Base station energy-saving optimization control method and device and readable storage medium

Abstract

The invention provides a base station energy-saving optimization control method and device and a readable storage medium, and the method comprises the steps: obtaining base station energy-saving related data, the base station energy-saving related data comprises at least one of equipment operation state data and service load data in a base station, external meteorological data of a position where the base station is located, and electricity price time-of-use data; according to the energy-saving related data of the base station, obtaining a service load rule and energy consumption related characteristics through an analysis algorithm, and constructing an energy consumption prediction model to obtain an energy consumption prediction result; based on an analysis result obtained through analysis and an energy consumption prediction result, a base station energy saving strategy is determined; and issuing the base station energy-saving strategy to the base station, and dynamically adjusting the base station energy-saving strategy based on real-time feedback information of the base station. On the premise of guaranteeing the communication service quality, the energy consumption of the base station is reduced to the maximum extent, and the overall energy efficiency and sustainable development capability of a communication network are improved.

Description

Technical Field

[0001] The present application relates to the field of energy-saving technology, and in particular to a base station energy-saving optimization control method, device and readable storage medium. Background Art

[0002] With the continuous expansion and evolution of modern communication networks, the number of communication base stations, as the key infrastructure of network coverage, is increasing, and the energy consumption problem is becoming more and more prominent. The continuous operation of base stations requires a large amount of electricity. The high energy consumption cost not only puts huge pressure on the operating costs of communication operators, but also does not meet the current social development requirements of energy conservation and emission reduction.

[0003] Traditional base station energy-saving methods are often limited to simple timed power on / off, fixed power adjustment and other extensive means, lacking comprehensive consideration of multi-dimensional factors such as the real-time operation status of the base station, complex business load conditions and changing external environment. For example, some energy-saving measures may affect the quality of communication services due to excessive energy saving during peak business hours, or fail to fully tap the energy-saving potential during low business hours, resulting in poor overall energy-saving effects and difficulty in achieving dynamic optimization.

[0004] Therefore, how to achieve base station energy saving more scientifically has become a problem that needs to be solved. Summary of the invention

[0005] The technical problem to be solved by the present application is to provide a base station energy-saving optimization control method, device and readable storage medium to solve the problems existing in the prior art in view of the above-mentioned deficiencies in the prior art.

[0006] In a first aspect, the present application provides a base station energy-saving optimization control method, the method

[0007] The law includes:

[0008] S1. Acquire base station energy-saving related data, wherein the base station energy-saving related data includes at least one of equipment operation status data inside the base station, service load data, external meteorological data at the location of the base station, and electricity price time-sharing data;

[0009] S2. According to the base station energy-saving related data, obtain the service load law and energy consumption related characteristics through the analysis algorithm, and build an energy consumption prediction model to obtain energy consumption prediction results;

[0010] S3. Determine the energy-saving strategy of the base station based on the analysis results and energy consumption prediction results obtained by the analysis;

[0011] S4. Send the base station energy-saving strategy to the base station, and dynamically adjust the base station energy-saving strategy based on real-time feedback information from the base station.

[0012] In some embodiments, in S2, obtaining a traffic load rule by analyzing an algorithm includes at least one of the following:

[0013] S211. Perform stationarity test, trend and seasonality analysis on business load data through time series analysis algorithm, build and train time series prediction model and predict future business load trend;

[0014] S212, taking the service load data as input, using a cluster analysis algorithm to classify different time periods according to the similarity of the service loads, and obtaining a service load distribution rule;

[0015] S213. By using an association rule mining algorithm, the business load data is combined with other related data to mine association rules, and the intrinsic relationship between different factors is obtained.

[0016] In some embodiments, in S2, the energy consumption correlation characteristics are obtained by analyzing the algorithm, including:

[0017] S221, determining the correlation coefficient between the equipment operation status data, the service load data, the external meteorological data and the energy consumption index through a correlation analysis algorithm, and determining a variable having a strong correlation with energy consumption;

[0018] S222. By using a regression analysis algorithm, a regression model is established with variables with strong correlation as independent variables and energy consumption as the dependent variable. The model structure is analyzed and optimized to obtain the relationship between energy consumption and various influencing factors.

[0019] In some embodiments, in S2, building an energy consumption prediction model includes:

[0020] S23, constructing an energy consumption prediction model based on the business load law and energy consumption correlation characteristics, wherein the energy consumption prediction model is constructed by selecting a single prediction model or adopting a model fusion strategy;

[0021] Among them, the single prediction model includes a support vector regression model or a random forest regression model, and the model fusion strategy includes a weighted average method or a Stacking ensemble learning method.

[0022] In some embodiments, in S3, the base station energy saving strategy includes: at least one of a device dynamic adjustment strategy, an intelligent power supply management strategy, and a collaborative optimization strategy.

[0023] In some embodiments, the device dynamically adjusts the strategy, including at least one of the following:

[0024] Dynamic adjustment of transmit power: Increase the transmit power of the radio unit during peak business hours and reduce it during low business hours.

[0025] Equipment shutdown and wake-up plan: Filter the devices that can be shut down based on energy consumption, functions and business scenarios, determine the shutdown period according to the business load rules, and formulate equipment shutdown and wake-up plans.

[0026] In some embodiments, the intelligent power supply management strategy includes at least one of the following:

[0027] Electricity price time-sharing response strategy: According to the electricity price time-sharing data, increase the charging capacity of base station equipment during the low electricity price period, and set high-energy consumption equipment to operate during the low electricity price period;

[0028] Power efficiency optimization strategy: Build a power module efficiency monitoring system based on power sensors and dynamically adjust the power module working mode according to the load conditions of the base station equipment.

[0029] In some embodiments, the collaborative optimization strategy includes at least one of the following:

[0030] Collaborative implementation of cross-base station resources: Based on the service load and coverage of the base station group, wireless parameters are adjusted to integrate resources during the service off-peak period, and service traffic is shared through load balancing algorithms during the service peak period.

