collaborative management of multiple servers and are difficult to achieve a globally optimal resource allocation.

[0006] In view of the above problems, there is an urgent need for a new type of intelligent server power management method that can overcome the limitations of the existing technology and provide a more accurate, efficient, and flexible power management solution. Summary of the Invention

[0007] The present invention aims to solve the technical problems existing in the existing server power management methods, such as low prediction accuracy, low efficiency of control strategies, slow system response speed, and poor adaptability. By innovatively combining multi-disciplinary frontier theories, the present invention proposes an intelligent server power management method and system, which can significantly improve the performance and efficiency of power management.

[0008] The present invention provides an intelligent server power management method, including:

[0009] An acquisition step, including:

[0010] Acquire the historical load data sequence and the current load data of the server;

[0011] A processing step, including:

[0012] Based on the historical load data sequence, perform topological dynamics load prediction to obtain a predicted load sequence;

[0013] According to the predicted load sequence, optimize it through group theory transformation to obtain an optimized predicted load sequence;

[0014] Based on the optimized predicted load sequence, generate a control strategy using a chaos control strategy;

[0015] Apply quantum probability field optimization to the control strategy to obtain an optimized control strategy;

[0016] Fine-tune the optimized control strategy through non-commutative geometry to obtain a final control strategy;

[0017] An output step, including:

[0018] Output the final control strategy for regulating the server power.

[0019] Preferably, the performing of the topological dynamics load prediction specifically includes:

[0020] Map the historical load data sequence to a high-dimensional topological space to obtain a sequence of mapped data points;

[0021] Based on the sequence of mapped data points, construct a dynamic system;

[0022] Utilize the Takens embedding theorem to reconstruct the phase space and estimate the vector field through the local linear embedding algorithm;

[0023] According to the vector field, predict the future load value;

[0024] Obtain the predicted load sequence through inverse mapping.

[0025] Preferably, the dynamic system is defined by the following differential equation:

[0026] ,

[0027] wherein, represents the state of the system on the m-dimensional manifold , is a vector field, is a vector field function, is an m-dimensional manifold, is 's tangent space. For each , is at the point a vector in the tangent space, represents time, is the derivative of y with respect to time t.

[0028] Preferably, the optimization by group theory transformation specifically includes:

[0029] Define a transformation group , where each is a transformation operation;

[0030] Apply the group action to each predicted value in the predicted load sequence to obtain the optimized predicted values;

[0031] Form the optimized predicted load sequence from all the optimized predicted values.

[0032] Preferably, the generation of the control strategy using the chaos control strategy specifically includes:

[0033] Construct a parameterized chaos function;

[0034] Embed the optimized predicted load sequence into the chaos system to obtain a control parameter sequence;

[0035] Generate a control strategy through a decision function based on the chaotic state and the control parameters.

[0036] Preferably, the chaos function is defined as:

[0037] ,

[0038] wherein, is the chaotic state at time t+1, is the parameterized chaos function, is the chaotic state at time t, is the control parameter.

[0039] Preferably, the quantum probability field optimization specifically includes:

[0040] Define the quantum state , where is the quantum state, is the complex amplitude, is the ground state, is the number of ground states;

[0041] Map the control strategy to the quantum state;

[0042] Obtain the optimized control strategy through quantum measurement operations.

[0043] Preferably, the non-commutative geometry fine-tuning specifically includes:

[0044] Define the non-commutative manifold , the points on which are represented by non-commutative coordinates;

[0045] Map the optimized control strategy to the non-commutative space;

[0046] Apply the non-commutative gradient operator and the action functional to obtain the final control strategy.

[0047] Preferably, the final control strategy is given by the following equation:

[0048] ,

[0049] where is the final control strategy, is the mapping function, is the point in the non-commutative space, is the non-commutative gradient operator, is the action functional.

