Base station air conditioner energy consumption management and energy storage method and system

Abstract

The invention discloses a base station air conditioner energy consumption management energy storage method and system, and the method comprises the steps: collecting multi-source data in real time according to the load data of a temperature sensor, a humidity sensor, an air conditioner operation state monitoring device and a communication device in a base station room, and generating a multi-source fusion data set; for the multi-source fusion data set, dynamic refrigeration requirements in a future time window are predicted, and a dynamic optimization model of air conditioner energy consumption is constructed through an energy consumption efficiency curve; according to the prediction result of the dynamic refrigeration demand and the energy consumption optimization model, the operation mode, the wind speed and the temperature set value of the air conditioner are dynamically adjusted, the energy storage scheduling strategy is optimized, and a combined optimization scheme of air conditioner operation and energy storage scheduling is generated; and monitoring the operation state of the air conditioner in real time, dynamically adjusting the joint optimization scheme, and generating a final base station air conditioner energy consumption management scheme. By utilizing the embodiment of the invention, the energy consumption management of the air conditioner has real-time performance and accuracy, and more efficient energy consumption control and energy storage scheduling optimization can be realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy storage, and in particular to a method and system for managing energy consumption of air conditioners in base stations. Background Art

[0002] With the rapid development of wireless communication technology, the surge in the number of base stations has also increased the demand for power resources. The cooling system required by the base station during its operation, especially the energy consumption of air conditioners, accounts for an important part of the base station's power consumption. The high energy consumption of base station air conditioners not only increases operating costs, but may also cause base stations to malfunction due to overheating, thereby affecting the stability and reliability of communication services. Therefore, how to effectively manage and optimize the energy consumption of base station air conditioners has become a technical problem that needs to be solved urgently in the current communications industry.

[0003] In traditional base station energy management, usually only a single data source is relied on, such as the real-time data of the temperature and humidity sensors of the air conditioner. This method cannot fully reflect the operating conditions within the base station. At the same time, the real-time load changes of the base station communication equipment and the temperature fluctuations of the external environment are not fully considered, resulting in inaccurate energy consumption prediction and difficulty in effectively adjusting the air conditioner operation strategy. Existing energy consumption management methods mostly rely on static models and lack dynamic response capabilities, resulting in the cooling demand of the air conditioner not being met in time under high load or extreme climate conditions, thereby increasing the excessive operation and energy consumption of the air conditioner. Summary of the invention

[0004] The purpose of the present invention is to provide a base station air conditioning energy consumption management energy storage method and system to solve the deficiencies in the prior art, so that the air conditioning energy consumption management can not only be real-time and accurate, but also can achieve more efficient energy consumption control and energy storage scheduling optimization through intelligent means.

[0005] An embodiment of the present application provides a base station air conditioning energy consumption management and energy storage method, the method comprising: According to the temperature sensors, humidity sensors, air conditioning operation status monitoring equipment and communication equipment load data in the base station room, the edge computing-based data collection and preprocessing method is adopted to collect multi-source data in real time, eliminate environmental noise and electromagnetic interference through adaptive filtering algorithm, and generate a multi-source fusion data set; For multi-source fusion data sets, a prediction model based on lightweight spatiotemporal neural network is used to predict the dynamic cooling demand in the future time window by combining the real-time load changes of base station communication equipment and the external ambient temperature fluctuations. A dynamic optimization model of air conditioning energy consumption is constructed through the energy efficiency curve. According to the prediction results of dynamic cooling demand and the energy consumption optimization model, an air conditioner operation strategy optimization algorithm based on reinforcement learning is adopted to dynamically adjust the operation mode, wind speed and temperature setting value of the air conditioner. At the same time, combined with the charge and discharge state of the energy storage unit, the energy storage scheduling strategy is optimized to generate a joint optimization plan for air conditioner operation and energy storage scheduling; The operation state of the air conditioner is monitored in real time. A fault warning model based on an anomaly detection algorithm is used to identify potential faults of the air conditioner. Through an adaptive control mechanism, the joint optimization plan is dynamically adjusted to ensure the operation stability and energy consumption efficiency of the air conditioner, and a final energy consumption management plan for the base station air conditioner is generated.

[0006] Optionally, based on the temperature sensor, humidity sensor, air conditioner operation state monitoring device and communication device load data in the base station computer room, a data acquisition and preprocessing method based on edge computing is adopted to collect multi-source data in real time. Through an adaptive filtering algorithm, environmental noise and electromagnetic interference are eliminated to generate a multi-source fusion data set, including: Based on the temperature sensor, humidity sensor, air conditioner operation state monitoring device and communication device load data in the base station computer room, an edge computing-based data acquisition framework is adopted to obtain multi-source data in real time through distributed data acquisition nodes. Each acquisition node is equipped with a lightweight data caching mechanism to ensure the real-time and continuous data acquisition; For the collected multi-source data, an adaptive filtering algorithm based on wavelet transform is used to separate the environmental noise and electromagnetic interference in the signal. Through a dynamic threshold adjustment mechanism, differential filtering processing is performed according to the noise characteristics of different sensor data to obtain a multi-source data set after preliminary denoising; For the denoised multi-source data set, a time alignment algorithm based on dynamic time warping is used to eliminate the timestamp difference of different sensor data; at the same time, a missing value filling method based on spatio-temporal correlation is used to interpolate and fill the missing data part to generate a complete data set with time synchronization; For the complete data set with time synchronization, a feature fusion method based on a multi-head attention mechanism is used to extract multi-dimensional features of temperature, humidity, air conditioner operation state and communication device load data. Through a feature weight adaptive allocation mechanism, a high-precision multi-source fusion data set is generated.

[0007] Optionally, for the multi-source fusion data set, a prediction model based on a lightweight spatio-temporal neural network is adopted. Combining the real-time load changes of the base station communication equipment and the external environmental temperature fluctuations, the dynamic cooling demand in the future time window is predicted, and a dynamic optimization model of air conditioner energy consumption is constructed through an energy consumption efficiency curve, including: For the multi-source fusion dataset, a prediction model based on a lightweight spatio-temporal neural network is adopted. The temporal features of the load change of the base station communication equipment are extracted through the temporal convolutional layer, and the local spatial features of the temperature distribution in the computer room are captured through the spatial graph convolutional layer to generate spatio-temporal feature representations; Combined with the external environmental temperature fluctuation data, an external environment fusion technology based on the attention mechanism is adopted to perform weighted fusion of the external temperature change features and the spatio-temporal feature representations, enhancing the prediction ability of the prediction model for dynamic cooling demand and generating environment-enhanced spatio-temporal feature representations; The environment-enhanced spatio-temporal feature representations are input into the output layer of the prediction model, and a time series decoder based on multi-step prediction is adopted to predict the dynamic cooling demand of the base station computer room within the future time window, generating the cooling demand prediction results; According to the historical operation data of the air conditioner, a method for fitting the energy consumption efficiency curve based on piecewise linear regression is adopted to construct a relationship model between the air conditioner energy consumption, the cooling capacity, and the operation mode, and combined with the cooling demand prediction results, a dynamic optimization model of the air conditioner energy consumption is generated.

[0008] Optionally, according to the prediction results of the dynamic cooling demand and the energy consumption optimization model, an air conditioner operation strategy optimization algorithm based on reinforcement learning is adopted to dynamically adjust the operation mode, wind speed, and temperature setting value of the air conditioner. At the same time, combined with the charge and discharge state of the energy storage unit, the energy storage scheduling strategy is optimized to generate a joint optimization plan for air conditioner operation and energy storage scheduling, including: According to the prediction results of the dynamic cooling demand and the energy consumption optimization model, an air conditioner operation strategy optimization algorithm based on deep reinforcement learning is adopted to construct a reward function with the goal of minimizing energy consumption and meeting the cooling demand. Through the double Q-network structure, the operation mode, wind speed, and temperature setting value of the air conditioner are dynamically adjusted to generate a preliminary air conditioner operation optimization strategy; Combined with the current state of charge, charge and discharge efficiency curve, and grid electricity price fluctuation data of the energy storage unit, a multi-dimensional constraint model of the energy storage state is constructed. Through the energy conservation constraint, charge and discharge rate constraint, and grid load balance constraint, the feasible solution space of the energy storage scheduling is defined; For the energy storage state constraint model, an energy storage scheduling method based on the distributed constraint optimization algorithm is adopted. Combined with the energy consumption demand in the air conditioner operation optimization strategy, the charge and discharge plan of the energy storage unit is dynamically adjusted. Through the time-segmented electricity price sensitivity analysis, the energy storage scheduling strategy is optimized to generate a preliminary energy storage scheduling plan; The air conditioner operation optimization strategy and the energy storage scheduling plan are co-optimized. An joint optimization framework based on game theory is adopted. Through the dynamic game between the air conditioner and the energy storage system, the energy consumption cost and the satisfaction of the cooling demand are balanced to generate a joint optimization plan for air conditioner operation and energy storage scheduling.

[0009] Optionally, for real-time monitoring of the air conditioner operation status, a fault warning model based on an anomaly detection algorithm is adopted to identify potential air conditioner faults. Through an adaptive control mechanism, the joint optimization scheme is dynamically adjusted to ensure the operation stability and energy consumption efficiency of the air conditioner, and a final base station air conditioner energy consumption management scheme is generated, including: Real-time monitor the operation status of the air conditioner, collect the operation parameters and energy consumption data of the air conditioner, and through the edge computing node, process the collected data in real time to generate a time series data set of the air conditioner operation status; For the time series data set, adopt an anomaly detection algorithm combining the isolation forest and the long short-term memory network to identify potential air conditioner faults, and generate a fault warning signal through a dynamic threshold adjustment mechanism; According to the fault warning signal, adopt an adaptive control mechanism based on fuzzy logic to dynamically adjust the joint optimization scheme, and through a multi-objective optimization model, balance the fault recovery cost and the energy consumption efficiency to generate an adaptive control strategy; Integrate the adaptive control strategy with the joint optimization scheme, adopt a dynamic adjustment method based on a feedback correction mechanism, and update the air conditioner operation parameters and the energy storage scheduling plan in real time to generate a final base station air conditioner energy consumption management scheme.

