Communication power supply adaptive optimization method and system based on intelligent prediction

Abstract

The invention discloses a communication power supply adaptive optimization method and system based on intelligent prediction, and relates to the technical field of power grid dispatching, and the method comprises the steps: collecting dynamic data and static data of a communication system, and constructing a load prediction model; on the basis of a load prediction model, combining static data to predict future load, combining dynamic data to calculate available energy of the system, and formulating a multi-path power supply scheduling strategy; and based on the prediction result, processing the prediction error, correcting the prediction result, and adjusting the power supply strategy under the condition of power interruption. The accurate load prediction enables the configuration of the standby power supply and the UPS system to be more reasonable, and reduces the risk of excessive configuration or insufficient configuration. The overall energy utilization efficiency is improved, and unnecessary energy waste is reduced. Efficient, reliable and intelligent operation of a power supply system is achieved, more stable and more economical power supply service is provided for users, and meanwhile a new solution is provided for intelligent management of an electric power system.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid dispatching, and in particular to a communication power adaptive optimization method and system based on intelligent prediction. Background Art

[0002] In the communication power supply system, load forecasting and optimization of power dispatching strategies are the core technologies to achieve efficient energy utilization and reliable system operation. Traditional methods mostly rely on statistical analysis of historical load data or prediction based on simple machine learning models, which have the following significant defects:

[0003] Existing technologies usually use single time series analysis (such as ARIMA) or shallow neural network models (such as BP neural network), which makes it difficult to effectively integrate dynamic data (such as real-time load, environmental parameters, equipment status) with static data (such as base station configuration, geographic information). This data fragmentation makes it impossible for the model to capture the temporal and spatial correlation of load changes, especially for sudden events (such as holiday traffic surges, extreme weather) and complex environmental factors (such as the impact of temperature and altitude on equipment power consumption), and the prediction error increases significantly. Summary of the invention

[0005] In view of the problems existing in the existing communication power adaptive optimization method based on intelligent prediction, the present invention is proposed.

[0006] Therefore, the problem to be solved by the present invention is how to regulate the power generation of the centralized control communication power supply and the battery and redesign the communication power supply.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] In the first aspect, an embodiment of the present invention provides a communication power adaptive optimization method based on intelligent prediction, which includes collecting dynamic data and static data of the communication system to construct a load prediction model; based on the load prediction model, combined with static data, predicting future loads, combining dynamic data to calculate the available energy of the system, and formulating a multi-power scheduling strategy; based on the prediction results, processing the prediction errors, correcting the prediction results, and adjusting the power supply strategy in the event of a power outage.

[0009] As a preferred solution of the communication power adaptive optimization method based on intelligent prediction described in the present invention, wherein: the dynamic data includes system load, temperature, humidity, mains status, energy consumption data, battery pack status, time, date type, special event mark, and communication system operation status;

[0010] The static data includes base station information, equipment specifications, altitude, and coverage area.

[0011] As a preferred solution of the communication power adaptive optimization method based on intelligent prediction described in the present invention, wherein: the architecture of the load prediction model includes a CNN layer, an LSTM layer, an Attention layer and a fully connected layer;

[0012] The CNN layer is used to capture local temporal patterns, and the LSTM layer is used to process long-term dependencies;

[0013] The Attention layer is used to focus on important time steps, and the fully connected layer is used to integrate features and output prediction results.

[0014] As a preferred solution of the communication power adaptive optimization method based on intelligent prediction described in the present invention, the future load is predicted based on the load prediction model, static and dynamic data are combined into an input vector, the input vector is input into the load prediction model, and the final output is performed, the output results are sorted, and the load prediction values ​​at the future T time points are obtained.

