Communication base station energy storage management system and method
By adopting time sequence feature processing and interrelated technology based on artificial intelligence in the communication base station energy storage management system, the real-time and holistic problems of traditional prediction methods are solved, and accurate prediction of short-term energy demand and grid stability are achieved. promote.
Patent Information
- Application Number
- CN202411288616.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-14
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-09-14
AI Technical Summary
The traditional communication base station energy storage management system is based on historical data prediction, lacks real-time data support, and cannot respond to emergencies and changes in a timely manner. It focuses too much on a single energy type or factor, ignores the integrity and mutual influence of the energy system, resulting in the prediction results and actual situations. The requirements do not match.
Through sensors and intelligent instruments, the time series of energy storage equipment generation, load demand and battery energy storage status are collected, and the timing characteristics based on artificial intelligence are processed and related to each other to generate prediction results of energy demand in the short term in the future, and intelligently adjust the battery Charging and discharging strategies to optimize energy use and grid load.
Accurate prediction of short-term energy demand is achieved, battery life is extended, energy waste and grid load fluctuations are reduced, and grid stability is enhanced.
Smart Images

Figure CN118801448B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of intelligent management, and more specifically, to a communication base station energy storage management system and method. Background Art
[0002] The communication base station energy storage management system is a system specially designed to monitor and manage the battery energy storage units in the communication base station. The main purpose of this system is to improve the service life of the battery, ensure the reliability of the power supply of the communication base station, and optimize the efficiency of the use of battery energy storage resources.
[0003] In order to more effectively plan and allocate energy resources, ensure a balance between supply and demand, and reduce energy waste, energy forecasting is necessary. This can help grid operators identify potential supply and demand imbalances in advance and take measures to avoid grid overload or power shortage, thereby improving the stability and reliability of the grid.
[0004] However, since traditional forecasting methods are usually based on historical data and lack the support of real-time data, they cannot respond to emergencies and changes in a timely manner, resulting in forecast results that do not match actual needs. Secondly, traditional forecasting may focus too much on a single energy type or a single factor, while ignoring the integrity of the energy system and the mutual influence between different energy sources.
[0005] Therefore, an optimized communication base station energy storage management system is expected. Summary of the invention
[0006] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a communication base station energy storage management system and method, which collects the time series of the power generation of the energy storage device, the load demand and the energy storage state of the battery through sensors and intelligent meters, wherein the energy storage state includes the charging state, discharging state, temperature, voltage and current of the battery, and uses artificial intelligence to perform time series feature processing and mutual correlation on the power generation of the device, the load demand and the energy storage state of the battery, so as to obtain the prediction result of the predicted value of the energy demand in the future short time. In this way, the energy demand in the short term can be accurately predicted, so as to intelligently adjust the battery charging and discharging strategy, extend the battery life, optimize the use of energy, reduce energy waste, and reduce the fluctuation of the power grid load, thereby enhancing the stability of the power grid.
[0007] According to one aspect of the present application, a communication base station energy storage management system is provided, which includes:
[0008] The communication base station parameter acquisition module is used to collect the time series of the power generation of the energy storage device, the load demand and the energy storage state of the battery through sensors and intelligent meters, wherein the energy storage state includes the charging state, discharging state, temperature, voltage and current of the battery;
[0009] A parameter timing arrangement module, used for arranging the time series of the power generation of the energy storage device and the time series of the load demand of the energy storage device into a power generation timing input vector and a load demand timing input vector according to the time dimension;
[0010] A feature extraction and association module is used to perform power generation time series coding and load demand time series coding on the power generation time series input vector and the load demand time series input vector to obtain a power generation time series feature vector and a load demand time series feature vector, and then perform feature association on the power generation time series feature vector and the load demand time series feature vector to obtain a power generation-load demand time series association feature matrix;
[0011] An energy storage state feature extraction module, used for arranging the time series of energy storage states of the energy storage device into an energy storage state time series input matrix and then extracting energy storage state time series features to obtain an energy storage state time series feature matrix;
[0012] A feature fusion module, used to obtain an energy demand fusion feature matrix based on the power generation-load demand time series correlation feature matrix and the energy storage state time series feature matrix;
[0013] A feature optimization module, used for performing similarity information correction based on the measurement dimension on the energy demand fusion feature vector expanded by the energy demand fusion feature matrix to obtain an optimized energy demand fusion feature vector;
[0014] The prediction result generation module is used to obtain the prediction result based on the optimized energy demand fusion feature vector.