[0010] Another embodiment of the present application provides a base station air conditioner energy consumption management energy storage system, and the system includes: A collection module, which is used to collect multi-source data in real time according to the temperature sensors, humidity sensors, air conditioner operation status monitoring devices and communication device load data in the base station computer room, adopt a data collection and preprocessing method based on edge computing, and eliminate environmental noise and electromagnetic interference through an adaptive filtering algorithm to generate a multi-source fusion data set; A prediction module, which is used to adopt a prediction model based on a lightweight spatio-temporal neural network for the multi-source fusion data set, combine the real-time load changes of the base station communication equipment and the external environmental temperature fluctuations to predict the dynamic cooling demand within a future time window, and construct a dynamic optimization model of the air conditioner energy consumption through an energy consumption efficiency curve; An adjustment module, which is used to adopt an air conditioner operation strategy optimization algorithm based on reinforcement learning according to the prediction result of the dynamic cooling demand and the energy consumption optimization model, dynamically adjust the operation mode, wind speed and temperature setting value of the air conditioner, and at the same time, combine the charge and discharge status of the energy storage unit to optimize the energy storage scheduling strategy and generate a joint optimization scheme for air conditioner operation and energy storage scheduling; A management module, which is used to real-time monitor the operation status of the air conditioner, adopt a fault warning model based on an anomaly detection algorithm to identify potential air conditioner faults, and through an adaptive control mechanism, dynamically adjust the joint optimization scheme to ensure the operation stability and energy consumption efficiency of the air conditioner, and generate a final base station air conditioner energy consumption management scheme.

[0011] Another embodiment of the present application provides a storage medium, in which a computer program is stored. Wherein, the computer program is configured to execute the method described in any one of the above when running.

[0012] Another embodiment of the present application provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to run the computer program to execute the method described in any one of the above.

[0013] Compared with the prior art, an energy storage method for base station air conditioner energy consumption management provided by the present invention collects multi-source data in real time according to the temperature sensor, humidity sensor, air conditioner operation status monitoring device and communication device load data in the base station computer room, and generates a multi-source fusion data set; for the multi-source fusion data set, predicts the dynamic cooling demand within a future time window, and constructs a dynamic optimization model of air conditioner energy consumption through an energy consumption efficiency curve; according to the prediction result of the dynamic cooling demand and the energy consumption optimization model, dynamically adjusts the operation mode, wind speed and temperature setting value of the air conditioner, optimizes the energy storage scheduling strategy, and generates a joint optimization plan for air conditioner operation and energy storage scheduling; real-time monitors the air conditioner operation status, dynamically adjusts the joint optimization plan, and generates a final base station air conditioner energy consumption management plan, so that the air conditioner energy consumption management can not only be real-time and accurate, but also can achieve more efficient energy consumption control and energy storage scheduling optimization through intelligent means. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a hardware structure block diagram of a computer terminal for an energy storage method for base station air conditioner energy consumption management provided by an embodiment of the present invention; Figure 2 It is a flow schematic diagram of an energy storage method for base station air conditioner energy consumption management provided by an embodiment of the present invention; Figure 3 It is a structure schematic diagram of an energy storage system for base station air conditioner energy consumption management provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0015] The embodiments described below with reference to the drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0016] An embodiment of the present invention first provides an energy storage method for base station air conditioner energy consumption management. This method can be applied to an electronic device, such as a computer terminal, specifically, such as an ordinary computer, etc.

[0017] The following takes running on a computer terminal as an example to describe it in detail. Figure 1 It is a hardware structure block diagram of a computer terminal for an energy storage method for base station air conditioner energy consumption management provided by an embodiment of the present invention. As Figure 1As shown in the figure, the computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the memory may include a non-volatile storage medium and an internal memory.

[0018] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, which when executed, can cause the processor to execute any one of the base station air conditioner energy consumption management energy storage methods.

[0019] The processor is used to provide computing and control capabilities to support the operation of the entire computer device.

[0020] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, it can cause the processor to execute any one of the base station air conditioner energy consumption management energy storage methods.

[0021] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that Figure 1 the structure shown in the figure is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0022] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0023] See Figure 2 , an embodiment of the present invention provides a base station air conditioner energy consumption management energy storage method, which may include the following steps: S201, according to the temperature sensor, humidity sensor, air conditioner operation status monitoring device, and communication device load data in the base station computer room, adopt a data acquisition and preprocessing method based on edge computing to collect multi-source data in real time, and through an adaptive filtering algorithm, eliminate environmental noise and electromagnetic interference to generate a multi-source fusion data set; This method involves collecting and preprocessing data from various sensors in the base station computer room in real time through edge computing-based data acquisition and preprocessing technologies. These sensors include temperature sensors, humidity sensors, air conditioner operation status monitoring devices, and load data of communication devices. Through an adaptive filtering algorithm, environmental noise and electromagnetic interference can be effectively eliminated, ensuring the accuracy and reliability of the collected data. The generated multi-source fusion dataset provides a good foundation for subsequent dynamic cooling demand prediction and energy consumption optimization, guaranteeing the operation efficiency of the entire system.

[0024] The combination of this data acquisition and preprocessing can monitor the environment and equipment operation status in the base station computer room in real time while ensuring the quality and effectiveness of the data. This not only improves the accuracy of energy consumption management, reduces the error risk, but also provides solid data support for intelligent management decisions. By fusing multiple data sources, the operation status of the base station can be more comprehensively grasped, thus effectively coping with various emergencies, optimizing the energy consumption management and energy storage scheduling strategies of the air conditioner, and ultimately achieving multiple goals such as energy conservation and consumption reduction and improving the operation stability of the base station.

[0025] Specifically, based on the temperature sensors, humidity sensors, air conditioner operation status monitoring devices, and communication device load data in the base station computer room, an edge computing-based data acquisition framework can be adopted to obtain multi-source data in real time through distributed data acquisition nodes. Each acquisition node is equipped with a lightweight data caching mechanism to ensure the real-time and continuous nature of data acquisition. The design of this step not only makes the data acquisition process efficient and real-time, but also ensures the continuity and integrity of the data. Through a distributed data acquisition architecture, the burden on the central processing node can be effectively reduced, while the flexibility of the entire system is improved to adapt to changing data collection requirements. The multi-source data obtained in real time can provide a timely and accurate basis for subsequent data processing and analysis, contributing to the formulation of more scientific air conditioner energy consumption management strategies.

[0026] In the specific implementation process, multiple distributed data acquisition nodes can be set up in the base station computer room. Each node is responsible for monitoring sensor data in a specific area. For example, one node can focus on the environment around the air conditioner equipment, while another may be located next to the communication device. Each node is equipped with a microprocessor and a memory module, capable of collecting the data stream output by the sensors in real time, such as temperature, humidity, the operation status of the air conditioner, and the load data of the communication device. By implementing edge computing on the node, the data can be preliminarily processed locally, reducing the latency of data transmission to the central server and improving the overall response speed.

[0027] The design of the data acquisition node also includes a lightweight data caching mechanism. This mechanism ensures that the node can continue to work and cache the latest data when the network is unstable or the central server fails. For example, when a certain node cannot connect to the server, it will temporarily store the data in memory and immediately send the data to the central system after the network is restored. This mechanism not only improves the reliability of data acquisition but also ensures seamless monitoring for a long time, enabling the environmental data in the base station computer room to be continuously and coherently recorded.

[0028] In this way, the base station can not only obtain various monitoring data in real time but also ensure the effectiveness of data acquisition under any conditions. This flexible edge-computing-based data acquisition architecture provides a solid foundation for subsequent data processing and analysis and lays an extremely important support for air-conditioning energy consumption management and energy storage scheduling.

[0029] For the multi-source data collected, an adaptive filtering algorithm based on wavelet transform is used to separate the environmental noise and electromagnetic interference in the signal. Through a dynamic threshold adjustment mechanism, differential filtering processing is carried out according to the noise characteristics of different sensor data to obtain a multi-source data set after preliminary denoising. By using the adaptive filtering algorithm of wavelet transform, the noise generated by environmental impact can be effectively removed, improving the signal-to-noise ratio of the data. This is a key link to ensure the accuracy of subsequent analysis results. The differential filtering processing can make the data of different types of sensors more reliable, enhance the quality of the entire multi-source fusion data set, and thus provide an accurate basis for subsequent data analysis and decision-making.

[0030] When implementing this step, first, the data from different sensors need to be input into the wavelet transform algorithm. Wavelet transform is an effective signal processing tool that can decompose a signal into multiple frequency levels, enabling us to clearly identify the difference between noise and useful signals. For example, the temperature sensor may be affected by air flow or electromagnetic interference, while the load data of communication equipment may fluctuate due to the interference of other electrical equipment. Through wavelet transform, the system can extract the real data characteristics from these frequency components.

[0031] Next, a dynamic threshold adjustment mechanism is introduced to perform differential processing according to different sensor characteristics. The traditional fixed threshold may not be able to adapt to the noise level under different environmental conditions. Therefore, the system will monitor the environmental noise situation in real time and continuously adjust the threshold of the filter dynamically according to the signal statistical information actually collected. For example, in a high-noise environment, the system may increase the threshold so that small fluctuations will not affect the final result, while in a relatively quiet environment, the threshold can be lowered to capture more subtle signal changes.

[0032] Finally, the multi-source data set after filtering will obtain a denoised signal set, which can be used for subsequent data analysis and dynamic prediction. The denoising result not only improves the data quality but also provides clear and reliable information for subsequent decision-making. This process ensures that the actual environmental state in the base station computer room can be accurately reflected, creating conditions for the precise implementation of air-conditioning energy consumption optimization and energy storage scheduling.