[0050] The server power intelligent management system that executes the above method includes:

[0051] An acquisition module for acquiring the historical load data sequence and the current load data of the server;

[0052] A processing module for:

[0053] Based on the historical load data sequence, perform topological dynamics load prediction to obtain a predicted load sequence;

[0054] According to the predicted load sequence, optimize it through group theory transformation to obtain an optimized predicted load sequence;

[0055] Based on the optimized predicted load sequence, generate a control strategy using a chaos control strategy;

[0056] Apply quantum probability field optimization to the control strategy to obtain an optimized control strategy;

[0057] Fine-tune the optimized control strategy through non-commutative geometry to obtain the final control strategy;

[0058] An output module, configured to output the final control strategy for regulating the server power supply.

[0059] Specifically, the beneficial effects of the present invention are mainly reflected in the following aspects:

[0060] The method of the present invention realizes high-precision prediction and intelligent control of server loads. This multi-dimensional and multi-level method not only overcomes the limitations of traditional methods but also fully utilizes the advantages of various theoretical tools to form a synergistic overall solution.

[0061] Specifically, in terms of load prediction, the method of the present invention can better capture the non-linear features and potential structures in load data through the combination of topological dynamics and group theory transformation, significantly improving the accuracy and robustness of prediction. In terms of control strategy generation, the introduction of chaos control theory enables the system to more flexibly respond to complex and changing load environments, while quantum probability field optimization expands the strategy search space, contributing to the discovery of better control schemes. The application of non-commutative geometry fine-tuning further enhances the system's response speed and refined regulation ability, enabling power management to more quickly and accurately adapt to load changes.

[0062] This innovative method brings significant effects in many aspects. First, the substantial improvement in prediction accuracy enables the system to optimize resource allocation in advance, effectively reducing energy waste. Second, the efficient control strategy generation mechanism not only improves energy utilization efficiency but also enhances the stability and reliability of the system. The fast response ability and strong adaptability ensure that the system can calmly handle various load conditions, including emergencies and abnormal loads.

[0063] In addition, the method of the present invention also demonstrates unique advantages in dealing with multi-server collaborative management. From a global optimization perspective, it can better balance the load and resource allocation among different servers, maximizing the overall benefits. This not only improves the overall energy efficiency of the data center but also enhances the service quality and system reliability.

[0064] Generally speaking, the server power intelligent management method and system proposed by the present invention have successfully solved the key problems in the prior art by innovatively integrating multidisciplinary theories. It has not only achieved significant improvements in various performance indicators, but more importantly, provided a comprehensive, balanced and efficient power management solution. The application of this method will bring substantial improvements to the operation of data centers, including reduction of energy costs, improvement of service quality, enhancement of system reliability and improvement of environmental friendliness. In the context of the continuous expansion of the scale of current data centers and the increasing energy consumption, the method of the present invention has important technical value and broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a flowchart of the method of the present invention.

[0066] Figure 2 It is a flowchart of the operation of the processing module of the present invention.

[0067] Figure 3 It is a logical block diagram of the system of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0068] Please refer to Figures 1 - 3 , first, according to one aspect of the present invention, a server power intelligent management method is provided. The purpose of this method is to achieve intelligent management of power by predicting the server load and generating optimized control strategies, thereby improving the energy efficiency of the data center.

[0069] The method of the present invention first obtains the historical load data sequence and the current load data of the server. These data are the basis for subsequent processing steps and reflect the load conditions of the server at different time points. Preferably, the historical load data sequence can include data for the past week, month or longer to capture the periodic changes and long-term trends of the load.

[0070] After obtaining the data, this method enters the processing stage. First, based on the historical load data sequence, topological dynamics load prediction is performed to obtain the predicted load sequence. This step utilizes advanced concepts of topology and dynamic system theory, maps the load data to a high-dimensional topological space, and makes predictions in this space.

[0071] Specifically, the process of topological dynamics load prediction is as follows:

[0072] First, define the historical load data sequence , where represents the load value at the th time point. Then, introduce the topological mapping function , where, is a n-dimensional manifold. Through this mapping, a sequence of data points after mapping is obtained:

[0073] ,

[0074] In this high-dimensional topological space, a dynamical system is constructed, which is represented by the following differential equation:

[0075] ,

[0076] where, represents the state of the system on the n-dimensional manifold , is the vector field, is the vector field function, is the m-dimensional manifold, is the tangent space of. For each , is a vector in the tangent space of at the point , represents time, is the derivative of y with respect to time t. This dynamical system describes the evolution law of the load in the high-dimensional space.