[0015] As a preferred solution of the communication power adaptive optimization method based on intelligent prediction described in the present invention, the available energy of the current system is evaluated based on dynamic data, including the sum of the available energy of the main power supply, all backup power supplies and the UPS system, which is expressed as:

[0016]

[0017] Among them, E total represents the total available energy, E main Indicates the available energy of the main power supply, represents the available energy of the ith backup power supply, E UPS Indicates the available energy of the UPS system;

[0018] Based on the predicted load and current available energy, the optimal power dispatch strategy is formulated, comprehensively considering energy cost, system reliability and energy utilization efficiency to minimize energy cost while maximizing system reliability and energy utilization efficiency. The objective function is defined as:

[0019] minF=w1*C energy +w2*(1 / R system )+w3*(1 / η energy )

[0020] Among them, C energy represents the energy cost, R system represents the system reliability, η energy represents energy utilization efficiency, w1, w2, w3 are weight coefficients;

[0021] At the same time, constraints are set to ensure that the power supply always meets the predicted load, the power switching time does not exceed the maximum allowable value, and the UPS backup time is not less than the minimum requirement. The improved particle swarm optimization algorithm is used to find the optimal scheduling strategy while meeting the constraints.

[0022] As a preferred solution of the communication power adaptive optimization method based on intelligent prediction described in the present invention, the prediction error is the relative deviation between the actual load and the predicted load, which is expressed as:

[0023] ε t =|L actual (t)-L pred (t)| / L pred (t)

[0024] Among them, ε t represents the relative prediction error at time t, L actual (t) represents the actual load at time t, ε threshold represents the predetermined error threshold, α represents the smoothing factor (0<α<1);

[0025] Set a threshold ε threshold , to determine whether prediction correction is needed, if ε t >ε threshold , then the correction is triggered;

[0026] The corrections include short-term corrections, long-term trend corrections and seasonal corrections.

[0027] As a preferred solution of the communication power adaptive optimization method based on intelligent prediction described in the present invention, wherein: based on the revised prediction value, the prediction of the subsequent time period is updated; if the revised prediction value shows a sudden increase in load, the backup power supply is started in advance to cope with possible peak demand, and the charging plan of the UPS system is adjusted to ensure sufficient backup energy; if the actual load at multiple consecutive time points is lower than the predicted value, the output of the main power supply is gradually reduced, and the originally planned backup power supply start time is postponed; if the revised prediction shows that load fluctuations are about to occur, the frequency of power switching is increased to more flexibly respond to load changes, adjust the energy distribution ratio, and balance the use of each power supply; if the prediction correction is triggered multiple times in a short period of time, the prediction cycle is shortened, the accuracy of the short-term prediction is improved, and the online update of the deep learning model is started to adapt to the new load pattern; if the revised prediction value indicates that the load pattern has changed significantly, the current scheduling strategy is re-evaluated to see whether it is still optimal, triggering a comprehensive power configuration optimization process.

[0028] In the second aspect, an embodiment of the present invention provides a communication power adaptive optimization system based on intelligent prediction, which includes: a collection module, which is used to collect dynamic data and static data of the communication system and construct a load prediction model; a construction module, which is used to predict future loads based on the load prediction model in combination with static data, calculate the available energy of the system in combination with dynamic data, and formulate a multi-power scheduling strategy; an optimization module, which is used to process prediction errors according to prediction results, correct prediction results, and adjust the power supply strategy in the event of power outages.

[0029] In a third aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the processor executes the computer program, any step of the above-mentioned communication power adaptive optimization method based on intelligent prediction is implemented.

[0030] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the above-mentioned communication power adaptive optimization method based on intelligent prediction is implemented.

[0031] The beneficial effect of the present invention is that the more accurate load prediction of the present invention makes the configuration of the backup power supply and UPS system more reasonable, reducing the risk of over-configuration or under-configuration. Accurate prediction and optimized scheduling strategy improve the overall energy utilization efficiency and reduce unnecessary energy waste. The efficient, reliable and intelligent operation of the power supply system is achieved, providing users with more stable and economical power supply services, and also providing a new solution for the intelligent management of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0033] Figure 1 A flowchart of a communication power adaptive optimization method based on intelligent prediction;

[0034] Figure 2 This is the overall design diagram of the DC power supply system. DETAILED DESCRIPTION

[0035] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0036] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0037] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0038] Example 1

[0039] Reference Figure 1 , which is the first embodiment of the present invention, and provides a communication power adaptive optimization method based on intelligent prediction, comprising:

[0040] S1: Collect dynamic and static data of the communication system and build a load prediction model;

[0041] Obtain communication power supply related data, including dynamic data and static data;

[0042] Dynamic data includes: system load, temperature, humidity, mains status, energy consumption data, battery pack status, time, date type, special event markers, and communication system operating status.