[0033] For the denoised multi-source data set, a time alignment algorithm based on dynamic time warping is used to eliminate the timestamp differences of different sensor data; at the same time, a missing value filling method based on spatio-temporal correlation is used to interpolate and fill the missing parts of the data, generating a complete time-synchronized data set. By eliminating the timestamp differences, the system can synchronously process different sensor data, making subsequent data analysis more accurate. Missing values are common in sensor data, so using spatio-temporal correlation to fill in the missing values ensures the integrity and coherence of the data set, which helps improve the training effect and prediction accuracy of the model.

[0034] In the implementation of this step, first, for the denoised data set, the dynamic time warping (DTW) algorithm is applied to eliminate the timestamp differences between different sensors. Considering that different sensors may collect data at different time frequencies, for example, the temperature sensor updates every 5 seconds, while the humidity sensor updates every 10 seconds, the DTW algorithm can reasonably match these time differences, thus creating consistency for the data of all sensors in a standard time frame. For example, the DTW algorithm will construct an optimal data trajectory for each time point to ensure that the readings of each sensor within the same time period are accurately linked.

[0035] Next, since sensor data may be missing during actual operation, a missing value filling method based on spatio-temporal correlation will be used to handle these missing parts. This method not only considers the historical data of the time series but also introduces spatial correlation, that is, the data of other adjacent sensors in the base station. For example, if the data of the temperature sensor is missing at a certain time point, the system will analyze the relevant data such as humidity and air flow at the same time point and use the interpolation method to predict the missing temperature value based on this information, thus filling in the data.

[0036] Through the processing of the above steps, the finally generated complete time-synchronized data set provides great convenience for subsequent data analysis. This data set not only has multi-dimensional features but also eliminates temporal interference, and can provide high-quality input data for the model training of dynamic cooling demand prediction and energy consumption optimization, ensuring the accuracy and reliability of subsequent analysis.

[0037] For the complete time-synchronized dataset, a feature fusion method based on the multi-head attention mechanism is adopted to extract multi-dimensional features of temperature, humidity, air-conditioning operating status, and communication equipment load data. Through the feature weight adaptive allocation mechanism, a high-precision multi-source fusion dataset is generated.

[0038] Feature fusion can comprehensively consider the influence of multiple sensors, extract more distinguishable features, and improve the accuracy and depth of the data analysis process. The introduction of the multi-head attention mechanism enables the system to focus on more relevant information, strengthens the data representation ability, and thus generates a high-precision multi-source fusion dataset, providing an accurate basis for subsequent dynamic cooling demand prediction.

[0039] In this implementation step, the system will perform feature extraction on the complete time-synchronized dataset and adopt an algorithm based on the multi-head attention mechanism for fusion. The advantage of the multi-head attention mechanism is that it can process multiple information sources in parallel, enabling better capture of the relationships between different features. For example, the system can set several attention heads to focus on temperature, humidity, and air-conditioning status respectively for in-depth analysis of these features. Each attention head independently processes the input data and generates a weighted representation of different signals in the output.

[0040] Next, the feature weight adaptive allocation mechanism will dynamically adjust according to the contribution of each feature in actual prediction. In the initial stage, the system will conduct a preliminary scoring of the features and then update the weights in real time based on the performance of the model during the training process. If the influence of temperature on the cooling demand is significant during a certain period, the weight of the temperature feature will increase, otherwise it may decrease. This flexible weight allocation mechanism ensures that the model always focuses on the most important features and maximizes the accuracy of the prediction.

[0041] The finally generated high-precision multi-source fusion dataset will be used for subsequent dynamic cooling demand prediction and energy consumption optimization models. This dataset not only comprehensively reflects the environmental conditions and equipment operating status in the base station computer room but also provides a clearer data basis for decision-making. The optimization of this process makes the subsequent air-conditioning energy consumption management and energy storage scheduling strategies more intelligent and effective, contributing to improving the overall energy efficiency and operating efficiency of the base station.

[0042] S202. For the multi-source fusion dataset, adopt a prediction model based on a lightweight spatio-temporal neural network, combine the real-time load changes of base station communication equipment and external environmental temperature fluctuations to predict the dynamic cooling demand within the future time window, and construct a dynamic optimization model for air-conditioning energy consumption through the energy consumption efficiency curve; This step aims to process the multi-source fusion dataset using a prediction model based on a lightweight spatio-temporal neural network to predict the dynamic cooling demand of the base station computer room within a future time window. This process combines the real-time load changes of base station communication equipment and external environmental temperature fluctuations. In the model, temporal features of load changes are extracted through a temporal convolutional layer, while local spatial features of the temperature distribution within the computer room are captured using a spatial graph convolutional layer, ultimately generating a spatio-temporal feature representation. Additionally, a dynamic optimization model for air conditioning energy consumption is constructed through an energy consumption efficiency curve to provide data support for the energy-saving operation of the air conditioning system.

[0043] Applying the lightweight spatio-temporal neural network to the prediction of dynamic cooling demand can significantly improve the response speed and accuracy of the energy consumption management system. By capturing and analyzing the load changes of base station communication equipment and the impact of the external environment in real time, this method helps to predict the demand for air conditioners in advance, thereby adjusting the operation strategy of air conditioners targetedly and reducing energy consumption. In addition, the establishment of the energy consumption efficiency curve provides a basis for formulating a more scientific energy consumption management plan, enabling the base station to achieve optimal energy efficiency and operation efficiency during operation, thus improving the overall service quality and stability.

[0044] Specifically, for the multi-source fusion dataset, a prediction model based on a lightweight spatio-temporal neural network can be adopted. Temporal features of the load changes of base station communication equipment are extracted through a temporal convolutional layer, and local spatial features of the temperature distribution within the computer room are captured through a spatial graph convolutional layer to generate a spatio-temporal feature representation. In the specific implementation process, it is first necessary to import the multi-source fusion dataset into the lightweight spatio-temporal neural network model. This model consists of two key parts: a temporal convolutional layer and a spatial graph convolutional layer. The working principle of the temporal convolutional layer is to perform a convolutional operation on the load data of base station communication equipment through a sliding window technique to capture the temporal features of load trends and periodic changes. For example, if the communication traffic suddenly increases during a certain period, this layer can automatically learn this change pattern through the convolutional kernel and extract the time points of load increase to form an effective feature representation.

[0045] Next, the spatial graph convolutional layer is responsible for analyzing the data of each temperature sensor within the computer room. By constructing a graph structure between the sensors, the model can capture the spatial relationships of the temperatures in each area. For example, assuming that the temperature on the left side of the computer room suddenly rises while the right side is normal, the spatial graph convolutional layer will learn this spatially non-uniform distribution feature through the connection relationships between nodes, enabling the model to understand the impact of temperature changes in local areas on the overall cooling demand.

[0046] Finally, through the joint processing of the temporal convolutional layer and the spatial graph convolutional layer, the generated spatio-temporal feature representation will fully reflect the interaction relationship between the load change and the temperature distribution. This feature representation not only provides the basic information for subsequent demand prediction but also provides the necessary data support for the policy optimization of the reinforcement learning algorithm.

[0047] Combined with the external environmental temperature fluctuation data, the external environment fusion technology based on the attention mechanism is adopted to perform weighted fusion of the external temperature change features and the spatio-temporal feature representation, enhancing the prediction ability of the prediction model for dynamic cooling demand and generating an environment-enhanced spatio-temporal feature representation. In this step, the system first needs to obtain the temperature fluctuation data of the external environment in real time. These external data include weather forecasts, real-time monitoring data from meteorological stations, etc., which have an important impact on the cooling demand in the computer room. Through the external environment fusion technology based on the attention mechanism, the external temperature change features can be effectively combined with the previously generated spatio-temporal feature representation. In this process, the attention mechanism will assign different weights to each feature to highlight the more important factors.

[0048] For example, assume that during a certain period, the external temperature rises rapidly due to climate change. The model will increase the weight of the external temperature feature through the attention mechanism, making the prediction model more sensitive to this change when judging the cooling demand. This dynamic weight assignment ensures that the model can respond to the changes in the external environment in real time and makes the prediction more accurate, thus effectively improving the energy efficiency of the air conditioner.

[0049] Through this weighted fusion process, the generated environment-enhanced spatio-temporal feature representation will accurately reflect the environmental changes inside and outside the computer room. This enhanced feature not only improves the sensitivity of the model to the changes in cooling demand but also provides richer data support for subsequent demand prediction and policy optimization, ensuring that the system can operate efficiently under different weather conditions.

[0050] Input the environment-enhanced spatio-temporal feature representation into the output layer of the prediction model, and adopt a time series decoder based on multi-step prediction to predict the dynamic cooling demand of the base station computer room within the future time window, generating a cooling demand prediction result. When implementing this step, the environment-enhanced spatio-temporal feature representation will be fed into the output layer of the prediction model as input. The decoder adopted here is a time series model based on multi-step prediction, and its structure allows the model to perform joint prediction on multiple future time points. This design enables the system to not only generate a prediction result for the next time point but also evaluate the changes in cooling demand within multiple time points simultaneously.

[0051] The system will automatically generate a future prediction of the cooling demand based on the learned spatio-temporal feature representation. For example, during the prediction process, if the model identifies that both the load and temperature were rising in the previous time period, then it will infer based on this trend that the cooling demand may continue to increase in the next few time periods. Such an ability enables the base station to prepare in advance and thus effectively adjust the operating mode of the air conditioner.

[0052] Finally, the prediction results of the cooling demand generated by the time series decoder will be fed back to the energy consumption management system. These results not only guide the operating strategy of the air conditioner but also help formulate more reasonable energy consumption management measures to ensure the stability and energy efficiency of the base station operation and reduce unnecessary energy waste.

[0053] According to the historical operation data of the air conditioner, a method of fitting the energy consumption efficiency curve based on piecewise linear regression is adopted to construct a relationship model between the air conditioner energy consumption, cooling capacity, and operating mode. Combining with the prediction results of the cooling demand, a dynamic optimization model of the air conditioner energy consumption is generated.