[0077] To reconstruct the phase space and estimate the vector field , this method adopts the Takens embedding theorem and the locally linear embedding (LLE) algorithm.

[0078] This combination takes advantage of non-linear dynamics and machine learning, and can effectively capture the dynamic characteristics of complex systems.

[0079] The equation for predicting future load values is:

[0080] ,

[0081] Finally, through the inverse mapping the predicted load values are obtained:

[0082] ,

[0083] is the representation of the predicted load on the high-dimensional manifold at the future time . is the predicted load value at the future time obtained through the inverse mapping .

[0084] The advantage of this topological dynamics method is that it can capture the non - linear features and high - dimensional structures in the load data, thus providing more accurate prediction results. In practical applications, an appropriate manifold dimension can be selected according to the specific situation of the server. . For example, for a server load with obvious daily and weekly cycles, one can choose , that is, use the hourly data of one week as the embedding dimension.

[0085] In server power management, the accuracy of load prediction is directly related to the improvement of energy efficiency. Traditional linear prediction methods often fail to capture the non - linear changes in the load, especially when facing complex periodic and random fluctuations. By introducing the topological dynamics method, these non - linear features can be better handled, thus improving the prediction accuracy.

[0086] For example, in a data center, the load of a server usually has obvious daily and weekly cycle characteristics and is also affected by burst traffic. By choosing an appropriate manifold dimension m, the hourly load data of one week can be embedded into a high - dimensional space to capture the periodicity and long - term trends of the load. This high - dimensional representation can more accurately reflect the change law of the load, helping the system make adjustments in advance to avoid power waste or overload situations.

[0087] After obtaining the predicted load sequence, this method is further optimized through group - theory transformation. This step introduces the concept of group theory and optimizes the prediction results by defining a set of transformation operations.

[0088] Historical load data can be collected in real - time through the server's monitoring system, recording metrics such as CPU usage, memory occupancy, network traffic, etc. at each time point. These data can be stored at the minute or hour level to form a time series.

[0089] First, pre - process the historical load data to remove outliers and noise. Then, according to the Takens embedding theorem, select an appropriate delay time and embedding dimension to map the one - dimensional time series onto a high - dimensional manifold. Next, use the LLE algorithm to reconstruct the phase space, estimate the vector field F, and based on this, conduct future load prediction.

[0090] Specifically, define a transformation group , where each is a transformation operation. The group action is defined as: .

[0091] For each predicted value , calculate:

[0092] ,

[0093] Among them, is the actual observed value. The output of this step is the optimized prediction sequence . The introduction of group theory transformation optimization enhances the robustness and adaptability of the prediction model. By defining an appropriate transformation group, various types of prediction errors can be handled, such as scale errors, offset errors, etc. For example, a linear transformation group can be defined, where and are real parameters. In practice, an appropriate transformation group can be selected according to the characteristics of historical data to achieve the best optimization effect.

[0094] In server power management, the results of load prediction may be affected by various factors, resulting in prediction errors. For example, due to the strong randomness and uncertainty of the server's workload, the prediction model may overfit or underfit. By introducing group theory transformation optimization, the prediction results can be adjusted under different transformations to ensure that the predicted values are closer to the actual load. For example, assume that the load of a certain server shows large fluctuations during certain time periods, and the prediction model fails to accurately capture these fluctuations. By defining a set of linear transformations , the predicted values can be scaled and translated to better match the changes in the actual load. By minimizing the error , the optimal transformation operation can be selected to make the prediction results more accurate.

[0095] In addition to historical load data, the actual load observation values also need to be recorded to calculate the prediction error. After each prediction, apply the group theory transformation optimization step to apply different transformation operations to the predicted value , and calculate the error under each transformation. Finally, select the transformation operation with the minimum error to obtain the optimized predicted value .

[0096] Through these steps, the method of the present invention can generate a high-quality load prediction sequence, providing a reliable basis for the generation of subsequent control strategies. This multi-level prediction and optimization method significantly improves the accuracy and reliability of the prediction, thus providing solid data support for the intelligent management of server power.