[0043] Static data includes: base station information, equipment specifications, altitude, and coverage area.

[0044] The base station information includes the number, type and geographical location of the base stations.

[0045] The collected data is cleaned, standardized and feature extracted to extract the data's time features, sliding window features, periodic features, cross features and external factor features.

[0046] Through these detailed data collection and preprocessing steps, high-quality, standardized input data are provided for the improved load forecasting model, thereby improving the model's prediction accuracy and robustness.

[0047] Based on the preprocessed data, external factors are integrated, including network traffic data, user behavior patterns, special event information, and weather data, and the integrated data is fused into a data vector.

[0048] Combining LSTM, CNN and Attention mechanism, a load forecasting model is constructed, and the data vector is input into the load forecasting model to better capture the multi-scale characteristics and long-term dependencies of the time series.

[0049] The architecture of the load prediction model includes: CNN layer, LSTM layer, Attention layer and fully connected layer. The CNN layer is used to capture local time patterns, the LSTM layer is used to process long-term dependencies, the Attention layer is used to focus on important time steps, and the fully connected layer is used to integrate features and output prediction results.

[0050] The model is further trained to enable it to accurately predict the load, and multiple indicators are used to evaluate the model performance until the target requirements are met. The load prediction model combines CNN, LSTM and Attention mechanisms, and incorporates external factors and graph attention network GAT to improve prediction accuracy and model robustness. First, the dynamic and static data of the communication system are collected. The dynamic data covers system load, environmental conditions, city power status, energy consumption, battery pack status, time and date type, special event markers and system operation status; static data includes base station information (such as quantity, type, geographical location), equipment specifications, altitude and coverage area, etc. The data is cleaned, standardized and feature extracted, especially the extraction of time features, sliding window features, periodic features, cross features and external factor features, to ensure the quality and consistency of the input model data. External factors, such as network traffic, user behavior, special events and weather conditions, are integrated into data vectors through specific algorithms to further enhance the richness of the model input. The model structure consists of multiple layers. The CNN layer captures the local time pattern in the input sequence, and the LSTM layer is responsible for processing the long-term dependencies in the sequence. Its core equation controls the information flow through the forget gate, input gate, and output gate. The Attention mechanism allows the model to focus on the more critical time steps in the time series, thereby improving the prediction accuracy. The fully connected layer integrates the features of the above layers and finally outputs the prediction results. In addition, the model also uses the Monte Carlo Dropout technology for uncertainty estimation, and the EWC algorithm to prevent catastrophic forgetting, ensuring the effective transfer learning of the model between new and old tasks. After pre-training on the source domain data, the model is fine-tuned on the target domain data, using the maximum mean difference MMD loss to reduce the distribution difference between the source and target domains and improve the generalization ability of the model. Finally, GAT is used to process the network topology and optimize the feature propagation through the attention coefficient between nodes, so that the final prediction results are closer to actual needs.

[0051] S2: Based on the load prediction model, combined with static data, predict future loads, combined with dynamic data to calculate the system available energy, and formulate multi-power supply scheduling strategies.

[0052] Based on the load prediction model, the future load is predicted, the static and dynamic data are combined into an input vector, the input vector is input into the load prediction model, and the final output is performed, and the output results are sorted to obtain the load prediction value at the future T time points.

[0053] Arrange the output results to obtain the load forecast value at the next T time points, expressed as:

[0054] L pred =[l1,l2,...,l T ]

[0055] Among them, l t represents the predicted load at time t.

[0056] Evaluate the available energy of the current system, including the sum of the available energy of the main power supply, all backup power supplies, and the UPS system, expressed as:

[0057]

[0058] Among them, E total represents the total available energy, E main Indicates the available energy of the main power supply, represents the available energy of the ith backup power supply, E UPS Indicates the available energy of the UPS system.