[0054] In this step, the system first needs to accumulate the historical operation data of the air conditioner, including the operating mode (such as cooling, dehumidification, etc.), cooling capacity, and the corresponding energy consumption. These data will be used for fitting the energy consumption efficiency curve based on piecewise linear regression. By analyzing the energy consumption data under different operating modes, the system can identify the energy consumption laws under each mode, thus providing a basis for subsequent optimization decisions.

[0055] For example, assume that there are significant differences in the energy consumption of the air conditioner under normal cooling and high-load cooling conditions. The system will construct different linear regression models for these two modes respectively, identify their energy efficiency ratios by analyzing historical data, and incorporate external influencing factors such as weather changes into the models. Such analysis can help the system clarify the complex relationship between energy consumption and cooling demand.

[0056] Finally, combining the prediction results of the cooling demand, the system will generate a dynamic optimization model of the air conditioner energy consumption. This model will be updated in real time, automatically adjusting the operating mode and energy output of the air conditioner according to the predicted cooling demand, so as to minimize energy consumption while ensuring that the cooling demand is met, and achieve efficient energy consumption management. This dynamic optimization model provides a scientific and effective solution for the energy-saving operation of the base station.

[0057] S203. According to the prediction results of the dynamic cooling demand and the energy consumption optimization model, adopt an air conditioner operating strategy optimization algorithm based on reinforcement learning to dynamically adjust the operating mode, wind speed, and temperature setting value of the air conditioner. At the same time, combining the charge and discharge status of the energy storage unit, optimize the energy storage scheduling strategy to generate a joint optimization plan for air conditioner operation and energy storage scheduling; Based on the prediction results of dynamic cooling demand and the energy consumption optimization model, this step utilizes an air conditioner operation strategy optimization algorithm based on reinforcement learning to dynamically adjust the operation mode, wind speed, and temperature setpoint of the air conditioner. Meanwhile, in combination with the charge and discharge status of the energy storage unit, the energy storage scheduling strategy is optimized to generate a joint optimization plan for air conditioner operation and energy storage scheduling. The reinforcement learning algorithm continuously learns and adjusts the strategy through environmental feedback, enabling the air conditioner to achieve the optimal operation mode in the face of real-time changing load demands and external environmental conditions. This optimization process not only considers the operation efficiency of the air conditioner itself but also, based on the status of the energy storage unit, enables the system to find the best balance between energy consumption and cooling effect.

[0058] Adopting an optimization algorithm based on reinforcement learning can significantly improve the intelligent level of the air conditioner system, enabling the air conditioner to dynamically adjust the operation strategy according to real-time data. The core of this method lies in using historical data and real-time status for learning and decision-making, which not only improves the energy efficiency of the air conditioner but also enhances the system's response ability to sudden demand changes. This flexible intelligent adjustment enables the base station to reduce energy consumption while ensuring the cooling demand, improve economic benefits, and is environmentally friendly, contributing to the achievement of sustainable development goals.

[0059] Specifically, according to the prediction results of dynamic cooling demand and the energy consumption optimization model, an air conditioner operation strategy optimization algorithm based on deep reinforcement learning can be adopted to construct a reward function with the goals of minimizing energy consumption and meeting cooling demand. Through a double Q-network structure, dynamically adjust the operation mode, wind speed, and temperature setpoint of the air conditioner to generate a preliminary air conditioner operation optimization strategy; In this step, the system first predicts dynamic cooling demand and energy consumption information by collecting historical data and real-time monitored data and using machine learning techniques. Then, an air conditioner operation strategy optimization algorithm based on deep reinforcement learning is constructed. The core of this algorithm lies in constructing a reward function that takes into account both the goals of minimizing energy consumption and meeting cooling demand. By using a double Q-network structure, the system can quickly learn the optimal action strategy in different working states, that is, how to reasonably adjust the operation mode, wind speed, and temperature setpoint of the air conditioner to minimize energy consumption while meeting the cooling demand.

[0060] This strategy optimization process not only makes the air conditioner system more intelligent but also enables dynamic adjustment based on real-time data, thereby effectively improving energy efficiency. Through deep reinforcement learning, the system can respond more flexibly to environmental changes and load fluctuations than traditional control methods, thus reducing energy consumption costs while ensuring indoor comfort. This intelligent operation strategy lays the foundation for subsequent energy storage scheduling and comprehensive energy consumption management.

[0061] In this step, the system will first establish a dynamic cooling demand prediction model. This model is based on data such as historical temperature, humidity, and air conditioner usage habits, and integrates machine learning algorithms to achieve real-time prediction of cooling demand. When the predicted ambient temperature is higher than 30°C, the model will predict an increase in the user's demand for air conditioner cooling, and at the same time consider the user's living habits, such as the change in the demand for comfort at different times of the day and night. This prediction result will be used as the input in the deep reinforcement learning algorithm.

[0062] Next, the system will use a deep reinforcement learning-based method to construct a reward function. The design of the reward function will consider both minimizing energy consumption and meeting the cooling demand, which can guide the model to continuously optimize the operation strategy of the air conditioner during the learning process. For example, if the system successfully reduces energy consumption and meets the user's cooling demand, it will receive a higher reward; conversely, if the energy consumption is too high or the cooling demand is not met, the reward will decrease. Through continuous iteration, the model will gradually learn the optimal strategy and then adjust the operation state of the air conditioner under different climate conditions.

[0063] Finally, a double Q-network structure will be used to dynamically adjust the operation mode, wind speed, and temperature setting value of the air conditioner. This structure alternates learning through two Q-networks, which can effectively reduce the variance in the policy update process and improve the stability and effectiveness of learning. For example, when the indoor temperature is higher than the set value, the system can intelligently increase the wind speed and lower the temperature setting, and generate a preliminary air conditioner operation optimization strategy by optimizing the operation mode of the air conditioner. This strategy will provide the necessary data basis for the energy storage scheduling in the subsequent steps.

[0064] Combined with the current state of charge, charge-discharge efficiency curve, and grid electricity price fluctuation data of the energy storage unit, a multi-dimensional constraint model of the energy storage state is constructed. Through energy conservation constraints, charge-discharge rate constraints, and grid load balance constraints, the feasible solution space of energy storage scheduling is defined; In this stage, the main task of the system is to establish a multi-dimensional constraint model in combination with the actual situation of the energy storage unit. This model will consider the current state of charge of the energy storage unit, the charge-discharge efficiency curve, and the fluctuation of the grid electricity price. Specifically, the energy conservation constraint ensures that the charge-discharge process of the energy storage unit conforms to the law of energy transfer, that is, the relationship between the charging power and the discharging power; the charge-discharge rate constraint limits the rate of the energy storage device during the charge-discharge process to ensure operation within a safe and stable range; the grid load balance constraint takes into account the load changes of the grid during the entire scheduling period to balance power supply and demand.

[0065] By constructing such a constraint model, the system can effectively define the feasible solution space of energy storage scheduling and provide a calculation basis for the subsequent scheduling strategy formulation. This constraint model that considers multi-dimensional factors ensures the scientificity and rationality of energy storage scheduling, so that under different electricity prices and climate conditions, the energy storage system can play its maximum efficiency and ensure the stability and economy of power supply.

[0066] At this stage, the system will integrate the state of charge and charge and discharge efficiency curves of the energy storage unit to establish a multi-dimensional constraint model for the energy storage state. The current state of charge can be monitored in real time by sensors to ensure that the system always accurately grasps the power of the energy storage unit. For example, if the current state of charge of the energy storage unit is 80%, the maximum charge and discharge limits must be considered when formulating the charge and discharge plan. At the same time, the charge and discharge efficiency curve will be used to evaluate the portion of power conversion lost under specific charging and discharging conditions to ensure that the optimized plan is practical.

[0067] Next, the energy conservation constraint requires that the total energy input and output must remain equal during the charging and discharging process of the energy storage unit. This means that if the energy storage unit discharges 100kWh of electricity at a certain moment, then at a certain subsequent moment, the difference between the charge and discharge amounts must be equal to the sum of the initial electricity. In addition, the charge and discharge rate constraints ensure that the energy storage system does not exceed its designed maximum capacity during charging and discharging. For example, discharge of more than 50kWh is not allowed in a short period of time to ensure the stability and safety of the system.

[0068] Finally, the grid load balance constraint ensures that the load of the grid is reasonably transferred within a given time, that is, it does not exceed the grid's carrying capacity during high-load periods. For example, during peak hours, the system should reasonably control the discharge amount to avoid excessive pressure on the grid. Through the construction of these comprehensive constraints, the system defines the feasible solution space for energy storage scheduling, providing the necessary theoretical basis for the subsequent optimization plan of energy storage scheduling.

[0069] For the energy storage state constraint model, a distributed constraint optimization algorithm-based energy storage scheduling method is used. Combined with the energy consumption demand in the air conditioning operation optimization strategy, the charging and discharging plan of the energy storage unit is dynamically adjusted. Through the time-segmented electricity price sensitivity analysis, the energy storage scheduling strategy is optimized to generate a preliminary energy storage scheduling plan. In this step, the system will adopt a distributed constraint optimization algorithm to conduct energy storage scheduling based on the energy storage state constraint model established in the previous step. This method can effectively integrate the energy consumption requirements in the air-conditioning operation optimization strategy and dynamically adjust the charge and discharge plans of the energy storage unit according to electricity price changes and load demands. The system will analyze the electricity price signals according to time period division to make them an important reference basis in the decision-making process. Through electricity price sensitivity analysis, the system can identify the optimal timing for charging and discharging in different time periods, so as to formulate a suitable scheduling plan.

[0070] By dynamically adjusting the charge and discharge plans of the energy storage unit, the system can better cope with electricity price fluctuations and load changes, and improve the utilization efficiency of the energy storage system. This can not only reduce the electricity cost during high electricity price periods, but also ensure the stable operation of the system during power shortage. In addition, optimizing the energy storage scheduling strategy will enhance the flexibility and elasticity of the overall power management, ensuring the coexistence of sustainability and economic benefits.