[0097] The method of the present invention is not limited to the above steps, and also includes generating a control strategy based on the optimized predicted load sequence, and further optimizing and fine-tuning the control strategy.

[0098] Generally speaking, the method provided by the present invention provides a novel and efficient solution for the intelligent management of server power by combining advanced mathematical theories and actual application requirements.

[0099] In a preferred embodiment of the present invention, based on the optimized predicted load sequence, the method further generates a control strategy using a chaos control strategy. This step introduces the concept of chaos theory and aims to handle the non-linearity and unpredictability that may exist in server loads.

[0100] Specifically, the method first constructs a parameterized chaos function. Preferably, the logistic map can be selected as the basis, and its definition is as follows:

[0101] ,

[0102] where is the chaos state at time t, and is the control parameter. In practical applications, the value range of usually lies between 3.57 and 4, within which the system exhibits obvious chaotic characteristics. Next, the method embeds the optimized predicted load sequence into the chaos system. This is achieved by defining a mapping function as follows:

[0103] ,

[0104] where is the optimized predicted load value. The choice of the mapping function has an important impact on the system performance. In an embodiment of the present invention, a linear mapping can be adopted:

[0105] ,

[0106] where and are parameters determined according to historical data. For example, it can be set that , , where and are respectively the maximum and minimum values of the historical load data. Based on the chaos state and the control parameter, the method generates a control strategy through a decision function. The decision function converts the chaos state and the control parameter into specific control actions:

[0107] ,

[0108] In practical applications, the decision function can be designed according to specific server power management requirements. For example, the control action can be defined as the adjustment amount of the server power supply voltage or current.

[0109] In server power management, the changes in load are often non-linear and unpredictable, especially when facing sudden traffic or complex application scenarios. Traditional control strategies may not be able to effectively handle these uncertainties, resulting in energy waste or performance degradation. By introducing chaos control strategies, more intelligent and adaptive control actions can be generated in complex load environments, ensuring that the server can operate efficiently under different working conditions.

[0110] For example, assume that the load of a certain server shows violent fluctuations during certain time periods. Traditional control strategies may frequently adjust the power supply, leading to energy waste. By embedding the load prediction results into a chaos system, the control parameters can be dynamically adjusted according to the current load situation , generating a more refined control strategy. This can not only avoid unnecessary adjustments but also better cope with sudden changes in load, improving the overall performance of the system.

[0111] It is necessary to obtain the optimized predicted load value , and calculate the corresponding control parameters through the mapping function . After each prediction, the optimized load value is input into the chaos system to generate a chaotic state . Then, through the decision function , the chaotic state is converted into specific control actions , such as adjusting the power supply voltage or current of the server. , such as adjusting the power supply voltage or current of the server.

[0112] After generating the initial control strategy, the method of the present invention further introduces quantum probability field optimization to improve the quality of the strategy. This step draws on the concepts of quantum mechanics and provides a new perspective for the optimization of control strategies.

[0113] First, define a quantum state:

[0114] ,

[0115] where is the ground state, is the complex amplitude, satisfying the normalization condition . In practical applications, can be set to a value that matches the dimension of the control strategy. For example, if the control strategy contains 10 discrete actions, can be set.

[0116] Next, this method maps the control strategy to the quantum state:

[0117] ,

[0118] Among them, is a unitary transformation dependent on the control strategy. In actual implementation, quantum gates such as rotation gates or phase gates can be selected to construct . For example, it can be defined as:

[0119] ,

[0120] Among them, is a Hermitian matrix representing the Hamiltonian of the system.

[0121] The optimized control strategy is given by the quantum measurement operation: ,

[0122] Among them, is a quantum measurement operation. In practical applications, an appropriate measurement basis can be selected to achieve the desired optimization effect. For example, the computational basis measurement can be used, and the measurement results can be mapped back to the control strategy space.

[0123] By introducing quantum probability field optimization, the method of the present invention can search for the optimal solution in a larger strategy space and potentially find high-quality control strategies that are difficult to discover by traditional methods.

[0124] In server power management, the optimization of control strategies is a complex multi-objective problem involving multiple variables and constraints. Traditional optimization methods may only find local optimal solutions and cannot achieve global optimality. By introducing quantum probability field optimization, the optimal solution can be searched in a larger strategy space, and more efficient control strategies can be found.