[0015] In the above-mentioned communication base station energy storage management system, the feature extraction and association module includes: a timing encoding unit, which is used to input the power generation timing input vector and the load demand timing input vector into the power generation timing encoder and the load demand timing encoder respectively to obtain the power generation timing feature vector and the load demand timing feature vector; and a timing association unit, which is used to perform time association on the power generation timing feature vector and the load demand timing feature vector to obtain the power generation-load demand timing association feature matrix.
[0016] In the above communication base station energy storage management system, the timing association unit is used to: input the power generation timing characteristic vector and the load demand timing characteristic vector into the characteristic timing association coding module to obtain the power generation-load demand timing association characteristic matrix.
[0017] In the above communication base station energy storage management system, the timing association unit is used to: calculate the transposed vector of the power generation timing characteristic vector and the load demand timing characteristic vector to obtain the power generation-load demand timing association characteristic matrix.
[0018] In the above-mentioned communication base station energy storage management system, the energy storage state feature extraction module is used to: arrange the time series of the energy storage state of the energy storage device according to the time dimension and the sample dimension into the energy storage state timing input matrix, and then input it into the energy storage state timing feature extractor based on the convolutional neural network model to obtain the energy storage state timing feature matrix.
[0019] In the above-mentioned communication base station energy storage management system, the feature optimization module includes: determining a decoding weight matrix of a decoder; performing matrix decoupling on the decoding weight matrix in units of row vectors to obtain a set of decoding weight vectors; calculating an energy demand fusion metric learning loss information item between each decoding weight vector in the set of decoding weight vectors and the energy demand fusion feature vector to obtain an energy demand fusion metric learning compensation representation vector; inputting the energy demand fusion metric learning compensation representation vector into a softmax activation function to obtain a probabilistic energy demand fusion metric learning compensation weight vector; calculating the positional point multiplication of the probabilistic energy demand fusion metric learning compensation weight vector and the energy demand fusion feature vector to obtain a point product feature vector, and calculating the weighted sum of the point product feature vector and the energy demand fusion feature vector to obtain the optimized energy demand fusion feature vector.
[0020] In the above-mentioned communication base station energy storage management system, the prediction result generation module is used to: input the optimized energy demand fusion feature vector into the decoder-based energy predictor to obtain a prediction result, and the prediction result is used to represent the predicted value of energy demand in the future short period of time.
[0021] According to another aspect of the present application, a communication base station energy storage management method is provided, which includes:
[0022] Collect the time series of energy storage equipment power generation, load demand and battery energy storage status through sensors and smart meters, wherein the energy storage status includes battery charge state, discharge state, temperature, voltage and current;
[0023] Arrange the time series of the power generation of the energy storage device and the time series of the load demand of the energy storage device according to the time dimension into a power generation time series input vector and a load demand time series input vector;
[0024] Performing power generation time series coding and load demand time series coding on the power generation time series input vector and the load demand time series input vector to obtain a power generation time series feature vector and a load demand time series feature vector, and then performing feature association on the power generation time series feature vector and the load demand time series feature vector to obtain a power generation-load demand time series association feature matrix;
[0025] Arranging the time series of the energy storage state of the energy storage device into an energy storage state time series input matrix and then extracting the energy storage state time series features to obtain an energy storage state time series feature matrix;
[0026] Based on the power generation-load demand time series correlation characteristic matrix and the energy storage state time series characteristic matrix, an energy demand fusion characteristic matrix is obtained;
[0027] Performing similarity information correction based on the measurement dimension on the energy demand fusion feature vector expanded from the energy demand fusion feature matrix to obtain an optimized energy demand fusion feature vector;
[0028] Based on the optimized energy demand fusion feature vector, a prediction result is obtained.
[0029] In the above-mentioned communication base station energy storage management method, the power generation timing input vector and the load demand timing input vector are subjected to power generation timing encoding and load demand timing encoding to obtain the power generation timing characteristic vector and the load demand timing characteristic vector, and then the power generation timing characteristic vector and the load demand timing characteristic vector are characteristically associated to obtain the power generation-load demand timing association characteristic matrix, including: inputting the power generation timing input vector and the load demand timing input vector into the power generation timing encoder and the load demand timing encoder respectively to obtain the power generation timing characteristic vector and the load demand timing characteristic vector; and, timing-associating the power generation timing characteristic vector and the load demand timing characteristic vector to obtain the power generation-load demand timing association characteristic matrix.