[0031] Coordinated implementation with the computer room air conditioning and refrigeration system: intelligently adjust the air conditioning's cooling power, operating mode, and fan speed based on the temperature distribution of the equipment in the computer room, business load, and energy consumption.

[0032] In a second aspect, the present application provides a base station energy-saving optimization control device, the device comprising:

[0033] A data acquisition module, configured to acquire base station energy-saving related data, wherein the base station energy-saving related data includes at least one of equipment operation status data inside the base station, service load data, external meteorological data at the location of the base station, and electricity price time-sharing data;

[0034] A data analysis module is configured to obtain service load rules and energy consumption related characteristics through an analysis algorithm based on the base station energy-saving related data, and to construct an energy consumption prediction model to obtain energy consumption prediction results;

[0035] An energy-saving strategy module, which is configured to determine an energy-saving strategy for a base station based on an analysis result obtained by analysis and an energy consumption prediction result;

[0036] The strategy execution module is configured to send the base station energy-saving strategy to the base station and dynamically adjust the base station energy-saving strategy based on real-time feedback information from the base station.

[0037] In a third aspect, the present application provides a base station energy-saving optimization control device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the base station energy-saving optimization control method described in the first aspect above.

[0038] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the base station energy-saving optimization control method described in the first aspect is implemented.

[0039] The present application provides a base station energy-saving optimization control method, device and readable storage medium, the method comprising: obtaining base station energy-saving related data, the base station energy-saving related data including at least one of the equipment operation status data inside the base station, business load data, external meteorological data at the location of the base station, and electricity price time-sharing data; according to the base station energy-saving related data, obtain the business load law and energy consumption related characteristics through the analysis algorithm, and construct an energy consumption prediction model to obtain energy consumption prediction results; determine the base station energy-saving strategy based on the analysis results and energy consumption prediction results obtained by analysis; send the base station energy-saving strategy to the base station, and dynamically adjust the base station energy-saving strategy based on the real-time feedback information of the base station. The present application comprehensively collects and deeply analyzes base station-related data, and accurately formulates and dynamically adjusts energy-saving optimization strategies based on the analysis results, thereby minimizing base station energy consumption and improving the overall energy efficiency and sustainable development capabilities of the communication network while ensuring the quality of communication services. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0041] Figure 1 A flowchart of a base station energy-saving optimization control method provided in an embodiment of the present application;

[0042] Figure 2 A schematic diagram of performing data analysis based on the base station energy-saving related data provided in an embodiment of the present application;

[0043] Figure 3 A schematic diagram of a base station energy saving strategy provided in an embodiment of the present application;

[0044] Figure 4 A schematic diagram of the structure of a base station energy-saving optimization control device provided in an embodiment of the present application;

[0045] Figure 5 A schematic diagram of the structure of another base station energy-saving optimization control device provided in an embodiment of the present application.

[0046] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0047] In order to enable those skilled in the art to better understand the technical solution of the present application, the implementation mode of the present application will be further described in detail below with reference to the accompanying drawings.

[0048] It should be understood that the specific embodiments and drawings described herein are only used to explain the present application, rather than to limit the present application.

[0049] It can be understood that, in the absence of conflict, the various embodiments in the present application and the various features in the embodiments can be combined with each other.

[0050] It can be understood that, for the convenience of description, the drawings of the present application only show the parts related to the present application, while the parts unrelated to the present application are not shown in the drawings.

[0051] It can be understood that each unit and module involved in the embodiments of the present application may correspond to only one physical structure, or may be composed of multiple physical structures, or multiple units and modules may be integrated into one physical structure.

[0052] It can be understood that the terms "first", "second", etc. in the embodiments of the present application are used to distinguish different objects, or to distinguish different processing of the same object, rather than to describe a specific order of objects.

[0053] It is understandable that, in the absence of conflict, the functions and steps marked in the flowcharts and block diagrams of the present application may occur in an order different from that marked in the drawings.

[0054] It is understood that the flowcharts and block diagrams of the present application illustrate the possible architectures, functions, and operations of the systems, devices, equipment, and methods according to the various embodiments of the present application. Among them, each box in the flowchart or block diagram may represent a unit, module, program segment, code, which contains executable instructions for implementing the specified functions. Moreover, each box or combination of boxes in the block diagram and flowchart may be implemented by a hardware-based system that implements the specified functions, or by a combination of hardware and computer instructions.

[0055] It can be understood that the units and modules involved in the embodiments of the present application can be implemented by software or hardware, for example, the units and modules can be located in a processor.

[0056] Existing base station energy-saving technologies have the following disadvantages:

[0057] (1) Poor business load adaptability:

[0058] Existing technologies, such as timed energy-saving strategies, only adjust base station equipment at fixed times (for example, uniformly reduce transmission power or shut down equipment at night), without fully considering the complex changes in service loads of different base stations at different times. For example, the peak and trough periods of service loads of base stations in urban commercial areas and residential areas are significantly different. This simple timed operation is difficult to meet actual service needs, and it is easy to over-energy-conserve and affect the quality of communication services, or miss better energy-saving opportunities, and fail to implement energy-saving measures accurately according to the dynamic service load.

[0059] (2) Extensive energy consumption control:

[0060] Although energy consumption management based on simple rules takes into account factors such as business traffic, the rules are too simple and only set a few fixed energy consumption adjustment gears. They cannot accurately capture the complex nonlinear relationship between business load and energy consumption, making it difficult to achieve refined energy consumption regulation, resulting in a significant reduction in energy-saving effects.

[0061] (3) Lack of coordination among equipment within the base station:

[0062] In the past, the focus was on energy-saving improvements of individual devices, such as improving the energy efficiency of a device (such as a power amplifier, chip, etc.) through hardware optimization, but the characteristics of the interconnection and collaborative work between the various devices in the base station were ignored. In fact, adjusting the operating status of a device (such as changing the transmission power of the RF unit) may affect other devices (such as the load of the power module, the processing capacity of the baseband processing unit, etc.). If the coordination of various devices is not comprehensively considered, it is difficult to achieve the overall optimal energy-saving state.