[0125] For example, assume that the power management of a certain server needs to be adjusted in multiple dimensions, such as supply voltage, current, fan speed, etc. Traditional optimization methods may only be able to adjust within a small range, resulting in limited performance improvement. By mapping the control strategy to the quantum state space, optimization can be carried out simultaneously in multiple dimensions to find a more global optimal solution. This can not only improve energy efficiency but also extend the service life of the server.

[0126] It is necessary to obtain the initial control strategy , and map it into the quantum state space. Update the quantum state through the unitary transformation , and perform quantum measurement after each iteration to extract the optimized control strategy . Repeat this process until a satisfactory optimization result is found.

[0127] Finally, the method of the present invention uses non-commutative geometry fine-tuning to further optimize the control strategy. This step introduces the concept of non-commutative geometry, providing a theoretical basis for the fine-tuning of the control strategy.

[0128] Specifically, a non-commutative manifold is defined , and the points on it are represented by non-commutative coordinates , satisfying: ,

[0129] where [,] represents the commutator, is the non-commutative parameter. In practical applications, the appropriate non-commutative parameter can be selected according to the characteristics of the system. For example, for a system with two main control variables, can be set, and is selected, where is a small positive real number, such as 0.01. This method maps the optimized control strategy to this non-commutative space:

[0130] ,

[0131] where is a mapping function. In practical implementation, an appropriate mapping function can be selected to maintain the main characteristics of the control strategy.

[0132] The final control strategy is given by the following equation:

[0133] ,

[0134] where is the non-commutative gradient operator, is an action functional defined on the non-commutative space. The introduction of the non-commutative gradient allows for finer adjustments in the policy space, potentially finding better strategies near the local optimal solution.

[0135] In server power management, the fine-tuning of the control strategy is an important link, especially in the face of complex load environments. Traditional fine-tuning methods may only be able to make adjustments within a small range, resulting in an unstable control strategy. By introducing non-commutative geometry fine-tuning, finer adjustments can be made in the control strategy space, ensuring that the system can maintain optimal performance under different load conditions.

[0136] For example, assume that the load of a certain server exhibits significant fluctuations during certain time periods. Traditional fine-tuning methods may lead to frequent adjustments of the control strategy, affecting the stability of the system. By introducing non-commutative geometry fine-tuning, more refined adjustments can be made in the control strategy space, avoiding unnecessary oscillations and ensuring that the system maintains stable performance under different load conditions. In addition, the introduction of non-commutative gradients can also help the system converge to the optimal solution faster, improving the overall control effect.

[0137] It is necessary to obtain the optimized control strategy , and map it into the non-commutative space. Through the non-commutative gradient operator calculate the gradient of the control strategy in the non-commutative space, and make adjustments according to the action functional . Finally, through the inverse mapping map the optimized control strategy back to the original space to obtain the final control strategy .

[0138] Through this series of optimization steps, the method of the present invention can generate a high-quality and robust server power control strategy. This multi-level and multi-angle optimization method fully considers the complexity and uncertainty of the server load, providing strong support for realizing intelligent and efficient power management.

[0139] In practical applications, the parameters and function selections in each step can be adjusted according to the specific server environment and management requirements. For example, an appropriate chaotic mapping function can be selected according to the characteristics of historical data, or the non-commutative parameters can be adjusted according to the physical constraints of the system. This flexibility enables this method to adapt to various different server power management scenarios and has broad application prospects.

[0140] The present invention also provides a server power intelligent management system corresponding to the above method. The design of this system aims to implement each step described in the method, thereby realizing efficient and intelligent server power management.

[0141] The system mainly includes an acquisition module 1, a processing module 2, and an output module 3. These modules work together to complete the whole process from data acquisition to the output of the final control strategy.

[0142] The acquisition module 1 is responsible for obtaining the historical load data sequence and the current load data of the server. This module can collect data through various methods, such as directly reading from the server's monitoring system or collecting data in real time through deployed sensors. Preferably, the acquisition module 1 can set the data acquisition frequency and time span to ensure sufficient historical data for subsequent analysis. For example, data can be collected once per minute and the historical data for the most recent month can be saved. Such settings can capture short-term fluctuations and long-term trends in the server load.