[0030] Compared with the prior art, the present application provides a communication base station energy storage management system and method, which collects the time series of energy storage equipment power generation, load demand and battery energy storage status through sensors and intelligent meters, wherein the energy storage status includes the battery charging state, discharge state, temperature, voltage and current, and uses artificial intelligence to perform time series feature processing and mutual correlation on the equipment power generation, load demand and battery energy storage status, so as to obtain the prediction result of the predicted value of energy demand in the future short time. In this way, the energy demand in the short term can be accurately predicted, so as to intelligently adjust the battery charging and discharging strategy, extend battery life, optimize energy use, reduce energy waste, and reduce the fluctuation of power grid load, thereby enhancing the stability of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.
[0032] Figure 1 It is a block diagram of a communication base station energy storage management system according to an embodiment of the present application.
[0033] Figure 2 Schematic diagram of the architecture of a communication base station energy storage management system according to an embodiment of the present application.
[0034] Figure 3 It is a block diagram of a feature extraction association module in a communication base station energy storage management system according to an embodiment of the present application.
[0035] Figure 4 Flow chart of a communication base station energy storage management method according to an embodiment of the present application. DETAILED DESCRIPTION
[0036] The embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, but rather these embodiments are provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.
[0037] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.
[0038] It should be noted that the terms "first\second\third" involved in the embodiments of the present application are only used to distinguish similar objects, and do not represent a specific order for the objects. It is understandable that the specific order or sequence of "first\second\third" can be interchanged where permitted. It should be understood that the objects distinguished by "first\second\third" can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.
[0039] As shown in the present disclosure and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.
[0040] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.
[0041] Communication base stations (also called base stations or wireless base stations) are fixed-location radio transmission equipment, and they are a key component of modern mobile communication networks. The main function of a base station is to act as a communication intermediary between mobile devices such as mobile phones, tablets, and smartphones and the telephone network. Specifically, base stations provide wireless signal coverage for mobile devices, ensuring that users can communicate within the coverage area; base stations handle voice, text messages, and data communications, including Internet access; base stations are equipped with signal amplifiers that can enhance signal strength and expand the coverage of wireless networks; base stations require a stable power supply, and some base stations are also equipped with backup power systems to prevent power outages from affecting services. The construction and maintenance of communication base stations are essential to ensuring the coverage, service quality, and user experience of mobile communication networks.
[0042] The communication base station energy storage management system is a system specially designed to monitor and manage the battery energy storage units in the communication base station. The main purpose of this system is to improve the service life of the battery, ensure the reliability of the power supply of the communication base station, and optimize the efficiency of the use of battery energy storage resources.
[0043] In order to more effectively plan and allocate energy resources, ensure a balance between supply and demand, and reduce energy waste, energy forecasting is necessary. This can help grid operators identify potential supply and demand imbalances in advance and take measures to avoid grid overload or power shortage, thereby improving the stability and reliability of the grid.
[0044] However, since traditional forecasting methods are usually based on historical data and lack the support of real-time data, they cannot respond to emergencies and changes in a timely manner, resulting in forecast results that do not match actual needs. Secondly, traditional forecasting may focus too much on a single energy type or a single factor, while ignoring the integrity of the energy system and the mutual influence between different energy sources.
[0045] Therefore, an optimized communication base station energy storage management system is expected, which collects the time series of energy storage equipment power generation, load demand and battery energy storage state through sensors and intelligent meters, wherein the energy storage state includes the battery charging state, discharging state, temperature, voltage and current, and uses artificial intelligence to process the time series characteristics of the equipment power generation, the load demand and the battery energy storage state and correlate them, so as to obtain the prediction result of the predicted value of the energy demand in the future short time. In this way, the energy demand in the short term can be accurately predicted, so as to intelligently adjust the battery charging and discharging strategy, extend the battery life, optimize the use of energy, reduce energy waste, and reduce the fluctuation of the power grid load, thereby enhancing the stability of the power grid.