[0063] (4) Limited resource coordination between base stations:

[0064] In base station cluster scenarios involving multiple base stations, existing technologies lack an effective cross-base station resource coordination mechanism. Simply adjusting the parameters of a base station device may disrupt the coverage balance and load balancing between base stations, affecting the service quality and energy consumption of surrounding base stations, and failing to integrate resources from the overall perspective of the base station cluster to achieve collaborative energy-saving optimization.

[0065] (5) Lack of in-depth data analysis:

[0066] Existing base station energy-saving management mostly relies on empirical operations or simple threshold settings. It does not fully utilize the large amount of data generated during the operation of the base station (equipment operating status, business load, environmental data, etc.) for in-depth mining and analysis. It is difficult to reveal the intrinsic connection between energy consumption and various influencing factors, and it is impossible to accurately grasp the trend of energy consumption changes, and it is impossible to formulate scientific and reasonable energy-saving strategies in advance.

[0067] (6) Low accuracy of prediction model:

[0068] In terms of energy consumption prediction, there is no systematic use of appropriate analysis algorithms to build an accurate energy consumption prediction model. Energy consumption is only roughly estimated or judged based on simple linear relationships. It cannot accurately reflect the actual energy consumption changes of base stations in different scenarios, making subsequent energy-saving control measures lack reliable data support, resulting in difficulty in balancing energy-saving effects and service quality assurance.

[0069] (7) Inflexible strategy adjustment:

[0070] Once the existing energy-saving strategies are formulated and implemented, they often lack effective feedback mechanisms and dynamic adjustment methods. When faced with changes in business demand (such as temporary large-scale events that cause sudden changes in base station business load) and changes in equipment operating conditions (such as sudden equipment failures or performance degradation), the energy-saving optimization strategies cannot be adjusted and optimized in a timely manner, and it is impossible to ensure that the energy-saving strategies are always in the best state to effectively save energy and ensure the quality of communication services.

[0071] To solve the above problems, the present application provides a method and system that can comprehensively collect and deeply analyze base station related data, and accurately formulate and dynamically adjust energy-saving optimization strategies based on the analysis results, so as to minimize the energy consumption of base stations while ensuring the quality of communication services, and improve the overall energy efficiency and sustainable development capabilities of the communication network.

[0072] The main purposes of this application include:

[0073] Achieve precise energy saving: By collecting multi-dimensional data inside the base station (equipment operating status, business load, weather, electricity price time-sharing, etc.), using a variety of analysis algorithms to deeply explore the business load rules and energy consumption related characteristics, and build an accurate energy consumption prediction model, and then formulate refined energy-saving strategies that fit different business scenarios, different time periods, and different environmental conditions based on the prediction results, it is possible to dynamically and accurately adjust equipment operating parameters and resource allocation according to the actual situation of the base station, minimize base station energy consumption, and ensure that the quality of communication services is not affected.

[0074] Strengthen equipment collaborative optimization: Within the base station, comprehensively consider the mutual influence between various devices, formulate equipment dynamic adjustment strategies (such as reasonable transmission power adjustment, equipment shutdown and wake-up plans, etc.) and intelligent power supply management strategies (such as dynamically adjusting the power module working mode according to the equipment load, etc.), to ensure that various devices work together to achieve the best overall energy saving;

[0075] At the base station group level, by formulating a collaborative optimization strategy, it is possible to integrate base station resources with overlapping coverage and light load during business off-peak periods, and coordinate surrounding base stations with relatively low loads to share part of the business traffic during business peak periods, thereby achieving effective coordination of resources between base stations and overall energy-saving optimization, and improving the comprehensive energy-saving benefits of the base station group.

[0076] Improve the accuracy of energy consumption prediction: With the help of big data analysis, the rich data collected is comprehensively processed and deeply mined. Various analysis algorithms such as time series analysis, cluster analysis, correlation analysis, and regression analysis are used to screen out variables that are strongly correlated with energy consumption. A scientific and reasonable energy consumption prediction model is constructed (including selecting a suitable single prediction model or using a model fusion strategy). The model is strictly trained, verified, and optimized to accurately predict the energy consumption changes of base stations under different conditions, providing a reliable data basis for the formulation of energy-saving strategies.

[0077] Establish a dynamic adjustment mechanism: After the energy-saving optimization strategy is converted into specific control instructions and sent to the base station equipment for execution, the execution status and actual effect are monitored in real time (by collecting performance indicator data such as signal coverage quality, business service quality, and equipment operating status), and the relevant parameters in the energy-saving optimization strategy are promptly fine-tuned according to the feedback information. When encountering problems that cannot be solved by parameter fine-tuning or when there are new business needs or changes in equipment operating conditions, the entire energy-saving optimization strategy can be comprehensively updated to ensure that the energy-saving strategy can flexibly adapt to various changes, always maintain the best state of balancing energy saving and service quality assurance, and ensure the long-term stable and efficient and energy-saving operation of the base station.

[0078] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0079] The present application provides a base station energy-saving optimization control method, the working process of which can be implemented by electronic devices, such as computers, handheld smart terminals, etc. For the convenience of explanation, the embodiments of the present application are described with the method execution subject being a computer.

[0080] Figure 1 A schematic diagram of a base station energy-saving optimization control method provided in an embodiment of the present application, such as Figure 1 As shown, the present application provides a base station energy-saving optimization control method, the method includes S1-S4, which are as follows:

[0081] S1. Acquire base station energy-saving related data, wherein the base station energy-saving related data includes at least one of equipment operation status data inside the base station, service load data, external meteorological data at the location of the base station, and electricity price time-sharing data;

[0082] The device operation status data includes at least one of the transmit power and receive power of the radio frequency unit, the gain and operating current of the power amplifier, the processing load of the baseband processing unit, and the input and output voltage and current of the power module;

[0083] The service load data includes at least one of voice call duration, call number, uplink and downlink traffic of data services, and user access and switching frequency;

[0084] The external meteorological data includes at least one of temperature, humidity, wind speed, wind direction, and air pressure;

[0085] The electricity price time-sharing data includes electricity price standards corresponding to different time periods and the time period division.