[0143] The processing module 2 is the core of the system and is responsible for performing a series of complex data processing and policy generation steps. This module can be further divided into several sub-modules, with each sub-module corresponding to a main step in the method.

[0144] First, the prediction sub-module 21 in the processing module 2 performs topological dynamics load prediction based on the historical load data sequence to obtain a predicted load sequence. This sub-module implements steps such as topological mapping, dynamic system construction, and prediction calculation described in the method. In actual implementation, the prediction sub-module 21 can use high-performance computing units, such as GPUs, to accelerate complex mathematical operations.

[0145] Next, the optimization sub-module 22 uses the group theory transformation method to optimize the predicted load sequence. This sub-module implements the definition and application of the transformation group, and improves the accuracy of the prediction results through a series of transformation operations. Preferably, the optimization sub-module 22 can dynamically adjust the parameters of the transformation group to adapt to different prediction scenarios.

[0146] The policy generation sub-module 23 generates a preliminary control policy based on the optimized predicted load sequence using chaos control theory. This sub-module implements steps such as chaos mapping, parameter adjustment, and decision functions. In practical applications, the policy generation sub-module 23 can include a parameter library that stores the optimal parameter settings for different load conditions to quickly respond to various load changes.

[0147] The quantum optimization sub-module 24 applies quantum probability field optimization to the preliminarily generated control policy. This sub-module implements operations such as quantum state construction, mapping, and measurement. Considering the complexity of quantum computing, the quantum optimization sub-module 24 can use a quantum simulator to simulate the behavior of the quantum system, thereby implementing the quantum optimization process on a classical computer.

[0148] Finally, the fine-tuning sub-module 25 finely adjusts the optimized control policy through non-commutative geometry methods to obtain the final control policy. This sub-module implements steps such as the definition of non-commutative manifolds, mapping, and gradient calculation. To improve computational efficiency, the fine-tuning sub-module 25 can use parallel computing technology to process multiple policy adjustment schemes simultaneously.

[0149] The output module 3 is responsible for outputting the final control strategy generated by the processing module 2 for regulating the server power supply. This module is not just simple data transmission but can also include some additional functions. For example, the output module 3 can format the control strategy to meet the requirements of a specific power management interface. In addition, the output module 3 can also include a caching mechanism to temporarily store the control strategy in case of network or system failures, ensuring the continuity and reliability of the system.

[0150] Preferably, the system can also include a feedback module 4 for collecting the actual effect data after the implementation of the control strategy. These data can be sent back to the acquisition module 1 to form a closed-loop system, continuously optimizing the performance of prediction and control.

[0151] In practical applications, the system can be flexibly deployed on different hardware platforms. For large data centers, the system can be deployed on dedicated high-performance servers to make full use of their powerful computing capabilities. For small and medium-sized server clusters, the system can adopt a distributed architecture, deploying different modules on multiple devices to balance the computing load. In addition, considering the trend of edge computing, the system can also design a lightweight version to be directly deployed on the servers that need to be managed to achieve localized intelligent power management.

[0152] Through this modular and scalable design, the intelligent server power management system of the present invention can adapt to server environments of various scales and types, providing users with efficient and flexible power management solutions. Each module of the system can be optimized and upgraded according to specific requirements to ensure that the overall performance is always in the best state.

[0153] To verify the superiority of the intelligent server power management method and system of the present invention, a series of comparative experiments were conducted. The following will detail the settings of the examples and comparative examples, the test methods, and the result analysis.

[0154] Example 1: The intelligent server power management method of the present invention

[0155] In this example, the complete method described in the present invention was adopted, including topological dynamics load prediction, group theory transformation optimization, chaotic control strategy generation, quantum probability field optimization, and non-commutative geometry fine-tuning. The test environment was a medium-sized data center with 100 servers, and the test period was 30 days.

[0156] Comparative Example 1: Traditional time series prediction method

[0157] This comparative example adopted the traditional ARIMA (Autoregressive Integrated Moving Average) model for load prediction and used a PID (Proportional-Integral-Derivative) controller for power management. This is a relatively common method in the industry at present.