[0071] At this stage, the system will adopt a distributed constraint optimization algorithm to schedule the energy storage state constraint model constructed in the previous step. The advantage of this optimization algorithm is that it can run in parallel on multiple processing nodes, improving the calculation efficiency. The system will collect power demand data and electricity price information in different time periods, and by converting these data into the input of the model, it will optimize the charge and discharge plans of the energy storage unit in real time. For example, if the system detects that the electricity price in a certain time period is at a peak, the distributed optimization algorithm will give priority to arranging the energy storage unit to discharge to reduce the overall electricity bill.

[0072] Combined with the energy consumption requirements in the air-conditioning operation optimization strategy, the energy storage scheduling method will be dynamically adjusted according to the expected energy consumption of the air-conditioning system in different time periods. For example, when the energy consumption demand of the air-conditioning is large during the day's peak period, the system may arrange the energy storage unit to discharge to meet the air-conditioning operation demand and avoid overloading the power grid. At this time, the system will continuously evaluate the charge and discharge strategy to ensure that the maximum power can be effectively utilized when needed.

[0073] Furthermore, the system will conduct time-segmented electricity price sensitivity analysis to optimize the energy storage scheduling strategy. By analyzing the electricity price changes in each time period, the system can identify when the charging cost is the lowest and when the discharge return is the highest, so as to formulate the optimal charge and discharge plan. For example, if the electricity price is low late at night, the system may arrange the energy storage unit to charge; while during the day's electricity price peak period, it will be adjusted to the discharge state. Through the comprehensive application of these strategies, the system finally generates a preliminary energy storage scheduling plan, laying a foundation for subsequent joint optimization.

[0074] Co - optimize the air - conditioner operation optimization strategy and the energy - storage scheduling scheme. Adopt a joint optimization framework based on game theory. Through the dynamic game between the air - conditioner and the energy - storage system, balance the energy - consumption cost and the satisfaction degree of the cooling demand, and generate a joint optimization scheme for air - conditioner operation and energy - storage scheduling.

[0075] In this step, the system integrates the operation optimization strategy of the air - conditioner and the scheduling scheme of the energy - storage, and conducts co - optimization using the framework of game theory. Game theory enables each system to dynamically adjust its decision according to the strategies of the other by defining the strategic interactions between different participants (the air - conditioner system and the energy - storage system). In this way, the system can achieve the best balance between the energy - consumption cost and the satisfaction degree of the cooling demand. For example, the air - conditioner system may be advised to reduce the load during high - electricity - price periods, while the energy - storage system will discharge appropriately according to the load - balancing requirements to take into account the needs of both systems.

[0076] This joint optimization scheme helps to improve the overall economic efficiency and resource utilization efficiency of the system. Through dynamic games, both sides make optimal decisions while influencing each other, ultimately achieving an effective reduction in the energy - consumption cost and sufficient satisfaction of the cooling demand. The advantage of this strategy is that the system can not only respond to the instantaneous change in electricity prices but also adapt to the operating requirements under different load conditions, thus enhancing the overall reliability and stability of the operation.

[0077] In this step, the system co - optimizes the air - conditioner operation optimization strategy and the energy - storage scheduling scheme, and establishes a dynamic game model using the framework of game theory. This game model regards the air - conditioner system and the energy - storage system as two "players", each optimizing its strategy to minimize the energy - consumption cost and maximize the satisfaction of the cooling demand. The decisions of both sides will influence each other. The operating state of the air - conditioner depends on the charge - discharge strategy of the energy - storage, and the scheduling of the energy - storage also needs to consider the change in the electricity demand of the air - conditioner.

[0078] The system will achieve dynamic adjustment of the strategy through multiple rounds of games. For example, during a certain period, if the electricity price surges, the air - conditioner system can choose to reduce the load to reduce the electricity bill, while the energy - storage system may discharge at this time to maximize the repayment of the grid load. This process is repeated. After each round of the game, the system will update the strategy of each participant according to the game result to improve the overall operating efficiency and economic benefits.

[0079] Finally, the system will integrate the game information of all parties to generate a joint optimization plan to ensure the coordinated operation of the air conditioner and the scheduling of energy storage. For example, when the electricity price remains high, the system may choose to reduce the air-conditioning load to lower the frequency of energy storage discharge, thereby extending the service life of the energy storage unit; while during periods of low electricity prices, the system can moderately increase air-conditioning cooling to improve user comfort. This flexible joint scheduling strategy achieves a good balance between energy consumption costs and cooling demand, providing a guarantee for the sustainability of electricity use.

[0080] S204, monitor the operating status of the air conditioner in real time, adopt a fault warning model based on an anomaly detection algorithm to identify potential air conditioner faults, and through an adaptive control mechanism, dynamically adjust the joint optimization plan to ensure the operating stability and energy consumption efficiency of the air conditioner, and generate the final base station air conditioner energy consumption management plan.

[0081] Real-time monitoring of the operating status of the air conditioner is a key link to ensure the efficient and stable operation of the base station air conditioner system. By integrating data from different sensors (such as temperature, humidity, energy consumption, and operating status, etc.), the system can track the operating conditions of the air conditioner in real time. Adopting a fault warning model based on an anomaly detection algorithm, the system can effectively identify potential faults. For example, when it is detected that the energy consumption of the air conditioner increases unusually within a short period of time, or the readings of some sensors deviate from the normal range, the system immediately generates a fault warning signal. On this basis, through an adaptive control mechanism, the system will dynamically adjust the joint optimization plan. Specifically, the system will automatically optimize the operating parameters of the air conditioner (such as wind speed, temperature setting value, and operating mode) according to real-time monitoring data and historical performance analysis to ensure the stability and energy efficiency of the air conditioner even when a fault occurs. Finally, this process will generate a comprehensive base station air conditioner energy consumption management plan to ensure that the system maintains optimal performance under various working conditions.

[0082] Implementing the real-time monitoring and fault warning mechanism greatly enhances the reliability and adaptability of the air conditioner system. When potential faults occur, the system can respond quickly and maintain normal operation by dynamically adjusting the joint optimization plan. This mechanism can effectively reduce the risk of equipment damage and maintenance costs, avoid operation interruptions caused by faults, and thus improve the continuity and reliability of the overall service. In addition, with the help of real-time data and intelligent adjustment, the air conditioner system can maintain good energy efficiency performance under rapid load changes and adverse environmental conditions, such as high temperature or high humidity, etc., reducing energy waste. For example, it can automatically increase the cooling capacity during high-load periods while reducing power consumption when the load drops, thus ensuring user comfort and saving operation costs. Finally, this dynamically adjusted joint optimization plan not only improves the energy consumption efficiency of the base station but also lays a solid foundation for the long-term stable operation of the equipment.

[0083] Specifically, the operating status of the air conditioner can be monitored in real time, the operating parameters and energy consumption data of the air conditioner are collected, and through the edge computing node, the collected data is processed in real time to generate a time series data set of the operating status of the air conditioner. In this stage, the system aggregates data from different sensors through the edge computing node, and collects multiple operating parameters of the air conditioner in real time, such as cooling load, energy consumption, current, temperature, and humidity. Each sensor continuously monitors specific metrics and uploads data to the edge computing node at fixed time intervals. The edge node is not only responsible for data collection, but also performs preliminary data processing, such as data cleaning and aggregation, to ensure the accuracy of the information. The generated time series data set will provide basic data for subsequent fault detection and energy efficiency analysis.

[0084] This real-time monitoring mechanism ensures that the operating status of the air conditioner can be accurately reflected, and thus provides key data support for fault warning, maintenance decision-making, and energy consumption optimization. Through more timely data collection and processing, abnormal situations during operation can be responded to more quickly. For example, if a certain sensor detects a sudden increase in the energy consumption of the air conditioner, the system will immediately record and analyze the possible reasons to prevent energy waste and equipment damage. This mechanism helps improve the operating efficiency of the entire air conditioner system and provides data support for subsequent decision-making.

[0085] In this stage, it is first necessary to deploy a variety of sensors inside the base station computer room, including temperature sensors, humidity sensors, wind speed sensors, and energy consumption monitors. Each sensor processes data through the edge computing node to ensure real-time and efficiency. For example, the temperature sensor can collect the temperature change in the computer room once every 1 second, while the energy consumption monitor is responsible for recording the power consumption of the air conditioner. The data of these sensors will be sent to the edge computing node in real time through wireless or wired networks to ensure the continuity and real-time of the data.

[0086] During the process of processing data by the edge computing node, it will first preprocess the multi-source data collected, including data cleaning and denoising. To improve the data quality, the edge node will apply an adaptive filtering algorithm to eliminate electromagnetic interference and environmental noise. For example, in some cases, the temperature sensor may be affected by external noise and generate inaccurate readings. By using the adaptive filtering algorithm, these abnormal data can be identified and corrected, thus generating an accurate multi-source data set. This process is crucial because high-quality data is the basis for subsequent analysis and decision-making.

[0087] Finally, the processed data will be formed into a time - series dataset, containing accurate records of the air - conditioner's operating status and energy - consumption data. Based on these time - series data, the system will generate real - time monitoring reports to help operation and maintenance personnel evaluate the working efficiency and energy - consumption situation of the air - conditioner. These data can be used to identify operation trends and potential problems. For example, if the energy consumption significantly increases during a certain period, it means that the air - conditioner may be at risk of decreased efficiency or malfunction. In this way, the system can ensure that the air - conditioner maintains an excellent state during continuous operation and provides an important basis for subsequent fault warning and optimization control.