[0086] Optionally, after S1, it also includes: preprocessing the base station energy-saving related data, and the preprocessing includes at least one of data cleaning, missing value processing and data standardization.

[0087] Among them, data cleaning includes removing abnormal data caused by sensor failure, network failure, transmission error or non-compliance with normal business usage rules;

[0088] The missing value processing includes processing the existing missing values ​​by using linear interpolation, mean filling or model prediction-based filling;

[0089] The data standardization includes standardizing the numerical data using the Min-Max standardization or Z-score standardization method to standardize the data within a reasonable standard range;

[0090] Specifically, all collected numerical data are standardized so that data with different dimensions and different value ranges are unified into a reasonable standard range, which is convenient for subsequent data analysis and model construction. Commonly used standardization methods such as Min-Max standardization map the data to the [0,1] interval, and the calculation formula is:

[0091]

[0092] Among them, X is the original data, X min and X max They are the minimum and maximum values ​​of the data set where the data is located;

[0093] Alternatively, Z-score standardization is used to transform the data into a distribution with a mean of 0 and a standard deviation of 1. The calculation formula is:

[0094]

[0095] Here, μ is the mean of the data set and σ is the standard deviation.

[0096] S2. According to the base station energy-saving related data, obtain the service load law and energy consumption related characteristics through the analysis algorithm, and build an energy consumption prediction model to obtain energy consumption prediction results;

[0097] Figure 2 A schematic diagram of performing data analysis based on the base station energy-saving related data provided in an embodiment of the present application, such as Figure 2 As shown, in some embodiments, in S2, obtaining a business load rule by analyzing an algorithm includes at least one of the following:

[0098] S211. Perform stationarity test, trend and seasonality analysis on business load data through time series analysis algorithm, build and train time series prediction model and predict future business load trend;

[0099] S212, taking the service load data as input, using a cluster analysis algorithm to classify different time periods according to the similarity of the service loads, and obtaining a service load distribution rule;

[0100] S213. By using an association rule mining algorithm, the business load data is combined with other related data to mine association rules, and the intrinsic relationship between different factors is obtained.

[0101] In some embodiments, in S2, the energy consumption correlation characteristics are obtained by analyzing the algorithm, including:

[0102] S221, determining the correlation coefficient between the equipment operation status data, the service load data, the external meteorological data and the energy consumption index through a correlation analysis algorithm, and determining a variable having a strong correlation with energy consumption;

[0103] Specifically, the numpy.corrcoef function in Python is used to calculate the Pearson correlation coefficient. For two variables X and Y, the function returns a 2×2 matrix, in which the elements located at the upper left corner and the lower right corner are the correlation coefficients between X and Y, and the value range is between -1 and 1; the closer the absolute value is to 1, the stronger the correlation is; if the coefficient is positive, it indicates a positive correlation, that is, when the variable increases, the energy consumption also increases, and vice versa, it is a negative correlation; for example, the calculated Pearson correlation coefficient between the base station transmit power and energy consumption is above 0.8, indicating that there is a strong positive correlation between the transmit power and energy consumption, and the change of transmit power has a significant impact on the energy consumption of the base station.

[0104] S222. By using a regression analysis algorithm, a regression model is established with variables with strong correlation as independent variables and energy consumption as the dependent variable. The model structure is analyzed and optimized to obtain the relationship between energy consumption and various influencing factors.

[0105] Specifically, a multivariate linear regression model is used for model fitting and parameter estimation. Assume that the transmission power, service load, equipment operation time, etc. are selected as independent variables x 1 ,x 2 ,x 3 ,…, energy consumption is the dependent variable y, and the regression equation y=β is established 0 +β 1 x 1 +β 2 x 2 +β 3 x 3 +…+ε(where β 0 is the intercept term, β 1 ,β 2 ,β 3 ,… are regression coefficients, ε is a random error term);

[0106] The model parameters (β 0 ,β 1 ,β 2 ,…) to estimate the observed value y and the predicted value The sum of squares of errors between (calculated by the regression equation) is minimized, and the estimated values ​​of each parameter and the corresponding standard errors, t values ​​and other statistics are obtained to determine whether the influence of each variable on the dependent variable is significant;

[0107] The regression model is tested, mainly including goodness of fit test (such as through R 2 The statistic measures how well the model fits the data, R 2The closer the value is to 1, the better the model fit is), significance test (F test is used to determine whether the entire regression equation is significant, and t test is used to determine whether each single independent variable has a significant impact on the dependent variable) and residual analysis (check whether the residual meets the assumptions of normality, independence and homogeneity of variance. If not, it means that there may be problems with the model);

[0108] The model is optimized according to the test results. If multicollinearity problems are found (the independent variables are highly correlated), principal component analysis, stepwise regression and other methods can be used to screen variables or reduce dimensionality. If there are problems such as heteroscedasticity in the residuals, the model can be improved by transforming the variables (such as taking logarithms, square roots, etc.) or using weighted least squares methods to ensure that the regression model can accurately reflect the relationship between energy consumption and various influencing factors, providing a reliable model basis for subsequent energy consumption forecasts.

[0109] In some embodiments, in S2, building an energy consumption prediction model includes:

[0110] An energy consumption prediction model is constructed based on the business load law and energy consumption correlation characteristics. The energy consumption prediction model is constructed by selecting a single prediction model or adopting a model fusion strategy, including training, verifying and optimizing the selected model, and using the trained model to predict the energy consumption of the base station;

[0111] Among them, the single prediction model includes a support vector regression model or a random forest regression model, and the model fusion strategy includes a weighted average method or a Stacking ensemble learning method.