[0059] Based on the predicted load and current available energy, the optimal power dispatch strategy is formulated, comprehensively considering energy cost, system reliability and energy utilization efficiency to minimize energy cost while maximizing system reliability and energy utilization efficiency. The objective function is defined as:

[0060] minF=w1*C energy +w2*(1 / R system )+w3*(1 / η energy )

[0061] Among them, C energy represents the energy cost, R system represents the system reliability, η energy represents energy utilization efficiency, w1, w2, w3 are weight coefficients;

[0062] At the same time, constraints are set to ensure that the power supply always meets the predicted load, the power switching time does not exceed the maximum allowable value, and the UPS backup time is not less than the minimum requirement. The improved particle swarm optimization algorithm is used to find the optimal scheduling strategy while meeting the constraints.

[0063] Furthermore, based on the load prediction model, first, the static data of the communication system, such as base station information, equipment specifications, altitude and coverage area, and the dynamic data, such as system load, temperature, humidity, mains status, energy consumption data, battery pack status, time and date type, special event markers and system operation status, are combined into a comprehensive input vector, which is then input into the load prediction model. The model integrates CNN, LSTM, Attention mechanism and graph attention network GAT. These advanced components are used to capture the multi-scale features, long-term dependencies and network topology in the time series, so as to obtain the load prediction value at each time point in the future. ,Next, the available energy of the current system is evaluated, which includes the total available energy of the main power source, all ,backup power sources and the UPS system. Based on the predicted load and the current available energy, the optimal power ,scheduling strategy is formulated. This strategy needs to comprehensively consider the energy cost, system reliability and energy utilization efficiency. The goal is to minimize the energy cost while maximizing the system reliability and energy utilization efficiency. To this end, several constraints are set to ensure that the power supply can meet the predicted load demand, the switching time between power sources does not exceed the maximum allowed value, and the UPS backup time is not less than the minimum requirement. The ,improved particle swarm optimization algorithm is used to find the optimal scheduling strategy under ,these constraints to achieve efficient and reliable power management.

[0064] S3: Based on the prediction results, process the prediction errors, correct the prediction results, and adjust the power supply strategy in the event of power outages.

[0065] The prediction error is the relative deviation between the actual load and the predicted load, expressed as:

[0066] ε t =|L actual (t)-L pred (t)| / L pred (t)

[0067] Among them, ε t represents the relative prediction error at time t, L actual (t) represents the actual load at time t, ε threshold represents the predetermined error threshold, α represents the smoothing factor (0<α<1);

[0068] Set a threshold ε threshold , to determine whether prediction correction is needed, if εt >ε threshold , then the correction is triggered;

[0069] Corrections include short-term corrections, long-term trend corrections and seasonal corrections.

[0070] If the revised forecast value shows a sudden increase in load, the backup power supply will be started in advance to cope with the possible peak demand, and the charging plan of the UPS system will be adjusted to ensure sufficient backup energy.

[0071] If the actual load at multiple consecutive time points is lower than the predicted value, the output of the main power supply will be gradually reduced and the originally planned start-up time of the backup power supply will be postponed.

[0072] If the revised forecast shows that load fluctuations are imminent, the frequency of power switching is increased to more flexibly respond to load changes, adjust the energy distribution ratio, and balance the use of each power source.

[0073] If the forecast correction is triggered multiple times in a short period of time, the forecast cycle is shortened, the accuracy of the short-term forecast is improved, and the online update of the deep learning model is initiated to adapt to the new load pattern.

[0074] If the revised forecast value indicates that the load pattern has changed significantly, the current scheduling strategy is re-evaluated to see if it is still optimal, triggering a comprehensive power configuration optimization process.

[0075] If the revised forecast value is close to the system's power supply upper limit, load peak reduction measures are activated, such as temporary shutdown of non-critical equipment, and an early warning signal is issued to remind managers that additional power support may be needed.