[0088] For the time - series dataset, an anomaly - detection algorithm combining Isolation Forest and Long Short - Term Memory network is adopted to identify potential faults of the air - conditioner, and a fault - warning signal is generated through a dynamic threshold - adjustment mechanism. In this step, the system will analyze the previously generated time - series dataset and apply an anomaly - detection technique that combines two algorithms: Isolation Forest and Long Short - Term Memory (LSTM) network. As an unsupervised learning algorithm, Isolation Forest can effectively identify outliers in the data, while the LSTM network is responsible for handling the long - term dependencies of time - series data and evaluating the temporal characteristics of the data. By combining these two algorithms, the system can more accurately detect potential fault symptoms. For example, when the energy - consumption data of the air - conditioner suddenly deviates from the normal range, the system will issue a warning in a timely manner.

[0089] The implementation of this fault - warning mechanism has significantly improved the reliability and maintenance efficiency of the air - conditioner system. By timely identifying potential faults, the risk of serious equipment failures can be effectively reduced, thereby reducing maintenance costs and extending the service life. For example, if the system detects a sudden surge in the energy consumption of the air - conditioner in a short period, it may indicate a problem with the equipment, and operation and maintenance personnel can intervene in advance to avoid further damage. This mechanism not only enhances the safety and stability of the system but also provides better services and realizes the efficient utilization of equipment.

[0090] In this step, first, the system will input the time - series dataset into an anomaly - detection model based on Isolation Forest. The Isolation Forest algorithm can effectively identify outliers that are significantly different from most data points, especially in the case of high - dimensional data. For example, when the energy - consumption data suddenly increases significantly during a specific period, the Isolation Forest algorithm will mark this point as a potential anomaly, thus triggering an alarm. This algorithm is applicable to various abnormal situations in the operation of the air - conditioner, such as equipment failures and sensor - reading errors, ensuring the discovery of potential problems at an early stage.

[0091] Next, in combination with the Long Short-Term Memory network (LSTM), the system will utilize the memory units of the LSTM to analyze the trends and periodic changes in time series data. LSTM is particularly suitable for processing continuous time series data and can capture long-term dependencies. For example, by modeling the temperature and energy consumption changes in the past few hours, LSTM can predict the normal operating level in the current environment. If it is detected that the actual data deviates from the range predicted by the LSTM model, the system will trigger a fault warning signal to notify the operation and maintenance personnel.

[0092] To improve the detection accuracy, the system implements a dynamic threshold adjustment mechanism. According to the changes in real-time monitoring data and environmental conditions, the system will automatically update the alarm threshold. For example, in the summer when the external temperature is extremely high, the energy consumption of the air conditioning system will naturally increase. If a fixed threshold is still used for fault detection at this time, it may lead to false alarms. Therefore, the dynamic threshold adjustment mechanism allows the system to automatically adjust the warning standard according to the changes in the operating environment to ensure the accuracy and sensitivity of fault detection.

[0093] Based on the fault warning signal, an adaptive control mechanism based on fuzzy logic is adopted to dynamically adjust the joint optimization plan. Through a multi-objective optimization model, the system balances the fault recovery cost and energy efficiency to generate an adaptive control strategy. In this step, the system will apply the fuzzy logic control mechanism to dynamically adjust the operating parameters of the air conditioner based on the fault warning signal generated in the previous step. The fuzzy logic controller can handle uncertainties and ambiguities and adjust the operating state of the air conditioner in real time by defining a series of fuzzy rules. For example, if the fault warning signal indicates that the energy consumption of the air conditioner has increased abnormally, the fuzzy logic controller can achieve adaptive adjustment by setting the rule "if the energy consumption is high and the temperature target is not achieved, then reduce the wind speed". This intelligent control mechanism can quickly take corresponding measures when a fault occurs, such as reducing the cooling capacity of the air conditioner, adjusting the temperature set point, or switching to the standby system when necessary, to ensure the stable operation and energy efficiency optimization of the air conditioner. The system can continuously update and optimize the control rules based on real-time data and historical performance to adapt to different operating environments and load requirements, ensuring optimal performance in all situations.

[0094] By implementing the adaptive control mechanism based on fuzzy logic, the system not only improves the response ability of the air conditioner in the face of potential faults but also significantly enhances the operating flexibility and energy efficiency. When the air conditioner is under heavy load or the environmental conditions change drastically, the system can quickly make adjustments to ensure that the equipment will not be damaged due to overload. For example, on extremely hot days, the air conditioner needs to maintain a high cooling efficiency. At this time, the system can automatically optimize the operating conditions so that the air conditioner can still maintain reasonable energy consumption under high load. This not only reduces energy waste but also helps to ensure the long-term stable use of the equipment.

[0095] At this stage, the system will activate an adaptive control mechanism based on fuzzy logic according to the fault warning signals generated in the previous step. Once a potential fault is detected, the system will analyze the type and severity of the fault through a fuzzy logic controller and adjust the operating mode of the air conditioner in real time. For example, if a certain sensor feedbacks that the cooling capacity of the air conditioner is insufficient, the fuzzy logic controller may recommend immediately increasing the temperature setting value or changing the wind speed to reduce the load on the equipment and avoid damage.

[0096] Meanwhile, the system will incorporate a joint optimization scheme, involving dynamic adjustments to the operating mode, wind speed, and temperature setting of the air conditioner. These adjustments are not only based on real-time fault warning signals but also take into account energy consumption efficiency. For example, when the cost of restoring the air conditioner's fault is higher than the energy consumption savings, the system may choose to continue using the current operating mode instead of forcing a repair. Through such decisions, the system achieves a trade-off between the cost of fault recovery and energy consumption efficiency, ensuring that the air conditioner can still maintain a relatively stable operating state in case of a fault.

[0097] When implementing the adaptive control strategy, the system will adopt a multi-objective optimization model, comprehensively considering various constraints and performance indicators. While ensuring the normal operation of the air conditioner, the system will strive to reduce energy consumption and maintain the equipment's lifespan. For example, if the air conditioner needs maintenance, the system will recommend charging and operating during periods of low electricity prices to maximize the reduction of maintenance costs and energy consumption. At the same time, this control strategy will be continuously updated to adapt to future operating environments and changes in equipment status, forming a self-optimizing dynamic process.

[0098] Integrate the adaptive control strategy with the joint optimization scheme, adopt a dynamic adjustment method based on a feedback correction mechanism, update the air conditioner operating parameters and energy storage scheduling plan in real time, and generate the final base station air conditioner energy consumption management plan.

[0099] At this stage, the system will integrate the adaptive control strategy with the previously formed joint optimization scheme to form a more intelligent and flexible management system. By introducing a feedback correction mechanism, the system can track the actual operating effect of the air conditioner and the charging and discharging performance of the energy storage unit, obtain performance data in real time and compare it with the preset targets. When the actual energy consumption or cooling effect deviates from the ideal state, the system will dynamically adjust the operating parameters and energy storage scheduling plan according to the feedback information. For example, if it is monitored that the energy consumption of the air conditioner during peak hours exceeds the expected target, the system can adjust the operating mode or lower the set temperature to optimize energy efficiency. Finally, the integrated scheme will form a comprehensive base station air conditioner energy consumption management plan to ensure that the system can maintain the best energy efficiency and stability under different working conditions and loads.

[0100] This integration step is crucial for enhancing the system's adaptability. In a rapidly changing environment and under varying load conditions, it can respond promptly and make adjustments to effectively avoid the risks of equipment overload or excessive energy consumption. For example, when the external environmental temperature suddenly rises, the system can timely adjust the operation strategy of the air conditioner to cope with the increased load, while ensuring the reasonable use of the energy storage unit to avoid overcharging and over-discharging. Through the feedback correction mechanism, the system continuously learns and optimizes, forming a self-improving management process, which not only reduces the complexity of operation and maintenance, but also improves the operation efficiency of the entire base station, ensures the minimization of energy consumption and the maximization of comfort, and provides strong support for the economy and sustainability of base station operation.

[0101] In this stage, the system will integrate between the adaptive control strategy and the joint optimization scheme to form a comprehensive air conditioner energy consumption management plan. The system will implement a feedback correction mechanism to continuously monitor the actual operating status of the air conditioner and compare it with the output of the prediction model to evaluate the effectiveness of the adaptive control strategy in real time. For example, if the actual energy consumption is significantly higher than expected, the system will automatically adjust the control strategy to match the actual conditions.

[0102] The integrated plan will update the operating parameters of the air conditioner, such as temperature settings and wind speeds, according to real-time data. At the same time, the energy storage scheduling plan will also be optimized during this process. By continuously monitoring the charge and discharge status of the energy storage unit, the system can make informed decisions to adjust the utilization of energy, such as charging during periods of low electricity prices and releasing energy from the energy storage unit for air conditioner use during peak electricity consumption periods. Such a strategy ensures that while meeting the cooling demand, the overall energy consumption cost is reduced.

[0103] The finally generated plan not only meets the immediate operating requirements but also provides data support for future energy efficiency improvement and equipment maintenance. The system can continuously optimize the control strategy based on historical operating data and real-time feedback to form a closed-loop management. This approach not only ensures the operating stability of the air conditioner system but also improves the overall energy consumption efficiency, thus providing a continuous and stable energy management plan for the base station to ensure the efficient and reliable operation of the equipment.

[0104] It can be seen that, according to the temperature sensors, humidity sensors, air conditioner operation status monitoring devices and communication equipment load data in the base station machine room, multi-source data is collected in real time to generate a multi-source fusion data set; for the multi-source fusion data set, the dynamic cooling demand within a future time window is predicted, and a dynamic optimization model of air conditioner energy consumption is constructed through the energy consumption efficiency curve; according to the prediction results of the dynamic cooling demand and the energy consumption optimization model, the operation mode, wind speed and temperature setting value of the air conditioner are dynamically adjusted, the energy storage scheduling strategy is optimized, and a joint optimization plan for air conditioner operation and energy storage scheduling is generated; the operation status of the air conditioner is monitored in real time, the joint optimization plan is dynamically adjusted, and a final base station air conditioner energy consumption management plan is generated, so that the air conditioner energy consumption management not only has real-time performance and accuracy, but also can achieve more efficient energy consumption control and energy storage scheduling optimization through intelligent means.