[0112] Specifically, firstly, data preparation is carried out to collect data including energy consumption and related influencing factors, ensure that these data have been cleaned in advance to remove outliers, fill in missing values, and standardize, and arrange the sorted data in chronological order or other reasonable logic; divide the data set into training set, validation set and test set according to a certain ratio, and the common division ratio can be 70% as training set, 15% as validation set, and 15% as test set;

[0113] In this application, the construction process of different models is as follows:

[0114] (1) Support Vector Regression (SVR) model parameter selection and training, specifically:

[0115] Determine the kernel function based on the data characteristics and use the Gaussian kernel function; based on the selected Gaussian kernel function, determine the value range of key parameters related to it for optimization; determine the penalty coefficient C (generally within the range of 2 -5 To 2 5Try different values ​​between , its role is to balance the complexity of the model and the training error. The larger the value, the heavier the penalty for the training error, the more complex the model, but it is easy to cause overfitting) and the γ parameter of the Gaussian kernel (usually between 2 -15 To 2 3 It takes values ​​between , which determines the width of the Gaussian kernel function and affects the model's fitting effect on the data);

[0116] Through the cross-validation method, search and test within the above-determined parameter value range to find the optimal parameter combination to achieve the best model performance. For example, K-fold cross-validation is used (common K values ​​can be 5 or 10), the training set is divided into K subsets, K-1 subsets are taken as new training sets each time, and the remaining 1 subset is used as the validation subset. The SVR model is trained with different C and γ parameter combinations, and the model performance is evaluated on the validation subset. Common evaluation indicators such as mean squared error MSE (Mean Squared Error) and mean absolute error MAE (Mean Absolute Error) are used to record the validation set performance indicators under different parameter combinations. Finally, the parameter combination that makes the validation set performance the best is selected as the final model parameter to complete the model training process on the entire training set.

[0117] Use the test set data to evaluate the performance of the trained SVR model and calculate the mean square error Where n is the number of test set samples, y i is the actual energy consumption value, is the energy consumption value predicted by the model), mean absolute error Mean absolute percentage error Indicators such as performance evaluation can be used to analyze the prediction accuracy and stability of the model on data that has not been trained, and to understand its advantages and disadvantages as well as applicable scenarios. For example, it can be determined whether the model is accurate in predicting energy consumption during business peak hours when processing current base station energy consumption data, or whether it has more advantages during business troughs.

[0118] (2) Determination and training of model parameters of random forest regression model, specifically:

[0119] Key parameter selection, including the number of decision trees, the number of randomly sampled features for each tree, and the maximum depth of each tree;

[0120] The number of decision trees: Generally, it needs to be determined by testing different values. Common values ​​range from dozens to hundreds of trees. You can first set an initial value (such as 100 trees), and then observe the performance of the model on the validation set (such as changes in indicators such as MSE and MAE), gradually increase or decrease the number of decision trees, weigh the computing cost and model performance, and determine a better number of decision trees (for example, 200 trees can achieve a good performance on the validation set and acceptable computing resource consumption);

[0121] The number of randomly sampled features for each tree is usually set according to the total number of input features. For example, if there are N input features in total, it can generally be set according to Or log 2 The number of features randomly extracted by each tree is selected according to the rule (N) to increase the diversity of decision trees and reduce the risk of overfitting.

[0122] Maximum depth of each tree: Controls the growth of the tree to avoid overfitting. You can set a suitable maximum depth value (such as 10-20 layers), or let the tree grow naturally until the number of node samples is small (such as the number of node samples is less than a certain threshold, such as 5 samples) and other stopping conditions appear. The specific adjustment is based on the data complexity and model performance. By observing the model error index on the validation set, it is judged whether the currently set maximum depth is appropriate. If overfitting occurs, the maximum depth is appropriately reduced;

[0123] Based on the training set data, each decision tree is constructed by sampling with replacement (Bootstrap sampling); in the process of tree construction, according to the selected random sampling features and sample data, according to certain node splitting rules (such as determining which feature and which value to split the node based on indicators such as information gain and Gini index, so that the data purity of the child nodes after the split is higher, for example, selecting the feature with the largest information gain and the corresponding value for splitting), the construction process is repeated until the maximum depth of the tree or other stopping conditions are reached, and finally a random forest model composed of multiple decision trees is formed;

[0124] The test set data is also used to evaluate the performance of the trained random forest regression model, and indicators such as MSE, MAE, and MAPE (Mean Absolute Percentage Error) are calculated. The accuracy and stability of the model in energy consumption prediction under different business scenarios (such as business peaks, troughs, and stable periods) and its robustness to outliers and noise are analyzed to clarify the advantages and limitations of the model.

[0125] (3) The operation of the weighted average method is as follows:

[0126] Before using the weighted average method for model fusion, it is necessary to first analyze the performance of the two single models, the support vector regression model and the random forest regression model, on the test set in detail, calculate their respective evaluation indicators such as MSE, MAE, and MAPE, and understand the prediction performance of each model in different business scenarios and different data characteristics. For example, through analysis, it is found that the MAPE of the energy consumption prediction of the SVR model during the business off-peak period is about 10%, while the MAPE during the business peak period is 18%; the MAPE of the random forest regression model during the business off-peak period is 12%, and the MAPE during the business peak period is 15%, so as to grasp the advantage period and overall performance difference of each model.

[0127] Determine the weight of fusion based on the performance of a single model; assign weights according to the inverse ratio of the performance indicators of each model on the test set; for example, calculate the average MAPE of the two models on the entire test set, assuming that the average MAPE of the SVR model is MAPE SVR , the average MAPE of the random forest regression model is MAPE RF , then the weight ω of the SVR model SVR According to the formula Calculate the weight ω of the random forest regression model RF =1-ω SVR ;

[0128] Fusion prediction: After determining the weights, for new input data that need to predict energy consumption (such as new business load, equipment parameters, environmental factors and other data combinations), the support vector regression model and the random forest regression model are used to predict and obtain their respective predicted energy consumption values. Then, the weighted sum is performed according to the determined weights to obtain the fused predicted energy consumption value y fusion , the calculation formula is In this way, the advantages of two single models are combined to improve the accuracy and stability of the overall energy consumption prediction.