[0046] Figure 1 It is a block diagram of a communication base station energy storage management system according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of the architecture of a communication base station energy storage management system according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to the communication base station energy storage management system 100 of the embodiment of the present application, it includes: a communication base station parameter acquisition module 110, which is used to collect the time series of the power generation of the energy storage device, the load demand and the energy storage state of the battery through sensors and intelligent meters, wherein the energy storage state includes the charging state, the discharging state, the temperature, the voltage and the current of the battery; a parameter timing arrangement module 120, which is used to arrange the time series of the power generation of the energy storage device and the time series of the load demand of the energy storage device according to the time dimension into a power generation timing input vector and a load demand timing input vector; a feature extraction association module 130, which is used to perform power generation timing encoding and load demand timing encoding on the power generation timing input vector and the load demand timing input vector to obtain the power generation timing feature vector and the load demand timing feature vector, and then perform the power generation timing feature vector on the power generation timing feature vector. The energy storage device is used to extract the energy storage state timing characteristics after arranging the time series of the energy storage state of the energy storage device into an energy storage state timing input matrix to obtain an energy storage state timing characteristic matrix; a feature fusion module 150 is used to obtain an energy demand fusion feature matrix based on the energy storage state timing characteristic matrix and the energy storage state timing characteristic matrix; a feature optimization module 160 is used to perform similarity information correction based on the metric dimension on the energy demand fusion feature vector expanded from the energy demand fusion feature matrix to obtain an optimized energy demand fusion feature vector; and a prediction result generation module 170 is used to obtain a prediction result based on the optimized energy demand fusion feature vector.
[0047] In the embodiment of the present application, the communication base station parameter acquisition module 110 is used to collect the time series of the power generation of the energy storage device, the load demand and the energy storage state of the battery through sensors and intelligent meters, wherein the energy storage state includes the charging state, discharging state, temperature, voltage and current of the battery. It should be understood that the power generation of the energy storage device, the load demand and the energy storage state of the battery are three key parameters in the communication base station energy storage management system. In detail, the power generation of the energy storage device reflects the amount of electricity generated by the energy storage device (such as solar panels, wind turbines, etc.) within a specific time. Understanding the power generation helps to evaluate the utilization efficiency of renewable energy and predict the power supply. The load demand refers to the power consumption or power demand of the communication base station in different time periods. Load data is crucial for predicting and planning power distribution to ensure that the needs of communication equipment are met. The charging state and the discharging state indicate whether the battery is currently storing or releasing energy; the temperature of the battery affects its performance and life, and too high or too low temperature may damage the battery; the voltage and current of the battery are key indicators for measuring battery performance, which are directly related to the health and safety status of the battery. Therefore, in order to manage energy more effectively and optimize the operation strategy of energy storage equipment, in the technical solution of the present application, sensors and smart meters are used to collect the time series of the power generation of energy storage equipment, load demand and battery energy storage status, wherein the energy storage status includes the battery charging status, discharging status, temperature, voltage and current. In this way, understanding the battery energy storage status is crucial to ensure that the communication base station can continue to operate when the power supply is unstable, and to prevent safety problems such as overcharging, over-discharging and overheating, thereby supporting the intelligent scheduling system to respond quickly according to the current energy supply and demand conditions.
[0048] In an embodiment of the present application, the parameter timing arrangement module 120 is used to arrange the time series of the power generation of the energy storage device and the time series of the load demand of the energy storage device according to the time dimension into a power generation timing input vector and a load demand timing input vector. Accordingly, considering that in the time series of the power generation of the energy storage device and the time series of the load demand of the energy storage device, the subsequent data points often have a dependency relationship with the previous data points, that is, the power generation of the energy storage device and the load demand of the energy storage device have a certain characteristic relationship on the time scale. Therefore, in order to retain the temporal variation trend of the power generation of the energy storage device and the load demand of the energy storage device, and to facilitate subsequent analysis and processing, in the technical solution of the present application, the time series of the power generation of the energy storage device and the time series of the load demand of the energy storage device are arranged according to the time dimension into a power generation timing input vector and a load demand timing input vector. That is to say, by arranging the time series of the power generation of the energy storage device and the time series of the load demand of the energy storage device according to the time dimension, the original time series of the power generation of the energy storage device and the load demand of the energy storage device can be maintained, ensuring that the trend and pattern of the power generation of the energy storage device and the load demand of the energy storage device over time can be observed during analysis, and these time series feature information can be captured and extracted, thereby improving the intelligence level of battery energy storage unit management and optimizing the balance of energy supply and demand.
[0049] In an embodiment of the present application, the feature extraction association module 130 is used to perform power generation timing encoding and load demand timing encoding on the power generation timing input vector and the load demand timing input vector to obtain a power generation timing feature vector and a load demand timing feature vector, and then perform feature association on the power generation timing feature vector and the load demand timing feature vector to obtain a power generation-load demand timing association feature matrix. Figure 3 is a block diagram of a feature extraction and association module in a communication base station energy storage management system according to an embodiment of the present application. Specifically, in the embodiment of the present application, Figure 3 As shown, the feature extraction association module 130 includes: a timing encoding unit 131, which is used to input the power generation timing input vector and the load demand timing input vector into the power generation timing encoder and the load demand timing encoder respectively to obtain the power generation timing feature vector and the load demand timing feature vector; and a timing association unit 132, which is used to perform time association on the power generation timing feature vector and the load demand timing feature vector to obtain the power generation-load demand timing association feature matrix.