[0076] In summary, the present invention significantly improves the accuracy of load prediction by comprehensively applying multiple correction methods, so that the prediction model can be dynamically adjusted and quickly adapt to load changes, thereby reducing prediction deviations. Based on these more accurate prediction results, the power scheduling strategy is optimized, and the power supply reliability of the system under normal and interruption conditions is improved. More accurate load prediction makes the configuration of the backup power supply and UPS system more reasonable, reducing the risk of over-configuration or under-configuration. At the same time, accurate predictions and optimized scheduling strategies improve the overall energy utilization efficiency and reduce unnecessary energy waste. The efficient, reliable and intelligent operation of the power supply system is achieved, providing users with more stable and economical power supply services, and also providing new solutions for the intelligent management of power systems.

[0077] Example 2

[0078] Reference Figure 1 Based on the first embodiment, this embodiment further provides a communication power adaptive optimization system based on intelligent prediction, including:

[0079] A collection module is used to collect dynamic and static data of the communication system and build a load prediction model;

[0080] Building modules for predicting future loads based on load prediction models combined with static data, calculating system available energy combined with dynamic data, and formulating multi-source power dispatch strategies;

[0081] The optimization module is used to process the prediction errors according to the prediction results, correct the prediction results, and adjust the power supply strategy in the event of power outages.

[0082] This embodiment also provides a computer device, which is suitable for the case of a communication power adaptive optimization method based on intelligent prediction, including a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the communication power adaptive optimization method based on intelligent prediction as proposed in the above embodiment.

[0083] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0084] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for realizing adaptive optimization of communication power supply based on intelligent prediction as proposed in the above embodiment is implemented.

[0085] The storage medium proposed in this embodiment and the data storage method proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be found in the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0086] Example 3

[0087] On the basis of the previous two embodiments, this embodiment provides a communication power adaptive optimization method based on intelligent prediction. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0088] Based on a typical communication base station, including main power supply, backup power supply and UPS system, it lasts for 30 days to simulate the load changes of a real communication base station, including daily fluctuations, weekend effects and emergencies.

[0089] The comparison objects include the present invention and the traditional fixed threshold scheduling system.

[0090] The prediction model of the present invention is trained using historical data of real communication base stations to generate 30 days of simulated load data, including normal fluctuations and emergencies, and power supply scheduling is performed using the present invention and traditional methods respectively.

[0091] The indicators of the two methods are recorded and compared. The experimental results are shown in Table 1:

[0092] Table 1 Comparative experimental results

[0093] Performance Indicators System of the present invention Traditional fixed threshold system Improvement Number of power outages (within 30 days) 2 times 7 times 71.4% reduction Average power outage time 5 minutes 15 minutes 66.7% shorter Energy efficiency 92％ 80％ 15% improvement Number of backup power supply startups (within 30 days) 6 times 10 times 40% reduction Average system response time 2 minutes 5 minutes 60% reduction UPS charge and discharge times (within 30 days) 45 times 60 times 25% reduction Successful response rate of sudden load events 95％ 70％ 35.7% improvement System operating cost (relative value) 85 100 15% reduction

[0094] This comparative test clearly demonstrates the significant advantages of the system of the present invention over the traditional system in various key performance indicators. In particular, the present invention has shown significant improvements in reducing the number of power outages, shortening the power outage time, improving energy utilization, reducing the number of equipment startups, and improving the ability to respond to emergencies. These data intuitively reflect the outstanding results of the present invention in improving power supply reliability, optimizing energy utilization, extending equipment life, and enhancing system adaptability.

[0095] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A communication power adaptive optimization method based on intelligent prediction, characterized in that: include, Collect dynamic and static data of the communication system and build a load prediction model; Based on the load prediction model, combined with static data, predict future loads, combined with dynamic data to calculate the system available energy, and formulate a multi-channel power supply scheduling strategy; Based on the prediction results, the prediction errors are processed, the prediction results are corrected, and the power supply strategy is adjusted in the event of power outages.

2. The communication power adaptive optimization method based on intelligent prediction according to claim 1, characterized in that: The dynamic data includes system load, temperature, humidity, mains status, energy consumption data, battery pack status, time, date type, special event markers, and communication system operation status; The static data includes base station information, equipment specifications, altitude, and coverage area.