[0105] Another embodiment of the present invention provides a base station air conditioner energy consumption management energy storage system. Refer to Figure 3 , the system may include: A collection module 301, configured to collect multi-source data in real time according to the temperature sensors, humidity sensors, air conditioner operation status monitoring devices and communication equipment load data in the base station machine room, adopt a data collection and preprocessing method based on edge computing, and eliminate environmental noise and electromagnetic interference through an adaptive filtering algorithm to generate a multi-source fusion data set; A prediction module 302, configured to adopt a prediction model based on a lightweight spatio-temporal neural network for the multi-source fusion data set, combine the real-time load changes of the base station communication equipment and the external environmental temperature fluctuations, predict the dynamic cooling demand within a future time window, and construct a dynamic optimization model of air conditioner energy consumption through the energy consumption efficiency curve; An adjustment module 303, configured to dynamically adjust the operation mode, wind speed and temperature setting value of the air conditioner according to the prediction results of the dynamic cooling demand and the energy consumption optimization model, adopt an air conditioner operation strategy optimization algorithm based on reinforcement learning, and at the same time, combine the charge and discharge status of the energy storage unit to optimize the energy storage scheduling strategy, and generate a joint optimization plan for air conditioner operation and energy storage scheduling; A management module 304, configured to monitor the operation status of the air conditioner in real time, adopt a fault warning model based on an anomaly detection algorithm to identify potential faults of the air conditioner, and dynamically adjust the joint optimization plan through an adaptive control mechanism to ensure the operation stability and energy consumption efficiency of the air conditioner, and generate a final base station air conditioner energy consumption management plan.

[0106] It can be seen that, according to the temperature sensor, humidity sensor, air conditioner operation status monitoring device and communication device load data in the base station machine room, multi-source data is collected in real time to generate a multi-source fusion data set; for the multi-source fusion data set, the dynamic cooling demand within a future time window is predicted, and a dynamic optimization model of air conditioner energy consumption is constructed through the energy consumption efficiency curve; according to the prediction results of the dynamic cooling demand and the energy consumption optimization model, the operation mode, wind speed and temperature setting value of the air conditioner are dynamically adjusted, the energy storage scheduling strategy is optimized, and a joint optimization plan for air conditioner operation and energy storage scheduling is generated; the operation status of the air conditioner is monitored in real time, the joint optimization plan is dynamically adjusted, and a final base station air conditioner energy consumption management plan is generated, so that the air conditioner energy consumption management not only has real-time performance and accuracy, but also can achieve more efficient energy consumption control and energy storage scheduling optimization through intelligent means.

[0107] An embodiment of the present invention further provides a storage medium, in which a computer program is stored, and the computer program is configured to execute the steps in any one of the above method embodiments when running.

[0108] Specifically, in this embodiment, the above storage medium may be configured to store a computer program for executing the following steps: S201, according to the temperature sensor, humidity sensor, air conditioner operation status monitoring device and communication device load data in the base station machine room, adopt a data acquisition and preprocessing method based on edge computing to collect multi-source data in real time, and eliminate environmental noise and electromagnetic interference through an adaptive filtering algorithm to generate a multi-source fusion data set; S202, for the multi-source fusion data set, adopt a prediction model based on a lightweight spatio-temporal neural network, combine the real-time load changes of the base station communication equipment and the external environmental temperature fluctuations, predict the dynamic cooling demand within a future time window, and construct a dynamic optimization model of air conditioner energy consumption through the energy consumption efficiency curve; S203, according to the prediction results of the dynamic cooling demand and the energy consumption optimization model, adopt an air conditioner operation strategy optimization algorithm based on reinforcement learning to dynamically adjust the operation mode, wind speed and temperature setting value of the air conditioner. At the same time, combine the charge and discharge status of the energy storage unit to optimize the energy storage scheduling strategy and generate a joint optimization plan for air conditioner operation and energy storage scheduling; S204, monitor the operation status of the air conditioner in real time, adopt a fault warning model based on an anomaly detection algorithm to identify potential faults of the air conditioner, and dynamically adjust the joint optimization plan through an adaptive control mechanism to ensure the operation stability and energy consumption efficiency of the air conditioner, and generate a final base station air conditioner energy consumption management plan.

[0109] It can be seen that according to the temperature sensor, humidity sensor, air conditioner operation status monitoring device and communication equipment load data in the base station machine room, multi-source data is collected in real time to generate a multi-source fusion data set; for the multi-source fusion data set, the dynamic cooling demand within the future time window is predicted, and a dynamic optimization model of air conditioner energy consumption is constructed through the energy consumption efficiency curve; according to the prediction results of the dynamic cooling demand and the energy consumption optimization model, the operation mode, wind speed and temperature setting value of the air conditioner are dynamically adjusted, the energy storage scheduling strategy is optimized, and a joint optimization plan for air conditioner operation and energy storage scheduling is generated; the operation status of the air conditioner is monitored in real time, the joint optimization plan is dynamically adjusted, and a final base station air conditioner energy consumption management plan is generated, so that the air conditioner energy consumption management can not only have real-time performance and accuracy, but also realize more efficient energy consumption control and energy storage scheduling optimization through intelligent means.

[0110] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0111] Specifically, the above electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0112] Specifically, in this embodiment, the above processor may be configured to execute the following steps through a computer program: S201, according to the temperature sensor, humidity sensor, air conditioner operation status monitoring device and communication equipment load data in the base station machine room, adopt a data acquisition and preprocessing method based on edge computing to collect multi-source data in real time, and eliminate environmental noise and electromagnetic interference through an adaptive filtering algorithm to generate a multi-source fusion data set; S202, for the multi-source fusion data set, adopt a prediction model based on a lightweight spatio-temporal neural network, combine the real-time load changes of the base station communication equipment and the external environmental temperature fluctuations, predict the dynamic cooling demand within the future time window, and construct a dynamic optimization model of air conditioner energy consumption through the energy consumption efficiency curve; S203, according to the prediction results of the dynamic cooling demand and the energy consumption optimization model, adopt an air conditioner operation strategy optimization algorithm based on reinforcement learning to dynamically adjust the operation mode, wind speed and temperature setting value of the air conditioner. At the same time, combined with the charge and discharge status of the energy storage unit, optimize the energy storage scheduling strategy to generate a joint optimization plan for air conditioner operation and energy storage scheduling; S204, monitor the operation status of the air conditioner in real time, adopt a fault warning model based on an anomaly detection algorithm to identify potential faults of the air conditioner, and dynamically adjust the joint optimization plan through an adaptive control mechanism to ensure the operation stability and energy consumption efficiency of the air conditioner, and generate a final base station air conditioner energy consumption management plan.

[0113] It can be seen that, according to the temperature sensors, humidity sensors, air conditioner operation status monitoring devices and communication equipment load data in the base station computer room, multi-source data are collected in real time to generate a multi-source fusion data set; for the multi-source fusion data set, the dynamic cooling demand within a future time window is predicted, and a dynamic optimization model of air conditioner energy consumption is constructed through the energy consumption efficiency curve; according to the prediction results of the dynamic cooling demand and the energy consumption optimization model, the operation mode, wind speed and temperature set value of the air conditioner are dynamically adjusted, the energy storage scheduling strategy is optimized, and a joint optimization scheme for air conditioner operation and energy storage scheduling is generated; the operation status of the air conditioner is monitored in real time, the joint optimization scheme is dynamically adjusted, and a final base station air conditioner energy consumption management scheme is generated, so that the air conditioner energy consumption management can not only be real-time and accurate, but also achieve more efficient energy consumption control and energy storage scheduling optimization through intelligent means.

[0114] The structure, features and effects of the present invention have been described in detail based on the embodiments shown in the drawings. The above are only the preferred embodiments of the present invention, but the present invention is not limited to the scope defined by the drawings. Any changes made according to the concept of the present invention, or equivalent embodiments modified into equivalent changes, which still do not exceed the spirit covered by the description and the drawings, shall be within the protection scope of the present invention.

Claims

1. A method for energy storage in energy consumption management of a base station air conditioner, characterized in that, The method includes: According to the temperature sensor, humidity sensor, air conditioner operation status monitoring device and communication equipment load data in the base station machine room, adopt a data acquisition and preprocessing method based on edge computing to collect multi-source data in real time. Through an adaptive filtering algorithm, eliminate environmental noise and electromagnetic interference, and generate a multi-source fusion data set; For the multi-source fusion data set, adopt a prediction model based on a lightweight spatio-temporal neural network. Combine the real-time load changes of the base station communication equipment and the external environmental temperature fluctuations to predict the dynamic cooling demand within the future time window, and construct a dynamic optimization model of air conditioner energy consumption through the energy consumption efficiency curve; According to the prediction results of the dynamic cooling demand and the energy consumption optimization model, adopt an air conditioner operation strategy optimization algorithm based on reinforcement learning to dynamically adjust the operation mode, wind speed and temperature setting value of the air conditioner. At the same time, combine the charge and discharge status of the energy storage unit to optimize the energy storage scheduling strategy, and generate a joint optimization plan for air conditioner operation and energy storage scheduling; Monitor the operation status of the air conditioner in real time, adopt a fault warning model based on an anomaly detection algorithm to identify potential air conditioner faults, and through an adaptive control mechanism, dynamically adjust the joint optimization plan to ensure the operation stability and energy consumption efficiency of the air conditioner, and generate the final base station air conditioner energy consumption management plan.