[0158] Comparative Example 2: Basic Machine Learning Method

[0159] This comparative example uses a load prediction model based on an LSTM (Long Short-Term Memory) neural network and combines a reinforcement learning algorithm to generate a power management strategy. This represents some new attempts in this field in recent years.

[0160] Test Metrics and Methods:

[0161] 1. Prediction Accuracy: The mean absolute percentage error (MAPE) is used to measure the accuracy of load prediction.

[0162] Calculation Method:

[0163] where, is the actual load value; is the predicted load value; is the number of test samples.

[0164] 2. Energy Efficiency Improvement: Compare the energy usage efficiency (PUE, Power Usage Effectiveness) before and after implementation.

[0165] Calculation Method: PUE = Total Equipment Energy Consumption / IT Equipment Energy Consumption

[0166] 3. Response Time: The response time of the system to sudden load changes, in seconds.

[0167] Test Method: Artificially create sudden load changes and record the time from the load change to the system's response.

[0168] 4. Policy Stability: The degree of fluctuation of the control policy, measured using the standard deviation.

[0169] Calculation Method: Sample the control policy within 30 days and calculate its standard deviation.

[0170] 5. System Robustness: The performance consistency under different load conditions, measured using the standard deviation of MAPE under different loads.

[0171] Calculation Method: Calculate MAPE under low, medium, and high loads respectively, and then calculate the standard deviation of these three MAPEs.

[0172] The test results are shown in Table 1 below:

[0173] Table 1 Comparison of Test Results between Example 1 and Comparative Example 1 and Comparative Example 2

[0174] Index Example 1 Comparative Example 1 Comparative Example 2 Prediction Accuracy (MAPE) 3.2 7.5 5.1 Energy Efficiency Improvement 18.5 8.3 12.7 Response Time (s) 2.3 5.7 3.9 Policy Stability 0.05 0.12 0.08 System Robustness 0.6 1.8 1.2

[0175] As can be clearly seen from the test results in Table 1, the method of the present invention (Example 1) is significantly superior to the traditional method (Comparative Example 1) and the basic machine learning method (Comparative Example 2) in all test metrics.

[0176] In terms of prediction accuracy, the method of the present invention reduced the MAPE to 3.2%, an increase of 57.3% compared to the traditional method and 37.3% compared to the basic machine learning method. This high-precision prediction benefits from the combination of topological dynamics load prediction and group theory transformation optimization, which can better capture the non-linear features and potential structures in the load data.

[0177] In terms of energy efficiency improvement, the method of the present invention achieved an 18.5% improvement, far exceeding the other two methods. This not only reflects more accurate load prediction but also demonstrates the advantages of chaos control strategies and quantum probability field optimization in generating efficient control strategies.

[0178] The significant reduction in response time (2.3 seconds, 5.7 seconds, and 3.9 seconds) demonstrates the excellent performance of this method in handling sudden load changes. This is mainly attributed to the non-commutative geometry fine-tuning technology, which can make more refined and rapid adjustments in the policy space.

[0179] The improvement in policy stability (standard deviations of 0.05, 0.12, and 0.08) indicates that the control strategies generated by this method are more stable, avoiding frequent large-scale adjustments, which is crucial for maintaining the long-term stability of server hardware.

[0180] The significant improvement in system robustness (0.6%, 1.8%, and 1.2%) proves that this method can maintain a high level of performance under different load conditions. This consistency is particularly important for coping with the complex and changing load environment of data centers.

[0181] Overall, these test results fully demonstrate the superiority of the method of the present invention. It not only performs well in each individual metric but, more importantly, shows a comprehensive and balanced performance improvement. This all-round improvement fully reflects the synergistic effect of multiple innovative points in the present invention. For example, the combination of topological dynamics and group theory improves prediction accuracy, the integration of chaos theory and quantum optimization enhances the efficiency and adaptability of policy generation, and the introduction of non-commutative geometry further improves the system's response speed and stability.

[0182] Such performance improvements are of great significance for the operation of modern data centers. Higher energy efficiency directly translates into reduced operating costs and improved environmental friendliness. Faster response times and more stable control strategies can improve service quality and reduce service interruptions caused by power problems. The improved system robustness provides greater confidence for data centers to handle various load conditions.