[0129] (4) The operation of the stacking ensemble learning method is as follows:

[0130] First, the support vector regression model and the random forest regression model are trained; the training is performed according to the training methods of the respective models to obtain the trained SVR basic model and the random forest basic model, ensuring that each basic model can achieve a good fitting effect on the training set. At the same time, the parameters are tuned and overfitting is prevented through the validation set, so that it has a certain generalization ability and can make reasonable energy consumption predictions for different data samples;

[0131] Generate new features, apply the trained basic model to the training set data, and obtain their respective predicted output values ​​for each sample in the training set. For example, each sample is predicted by the SVR model to obtain an energy consumption prediction value. The random forest regression model predicts The predicted output values ​​of these basic models are used as new features to construct a new dataset together with some of the original important input features. Assuming that there are n original input features, after the predicted values ​​of the two basic models, the SVR model and the random forest regression model, are added as new features, each sample in the new dataset will have n+2 features, thereby generating a new feature dataset for meta-learner training.

[0132] Meta-learner training and selection: Select a suitable meta-learner. Taking the linear regression model as an example, divide the new feature data set into a new training set and a validation set (the division ratio can be based on actual conditions, such as 80% training set and 20% validation set). Use the training set data to train the meta-learner according to the conventional training method of linear regression (such as least squares method to estimate parameters, etc.), so that it can learn how to best predict the final energy consumption value based on these "new features" output by the basic model. At the same time, adjust and optimize the parameters and performance of the meta-learner through the validation set, and select the meta-learner parameter combination that makes the validation set performance the best, to ensure that the meta-learner can effectively integrate the advantages of the basic model and improve the overall prediction effect.

[0133] Fusion prediction: In the prediction stage, for new data samples to be predicted, each basic model (SVR model and random forest regression model) is first used to predict them, and their respective prediction results are obtained (such as ), and input these results as new features into the trained meta-learner, which then outputs the final fused predicted energy consumption value based on the learned relationship. This allows the advantages of multiple basic models to be integrated through the Stacking ensemble learning method, further improving the accuracy and generalization ability of energy consumption prediction, and providing a more reliable basis for energy consumption prediction for subsequent applications such as base station energy-saving optimization control.

[0134] S3. Determine the energy-saving strategy of the base station based on the analysis results and energy consumption prediction results obtained by the analysis;

[0135] Figure 3 A schematic diagram of a base station energy saving strategy provided in an embodiment of the present application, such as Figure 3 As shown, in some embodiments, in S3, the base station energy-saving strategy includes: at least one of a device dynamic adjustment strategy, an intelligent power supply management strategy, and a collaborative optimization strategy.

[0136] In some embodiments, the device dynamically adjusts the strategy, including at least one of the following:

[0137] Dynamic adjustment of transmit power: Increase the transmit power of the radio unit during peak business hours and reduce it during low business hours.

[0138] Equipment shutdown and wake-up plan: Filter the devices that can be shut down based on energy consumption, functions and business scenarios, determine the shutdown period according to the business load rules, and formulate equipment shutdown and wake-up plans.

[0139] The present application formulates a dynamic device adjustment strategy based on the business load analysis and energy consumption prediction results. The dynamic device adjustment strategy includes gradually reducing the base station radio frequency unit transmission power according to the preset power adjustment step during the business off-peak period, and monitoring the signal strength and user business experience in real time to ensure that the communication service requirements are met, and determining the list of equipment that can be shut down and the corresponding shutdown period, setting a timed wake-up mechanism, and automatically waking up the corresponding equipment according to the recovery of the business load;

[0140] The device dynamically adjusts the strategy implementation, specifically:

[0141] Dynamic adjustment of transmission power: By analyzing historical service load data to divide time periods, and combining energy consumption prediction to set transmission power reference values ​​for each period (higher power is set for high-load periods, and power is greatly reduced for low-load periods); relying on signal monitoring points in the coverage area and the network management system to monitor signal and service experience indicators in real time, setting thresholds to ensure quality, adjusting power in the event of anomalies, and operation and maintenance personnel to investigate causes and optimize;

[0142] Implementation of equipment shutdown and wake-up plan: Check base station equipment and select equipment that can be shut down based on energy consumption, function and business scenarios; determine shutdown periods and formulate plans based on business load patterns, set wake-up conditions based on business load forecasts, monitor equipment wake-up status, and promptly troubleshoot and repair any abnormalities to ensure communication services.

[0143] In some embodiments, the intelligent power supply management strategy includes at least one of the following:

[0144] Electricity price time-sharing response strategy: According to the electricity price time-sharing data, increase the charging capacity of base station equipment during the low electricity price period, and set high-energy consumption equipment to operate during the low electricity price period;

[0145] Power efficiency optimization strategy: Build a power module efficiency monitoring system based on power sensors and dynamically adjust the power module working mode according to the load conditions of the base station equipment.

[0146] The present application also formulates an intelligent power supply management strategy, which includes appropriately increasing the charging amount of base station equipment during the low electricity price period based on the electricity price time-sharing data, arranging non-critical and energy-intensive equipment to operate during the low electricity price period, and real-time monitoring of the conversion efficiency of the power module, dynamically adjusting the working mode of the power module according to the equipment load condition, so that it maintains a high conversion efficiency state;

[0147] Implementation of intelligent power supply management strategy, specifically:

[0148] Implementation of electricity price time-of-use response strategy: Understand the electricity price time-of-use policy, analyze the business load and energy consumption characteristics of different electricity price periods; charge energy storage equipment more during off-peak hours and reasonably dispatch auxiliary equipment to operate; adjust the duration or power accordingly during peak hours and normal times; monitor electricity costs in real time and dynamically optimize strategies as needed.