3. The communication power adaptive optimization method based on intelligent prediction according to claim 2, characterized in that: The architecture of the load prediction model includes a CNN layer, an LSTM layer, an Attention layer, and a fully connected layer; The CNN layer is used to capture local temporal patterns, the LSTM layer is used to process long-term dependencies; the Attention layer is used to focus on important time steps, and the fully connected layer is used to integrate features and output prediction results.

4. The communication power adaptive optimization method based on intelligent prediction according to claim 3, characterized in that: Based on the load prediction model, the future load is predicted, the static and dynamic data are combined into an input vector, the input vector is input into the load prediction model, and the final output is performed, and the output results are sorted to obtain the load prediction value at the future T time points.

5. The communication power adaptive optimization method based on intelligent prediction according to claim 4, characterized in that: Based on dynamic data, the available energy of the current system is evaluated, including the sum of the available energy of the main power supply, all backup power supplies and UPS systems, expressed as: Among them, E total represents the total available energy, E main Indicates the available energy of the main power supply, represents the available energy of the ith backup power supply, E UPS Indicates the available energy of the UPS system; Based on the predicted load and current available energy, the optimal power dispatch strategy is formulated, comprehensively considering energy cost, system reliability and energy utilization efficiency to minimize energy cost while maximizing system reliability and energy utilization efficiency. The objective function is defined as: minF=w1*C energy +w2*(1 / R system )+w3*(1 / η energy ) Among them, C energy represents the energy cost, R system represents the system reliability, η energy represents energy utilization efficiency, w1, w2, w3 are weight coefficients; At the same time, constraints are set to ensure that the power supply always meets the predicted load, the power switching time does not exceed the maximum allowable value, and the UPS backup time is not less than the minimum requirement. The improved particle swarm optimization algorithm is used to find the optimal scheduling strategy while meeting the constraints.

6. The communication power adaptive optimization method based on intelligent prediction according to claim 5, characterized in that: The prediction error is the relative deviation between the actual load and the predicted load, expressed as: ε t =|L actual (t)-L pred (t)| / L pred (t) Among them, ε t represents the relative prediction error at time t, L actual (t) represents the actual load at time t, ε threshold represents the predetermined error threshold, α represents the smoothing factor (0<α<1); Set a threshold ε threshold , to determine whether prediction correction is needed, if ε t >ε threshold , then the correction is triggered; The corrections include short-term corrections, long-term trend corrections and seasonal corrections.

7. The communication power adaptive optimization method based on intelligent prediction according to claim 6, characterized in that: Based on the revised forecast value, update the forecast for subsequent time periods, If the revised forecast value shows a sudden increase in load, the backup power supply is started in advance to cope with the possible peak demand, and the charging plan of the UPS system is adjusted to ensure sufficient backup energy; If the actual load at multiple consecutive time points is lower than the predicted value, the output of the main power supply will be gradually reduced, and the originally planned start-up time of the backup power supply will be postponed; If the revised forecast shows that load fluctuations are imminent, the frequency of power switching is increased to more flexibly respond to load changes, adjust the energy distribution ratio, and balance the use of each power source; If the forecast correction is triggered multiple times in a short period of time, the forecast cycle is shortened, the accuracy of the short-term forecast is improved, and the online update of the deep learning model is initiated to adapt to the new load pattern; If the revised forecast value indicates that the load pattern has changed significantly, the current scheduling strategy is re-evaluated to see if it is still optimal, triggering a comprehensive power configuration optimization process.

8. A communication power adaptive optimization system based on intelligent prediction, based on the communication power adaptive optimization method based on intelligent prediction according to any one of claims 1 to 7, characterized in that: include, A collection module is used to collect dynamic and static data of the communication system and build a load prediction model; A construction module is used to predict future loads based on the load prediction model in combination with static data, calculate the available energy of the system in combination with dynamic data, and formulate a multi-channel power supply scheduling strategy; The optimization module is used to process the prediction errors according to the prediction results, correct the prediction results, and adjust the power supply strategy in the event of power outages.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the communication power adaptive optimization method based on intelligent prediction according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the communication power adaptive optimization method based on intelligent prediction described in any one of claims 1 to 7 are implemented.

Images