[0183] In summary, the server power intelligent management method and system of the present invention exhibit significant technical advantages and application values, providing an innovative and efficient solution for the power management of data centers.

[0184] It should be noted that the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A server power intelligent management method, characterized in that: include: The acquisition steps include: Get the historical load data sequence and current load data of the server; Processing steps include: Based on the historical load data sequence, performing topological dynamic load prediction to obtain a predicted load sequence; According to the predicted load sequence, optimizing is performed through group theory transformation optimization to obtain an optimized predicted load sequence; Based on the optimized predicted load sequence, generating a control strategy using a chaos control strategy; Applying quantum probability field optimization to the control strategy to obtain an optimized control strategy; The optimized control strategy is fine-tuned by non-commutative geometry to obtain a final control strategy; Output steps include: Outputting the final control strategy for adjusting the server power supply; The quantum probability field optimization specifically includes: Defining quantum states Among them, |ψ> is the quantum state, c i is the complex amplitude, |i> is the ground state, and N is the number of ground states; Mapping control strategies to quantum states; The optimized control strategy is obtained through quantum measurement operations.

2. The method according to claim 1, characterized in that The execution of topology dynamics load prediction specifically includes: Mapping the historical load data sequence to a high-dimensional topological space to obtain a sequence of mapped data points; constructing a dynamic system based on the sequence of mapped data points; The phase space is reconstructed using Takens embedding theorem and the vector field is estimated using a local linear embedding algorithm. predicting future load values ​​based on the vector field; The predicted load sequence is obtained through inverse mapping.

3. The method according to claim 2, characterized in that The dynamical system is defined by the following differential equation: Where y∈M represents the state of the system on the m-dimensional manifold M, F:M→TM is a vector field, F is a vector field function, M is an m-dimensional manifold, TM is the tangent space of M, and for each y∈M, F(y)∈T y M is a vector in the tangent space of M at point y, t represents time, is the derivative of y with respect to time t, T y M represents the tangent space of M at point y.

4. The method according to claim 1, characterized in that: The optimization by group theory transformation optimization specifically includes: Define the transformation group G = {g1, g2, ..., g k }, where each g i is a transformation operation; Applying group action to each predicted value in the predicted load sequence to obtain an optimized predicted value; All optimized forecast values ​​are combined into an optimized forecast load sequence.

5. The method according to claim 1, characterized in that The method of generating a control strategy by utilizing a chaos control strategy specifically includes: Construct parameterized chaotic functions; The optimized predicted load sequence is embedded into the chaotic system to obtain the control parameter sequence; Based on the chaotic state and control parameters, the control strategy is generated through the decision function.

6. The method according to claim 5, characterized in that The chaos function is defined as: With t+1 =f(z t ,μ), Among them, z t+1 is the chaotic state at time t+1, f is the parameterized chaotic function, z t is the chaotic state at time t, and μ is the control parameter.

7. The method according to claim 1, characterized in that The non-commutative geometry fine-tuning specifically includes: Defining a noncommutative manifold Points on it are represented by non-commutative coordinates; Map the optimized control strategy to the non-swap space; Applying the non-commutative gradient operator and the action functional, we obtain the final control policy.

8. The method according to claim 1, characterized in that The final control strategy is given by the following equation: in, is the final control strategy, Φ is the mapping function, X t is a point in noncommutative space, is a non-commutative gradient operator and S is the action functional.

9. A server power intelligent management system for executing the method according to any one of claims 1 to 8, characterized in that: include: An acquisition module is used to acquire the historical load data sequence and current load data of the server; Processing modules for: Based on the historical load data sequence, performing topological dynamics load prediction to obtain a predicted load sequence; According to the predicted load sequence, optimizing is performed through group theory transformation optimization to obtain an optimized predicted load sequence; Based on the optimized predicted load sequence, generating a control strategy using a chaos control strategy; Applying quantum probability field optimization to the control strategy to obtain an optimized control strategy; The optimized control strategy is fine-tuned by non-commutative geometry to obtain a final control strategy; The output module is used to output the final control strategy for adjusting the server power supply.

Images