[0149] Implementation of power efficiency optimization strategy: Install power sensors to build a power module efficiency monitoring system and record data in real time; according to the load of base station equipment, dynamically adjust the power module working mode (reduce voltage and switching frequency when low load) through intelligent decision-making algorithms to improve efficiency.

[0150] In some embodiments, the collaborative optimization strategy includes at least one of the following:

[0151] Collaborative implementation of cross-base station resources: Based on the service load and coverage of the base station group, wireless parameters are adjusted to integrate resources during the service off-peak period, and service traffic is shared through load balancing algorithms during the service peak period.

[0152] Coordinated implementation with the computer room air conditioning and refrigeration system: intelligently adjust the air conditioning's cooling power, operating mode, and fan speed based on the temperature distribution of the equipment in the computer room, business load, and energy consumption.

[0153] The present application also formulates a collaborative optimization strategy, which includes cross-base station resource coordination, integrating base station resources with overlapping coverage and light load during the business off-peak period for a base station group composed of multiple adjacent base stations, and coordinating surrounding base stations with relatively low loads to share part of the business traffic during the business peak period; and working in coordination with the air conditioning and refrigeration system of the computer room where the base station is located, intelligently adjusting the cooling power, operation mode and fan speed of the air conditioner according to the temperature distribution of the equipment in the computer room, the business load and energy consumption;

[0154] Collaborative optimization strategy implementation, specifically:

[0155] Collaborative implementation of cross-base station resources: Analyze the service load and coverage of the base station group and classify the base stations; adjust wireless parameters to integrate resources during off-peak hours, use load balancing algorithms to share service traffic during peak hours, monitor the entire process and fine-tune parameters as needed to ensure efficient and energy-saving operation;

[0156] Coordinated implementation with the computer room air conditioning and refrigeration system: deploy sensors in the computer room to collect environmental, business load and energy consumption data, and build models to analyze the relationship between temperature, business and energy consumption; based on the model and real-time data, intelligently adjust the air conditioning refrigeration power, mode and fan speed, monitor the adjustment effect, and make timely adjustments in case of problems to ensure coordinated energy saving.

[0157] S4. Send the base station energy-saving strategy to the base station, and dynamically adjust the base station energy-saving strategy based on real-time feedback information from the base station.

[0158] Specifically, S4 includes:

[0159] S41, converting the energy-saving optimization strategy into a specific control instruction, sending the control instruction to the base station device for execution through the communication interface and protocol compatible with the base station devices and related systems, and monitoring the execution of the instructions by the devices and systems in real time;

[0160] S42, continuously collecting various performance indicator data of the base station after executing the energy-saving optimization strategy, the performance indicator data including signal coverage quality indicator, business service quality indicator and equipment operation status indicator, comparing and analyzing the real-time monitored performance indicator with the pre-set performance threshold, and evaluating the actual effect of the energy-saving strategy;

[0161] S43. Analyze feedback information based on performance indicator data and energy-saving effect evaluation results, and make targeted fine-tuning of relevant parameters in the energy-saving optimization strategy. When problems that cannot be solved by parameter fine-tuning arise, or when there are new business requirements or changes in equipment operating conditions, comprehensively update the entire energy-saving optimization strategy. Through a continuous cycle optimization mechanism, ensure that the energy-saving strategy is always in the best state to effectively save energy and ensure the quality of communication services.

[0162] The present application provides a base station energy-saving optimization control method. During the implementation process, it is mainly based on the collection and analysis of the operating data of existing base station equipment, and then the equipment operating status is adjusted through control instructions. The operations involved are basically expanded and optimized based on the existing base station system architecture and equipment functions. It does not require large-scale replacement or modification of the core equipment of the base station. It can be well compatible and coordinated with existing base station equipment, communication networks, and power supply systems, reducing the difficulty and cost of hardware modification during the implementation process.

[0163] In addition, it can adapt to the differences in different base station scenarios: whether it is a base station in different geographical locations such as cities and villages, or a base station of different types (such as macro base stations, micro base stations, etc.) and different business load characteristics (such as base stations in commercial areas, residential areas, etc.), the technical solution of this application can collect corresponding specific data, use data analysis to explore the respective business load rules and energy consumption correlation characteristics, and then formulate energy-saving optimization strategies that fit the actual situation. It has strong adaptability and can be applied in a variety of base station scenarios.

[0164] In addition, the entire energy-saving optimization control process of the present application has certain automation and intelligent characteristics. For example, the energy consumption prediction model can automatically give prediction results based on data changes, and the dynamic adjustment of energy-saving strategies can also automatically optimize some parameters according to preset rules and algorithms. This reduces the workload of manual intervention to a certain extent, reduces the risks caused by human operational errors, and improves the efficiency and quality of operation and maintenance management.

[0165] It should be understood that, although the various steps in the flowcharts in the above-described embodiments are sequentially displayed according to the indications of the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps is not strictly limited in order, and they can be executed in other orders. Moreover, at least a portion of the steps in the figure may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily to be carried out sequentially, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0166] Figure 4 A schematic diagram of a base station energy-saving optimization control device provided in an embodiment of the present application, such as Figure 4 As shown, the present application provides a base station energy-saving optimization control device, the device comprising:

[0167] A data acquisition module 11 is configured to acquire base station energy-saving related data, wherein the base station energy-saving related data includes at least one of equipment operation status data inside the base station, service load data, external meteorological data at the location of the base station, and electricity price time-sharing data;

[0168] A data analysis module 12 is configured to obtain service load rules and energy consumption related characteristics through an analysis algorithm based on the base station energy-saving related data, and to construct an energy consumption prediction model to obtain energy consumption prediction results;

[0169] An energy-saving strategy module 13, which is configured to determine a base station energy-saving strategy based on the analysis results and energy consumption prediction results obtained by analysis;

[0170] The strategy execution module 14 is configured to send the base station energy-saving strategy to the base station and dynamically adjust the base station energy-saving strategy based on the real-time feedback information of the base station.