2. The method according to claim 1, wherein The step of according to the temperature sensor, humidity sensor, air conditioner operation status monitoring device and communication equipment load data in the base station machine room, adopting a data acquisition and preprocessing method based on edge computing to collect multi-source data in real time, through an adaptive filtering algorithm, eliminating environmental noise and electromagnetic interference, and generating a multi-source fusion data set includes: According to the temperature sensor, humidity sensor, air conditioner operation status monitoring device and communication equipment load data in the base station machine room, adopt a data acquisition framework based on edge computing, and obtain multi-source data in real time through distributed data acquisition nodes. Each acquisition node is equipped with a lightweight data caching mechanism to ensure the real-time and continuous data acquisition; For the collected multi-source data, adopt an adaptive filtering algorithm based on wavelet transform to separate the environmental noise and electromagnetic interference in the signal. Through a dynamic threshold adjustment mechanism, perform differential filtering processing according to the noise characteristics of different sensor data to obtain a preliminarily denoised multi-source data set; For the denoised multi-source data set, adopt a time alignment algorithm based on dynamic time warping to eliminate the timestamp differences of different sensor data; at the same time, use a missing value filling method based on spatio-temporal correlation to interpolate and fill the missing data part to generate a time-synchronized complete data set; For the time-synchronized complete data set, adopt a feature fusion method based on a multi-head attention mechanism to extract multi-dimensional features of temperature, humidity, air conditioner operation status and communication equipment load data, and generate a high-precision multi-source fusion data set through a feature weight adaptive allocation mechanism.

3. The method according to claim 2, wherein The step of for the multi-source fusion data set, adopting a prediction model based on a lightweight spatio-temporal neural network, combining the real-time load changes of the base station communication equipment and the external environmental temperature fluctuations to predict the dynamic cooling demand within the future time window, and constructing a dynamic optimization model of air conditioner energy consumption through the energy consumption efficiency curve includes: For a multi-source fusion dataset, a prediction model based on a lightweight spatio-temporal neural network is adopted. Temporal features of the load change of base station communication equipment are extracted through a temporal convolutional layer, and local spatial features of the temperature distribution in the computer room are captured through a spatial graph convolutional layer to generate spatio-temporal feature representations; Combined with external environmental temperature fluctuation data, an external environment fusion technology based on an attention mechanism is adopted to perform weighted fusion of external temperature change features and spatio-temporal feature representations, enhancing the prediction ability of the prediction model for dynamic cooling demand and generating environment-enhanced spatio-temporal feature representations; The environment-enhanced spatio-temporal feature representations are input into the output layer of the prediction model, and a time series decoder based on multi-step prediction is adopted to predict the dynamic cooling demand of the base station computer room within a future time window, generating a cooling demand prediction result; According to the historical operation data of the air conditioner, a method for fitting the energy consumption efficiency curve based on piecewise linear regression is adopted to construct a relationship model between the air conditioner energy consumption, cooling capacity, and operation mode, and combined with the cooling demand prediction result, a dynamic optimization model of the air conditioner energy consumption is generated.

4. The method according to claim 3, characterized in that, Based on the prediction result of the dynamic cooling demand and the energy consumption optimization model, an air conditioner operation strategy optimization algorithm based on reinforcement learning is adopted to dynamically adjust the operation mode, wind speed, and temperature setting value of the air conditioner. At the same time, combined with the charge and discharge state of the energy storage unit, the energy storage scheduling strategy is optimized to generate a joint optimization plan for air conditioner operation and energy storage scheduling, including: Based on the prediction result of the dynamic cooling demand and the energy consumption optimization model, an air conditioner operation strategy optimization algorithm based on deep reinforcement learning is adopted to construct a reward function with the goal of minimizing energy consumption and meeting the cooling demand. Through a double Q-network structure, the operation mode, wind speed, and temperature setting value of the air conditioner are dynamically adjusted to generate a preliminary air conditioner operation optimization strategy; Combined with the current state of charge, charge and discharge efficiency curve, and grid electricity price fluctuation data of the energy storage unit, a multi-dimensional constraint model of the energy storage state is constructed. Through energy conservation constraints, charge and discharge rate constraints, and grid load balance constraints, a feasible solution space for energy storage scheduling is defined; For the energy storage state constraint model, an energy storage scheduling method based on a distributed constraint optimization algorithm is adopted. Combined with the energy consumption demand in the air conditioner operation optimization strategy, the charge and discharge plan of the energy storage unit is dynamically adjusted. Through time-segmented electricity price sensitivity analysis, the energy storage scheduling strategy is optimized to generate a preliminary energy storage scheduling plan; The air conditioner operation optimization strategy and the energy storage scheduling plan are co-optimized. A joint optimization framework based on game theory is adopted. Through the dynamic game between the air conditioner and the energy storage system, the energy consumption cost and the satisfaction of the cooling demand are balanced to generate a joint optimization plan for air conditioner operation and energy storage scheduling.

5. The method according to claim 4, characterized in that, For real-time monitoring of the air conditioner operation state, a fault warning model based on an anomaly detection algorithm is adopted to identify potential air conditioner faults. Through an adaptive control mechanism, the joint optimization plan is dynamically adjusted to ensure the operation stability and energy consumption efficiency of the air conditioner, generating a final base station air conditioner energy consumption management plan, including: Real-time monitoring of the air conditioner operation state, collecting the operation parameters and energy consumption data of the air conditioner, and through edge computing nodes, the collected data is processed in real time to generate a time series dataset of the air conditioner operation state; For time series datasets, an anomaly detection algorithm combining isolation forest and long short-term memory network is adopted to identify potential faults of air conditioners, and a fault warning signal is generated through a dynamic threshold adjustment mechanism; According to the fault warning signal, an adaptive control mechanism based on fuzzy logic is adopted to dynamically adjust the joint optimization scheme. Through a multi-objective optimization model, the fault recovery cost and energy consumption efficiency are balanced to generate an adaptive control strategy; The adaptive control strategy and the joint optimization scheme are integrated, and a dynamic adjustment method based on a feedback correction mechanism is adopted to update the air conditioner operation parameters and energy storage scheduling plan in real time, generating the final energy consumption management scheme for base station air conditioners.

6. A base station air conditioner energy consumption management energy storage system, characterized in that, The system includes: A collection module, which is used to collect multi-source data in real time according to the temperature sensors, humidity sensors, air conditioner operation status monitoring devices and communication device load data in the base station computer room, adopting a data collection and preprocessing method based on edge computing. Through an adaptive filtering algorithm, environmental noise and electromagnetic interference are eliminated to generate a multi-source fusion dataset; A prediction module, which is used to adopt a prediction model based on a lightweight spatio-temporal neural network for the multi-source fusion dataset, combined with the real-time load changes of base station communication devices and external environmental temperature fluctuations, to predict the dynamic cooling demand within a future time window, and through an energy consumption efficiency curve, construct a dynamic optimization model of air conditioner energy consumption; An adjustment module, which is used to adopt an air conditioner operation strategy optimization algorithm based on reinforcement learning according to the prediction results of the dynamic cooling demand and the energy consumption optimization model, dynamically adjust the operation mode, wind speed and temperature setting value of the air conditioner. At the same time, combined with the charge and discharge status of the energy storage unit, optimize the energy storage scheduling strategy to generate a joint optimization scheme for air conditioner operation and energy storage scheduling; A management module, which is used to monitor the operation status of the air conditioner in real time, adopt a fault warning model based on an anomaly detection algorithm to identify potential faults of the air conditioner, and through an adaptive control mechanism, dynamically adjust the joint optimization scheme to ensure the operation stability and energy consumption efficiency of the air conditioner, generating the final energy consumption management scheme for base station air conditioners.

7. The system according to claim 6, characterized in that The collection module specifically includes: According to the temperature sensors, humidity sensors, air conditioner operation status monitoring devices and communication device load data in the base station computer room, adopt a data collection framework based on edge computing, and obtain multi-source data in real time through distributed data collection nodes. Each collection node is equipped with a lightweight data caching mechanism to ensure the real-time and continuous data collection; For the collected multi-source data, adopt an adaptive filtering algorithm based on wavelet transform to separate the environmental noise and electromagnetic interference in the signal. Through a dynamic threshold adjustment mechanism, differential filtering processing is carried out according to the noise characteristics of different sensor data to obtain a preliminarily denoised multi-source dataset; For the denoised multi-source dataset, adopt a time alignment algorithm based on dynamic time warping to eliminate the timestamp differences of different sensor data; at the same time, use a missing value filling method based on spatio-temporal correlation to interpolate and fill the missing data part to generate a time-synchronized complete dataset; For the complete time-synchronized dataset, a feature fusion method based on the multi-head attention mechanism is adopted to extract multi-dimensional features of temperature, humidity, air conditioner operating status, and communication device load data. Through the feature weight adaptive allocation mechanism, a high-precision multi-source fusion dataset is generated.

8. The system according to claim 7, wherein The prediction module is specifically used for: For the multi-source fusion dataset, a prediction model based on a lightweight spatio-temporal neural network is adopted. Through the temporal convolutional layer, the temporal features of the load change of the base station communication device are extracted, and through the spatial graph convolutional layer, the local spatial features of the temperature distribution in the computer room are captured to generate a spatio-temporal feature representation; Combined with the external environmental temperature fluctuation data, an external environment fusion technology based on the attention mechanism is adopted to perform weighted fusion of the external temperature change features and the spatio-temporal feature representation, enhancing the prediction ability of the prediction model for dynamic cooling demand and generating an environment-enhanced spatio-temporal feature representation; Input the environment-enhanced spatio-temporal feature representation into the output layer of the prediction model, and adopt a time series decoder based on multi-step prediction to predict the dynamic cooling demand of the base station computer room within the future time window, generating a cooling demand prediction result; According to the historical operation data of the air conditioner, a method for fitting the energy consumption efficiency curve based on piecewise linear regression is adopted to construct a relationship model between the air conditioner energy consumption, cooling capacity, and operation mode, and combined with the cooling demand prediction result, a dynamic optimization model of the air conditioner energy consumption is generated.

9. A storage medium, characterized in that, A computer program is stored in the storage medium, wherein the computer program is set to execute the method according to any one of claims 1-5 when running.

10. An electronic device, comprising a memory and a processor, characterized in that, A computer program is stored in the memory, and the processor is set to run the computer program to execute the method according to any one of claims 1-5.

Images