[0171] Regarding the limitation on the base station energy-saving optimization control device, reference may be made to the limitation on the base station energy-saving optimization control method in the above-mentioned embodiments of the present application, and this embodiment will not be repeated here.

[0172] Figure 5 Another schematic diagram of the base station energy-saving optimization control device provided in the embodiment of the present application is as follows Figure 5 As shown, the device includes a memory 22 and a processor 21, the memory stores a computer program, and the processor is configured to run the computer program to execute the methods in the above embodiments of the present application.

[0173] The memory is connected to the processor, the memory may be a flash memory or a read-only memory or other memory, and the processor may be a central processing unit or a single-chip microcomputer.

[0174] In some embodiments, the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the methods in the above embodiments of the present application are implemented.

[0175] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules or other data). Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable read only memory), flash memory or other memory technology, CD-ROM (Compact Disc Read-Only Memory), digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0176] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present application, but the present application is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and substance of the present application, and these modifications and improvements are also considered to be within the scope of protection of the present application.

Claims

1. A base station energy-saving optimization control method, characterized in that: The method comprises: S1. Acquire base station energy-saving related data, wherein the base station energy-saving related data includes at least one of equipment operation status data inside the base station, service load data, external meteorological data at the location of the base station, and electricity price time-sharing data; S2. According to the base station energy-saving related data, obtain the service load law and energy consumption related characteristics through the analysis algorithm, and build an energy consumption prediction model to obtain energy consumption prediction results; S3. Determine the energy-saving strategy of the base station based on the analysis results and energy consumption prediction results obtained by the analysis; S4. Send the base station energy-saving strategy to the base station, and dynamically adjust the base station energy-saving strategy based on real-time feedback information from the base station.

2. The base station energy-saving optimization control method according to claim 1, characterized in that: In S2, the business load pattern is obtained by analyzing the algorithm, including at least one of the following: S211. Perform stationarity test, trend and seasonality analysis on business load data through time series analysis algorithm, build and train time series prediction model and predict future business load trend; S212, taking the service load data as input, using a cluster analysis algorithm to classify different time periods according to the similarity of the service loads, and obtaining a service load distribution rule; S213. By using an association rule mining algorithm, the business load data is combined with other related data to mine association rules, and the intrinsic relationship between different factors is obtained.

3. The base station energy-saving optimization control method according to claim 1, characterized in that: In S2, the energy consumption correlation characteristics are obtained through the analysis algorithm, including: S221, determining the correlation coefficient between the equipment operation status data, the service load data, the external meteorological data and the energy consumption index through a correlation analysis algorithm, and determining a variable having a strong correlation with energy consumption; S222. By using a regression analysis algorithm, a regression model is established with variables with strong correlation as independent variables and energy consumption as the dependent variable. The model structure is analyzed and optimized to obtain the relationship between energy consumption and various influencing factors.

4. The base station energy-saving optimization control method according to claim 1, characterized in that: In S2, an energy consumption prediction model is constructed, including: S23, constructing an energy consumption prediction model based on the business load law and energy consumption correlation characteristics, wherein the energy consumption prediction model is constructed by selecting a single prediction model or adopting a model fusion strategy; Among them, the single prediction model includes a support vector regression model or a random forest regression model, and the model fusion strategy includes a weighted average method or a Stacking ensemble learning method.

5. The base station energy-saving optimization control method according to any one of claims 1 to 4, characterized in that: In S3, the base station energy-saving strategy includes: at least one of a device dynamic adjustment strategy, an intelligent power supply management strategy, and a collaborative optimization strategy.

6. The base station energy-saving optimization control method according to claim 5, characterized in that: The device dynamic adjustment strategy includes at least one of the following: Dynamic adjustment of transmit power: Increase the transmit power of the radio unit during peak business hours and reduce it during low business hours. Equipment shutdown and wake-up plan: Filter the devices that can be shut down based on energy consumption, functions and business scenarios, determine the shutdown period according to the business load rules, and formulate equipment shutdown and wake-up plans.

7. The base station energy-saving optimization control method according to claim 5, characterized in that: The intelligent power supply management strategy includes at least one of the following: Electricity price time-sharing response strategy: According to the electricity price time-sharing data, increase the charging capacity of base station equipment during the low electricity price period, and set high-energy consumption equipment to operate during the low electricity price period; Power efficiency optimization strategy: Build a power module efficiency monitoring system based on power sensors and dynamically adjust the power module working mode according to the load conditions of the base station equipment.

8. The base station energy-saving optimization control method according to claim 5, characterized in that: The collaborative optimization strategy includes at least one of the following: Collaborative implementation of cross-base station resources: Based on the service load and coverage of the base station group, wireless parameters are adjusted to integrate resources during the service off-peak period, and service traffic is shared through load balancing algorithms during the service peak period. Coordinated implementation with the computer room air conditioning and refrigeration system: intelligently adjust the air conditioning's cooling power, operating mode, and fan speed based on the temperature distribution of the equipment in the computer room, business load, and energy consumption.

9. A base station energy-saving optimization control device, characterized in that: The device comprises: A data acquisition module, configured to acquire base station energy-saving related data, wherein the base station energy-saving related data includes at least one of equipment operation status data inside the base station, service load data, external meteorological data at the location of the base station, and electricity price time-sharing data; A data analysis module is configured to obtain service load rules and energy consumption related characteristics through an analysis algorithm based on the base station energy-saving related data, and to construct an energy consumption prediction model to obtain energy consumption prediction results; An energy-saving strategy module, which is configured to determine an energy-saving strategy for a base station based on an analysis result obtained by analysis and an energy consumption prediction result; The strategy execution module is configured to send the base station energy-saving strategy to the base station and dynamically adjust the base station energy-saving strategy based on real-time feedback information from the base station.

10. A base station energy-saving optimization control device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the base station energy-saving optimization control method according to any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the base station energy-saving optimization control method according to any one of claims 1 to 8 is implemented.

Images