[0050] Specifically, the time series encoding unit 131 is used to input the power generation time series input vector and the load demand time series input vector into the power generation time series encoder and the load demand time series encoder respectively to obtain the power generation time series feature vector and the load demand time series feature vector. Accordingly, considering that the power generation time series input vector and the load demand time series input vector have characteristic information in time series, which includes the time series characteristic information about the power generation and the load demand, such as the change trend, periodic change and seasonal change of the power generation and the load demand in time series. The time series encoder can extract and capture the time series pattern and structural characteristics in the sequence data from the time series data. Therefore, in order to extract the characteristic information about the power generation and the load demand on the time scale from the power generation timing input vector and the load demand timing input vector, in the technical solution of the present application, the power generation timing input vector and the load demand timing input vector are respectively input into the power generation timing encoder and the load demand timing encoder to obtain the power generation timing characteristic vector and the load demand timing characteristic vector, thereby providing support and assistance for the subsequent prediction of future short-term energy demand.
[0051] Specifically, the timing association unit 132 is used to perform timing association on the power generation timing characteristic vector and the load demand timing characteristic vector to obtain the power generation-load demand timing association characteristic matrix. Specifically, in the embodiment of the present application, the timing association unit is used to: input the power generation timing characteristic vector and the load demand timing characteristic vector into the characteristic timing association coding module to obtain the power generation-load demand timing association characteristic matrix. It should be understood that there is a mutual correlation between the power generation timing characteristic vector and the load demand timing characteristic vector. For example, the power generation timing characteristic vector may show that the power generation is higher in certain time periods (such as daytime), while the load demand timing characteristic vector may show a higher power demand in the same time period, indicating a direct correlation between supply and demand; if the power generation timing characteristic shows that the performance of a certain power generation equipment has decreased under certain conditions, and the load demand timing characteristic shows a high demand under the same conditions, this may indicate that predictive maintenance of the power generation equipment is required. Therefore, in order to better integrate the correlation and mutual influence between the power generation timing characteristic vector and the load demand timing characteristic vector, in the technical solution of the present application, the power generation timing characteristic vector and the load demand timing characteristic vector are input into the characteristic timing association coding module to obtain the power generation-load demand timing association characteristic matrix, so as to better understand the timing correlation between the two for subsequent precise analysis and processing.
[0052] More specifically, in an embodiment of the present application, the timing association unit is used to: calculate the transposed vector of the power generation timing characteristic vector and the load demand timing characteristic vector to obtain the power generation-load demand timing association characteristic matrix.
[0053] In the embodiment of the present application, the energy storage state feature extraction module 140 is used to arrange the time series of the energy storage state of the energy storage device into an energy storage state timing input matrix and then extract the energy storage state timing feature to obtain the energy storage state timing feature matrix. Specifically, in the embodiment of the present application, the energy storage state feature extraction module is used to: arrange the time series of the energy storage state of the energy storage device into the energy storage state timing input matrix according to the time dimension and the sample dimension and then input the energy storage state timing feature extractor based on the convolutional neural network model to obtain the energy storage state timing feature matrix. It should be understood that considering that the energy storage state of the energy storage device changes with time, and the energy storage states of the energy storage device (battery charging state, discharge state, temperature, voltage and current) also have mutual correlation and influence relationship in timing. Therefore, in order to analyze and understand the relationship and timing characteristics between these states more accurately in the future, in the technical solution of the present application, the time series of the energy storage state of the energy storage device is arranged into the energy storage state timing input matrix according to the time dimension and the sample dimension. Furthermore, considering that the energy storage state timing input matrix contains implicit feature information about the energy storage state, for example, it reveals the battery's charge and discharge periodicity, such as daily cycles, weekly cycles or seasonal changes, and the battery's usage intensity, that is, whether the battery is lightly used or frequently charged and discharged with high intensity. Based on this, in the technical solution of the present application, the energy storage state timing input matrix is input into an energy storage state timing feature extractor based on a convolutional neural network model to obtain the energy storage state timing feature matrix, so as to capture the implicit feature information contained in the energy storage state timing input matrix, thereby enhancing the other expression ability of the energy storage state timing feature matrix.
[0054] In an embodiment of the present application, the feature fusion module 150 is used to obtain an energy demand fusion feature matrix based on the power generation-load demand time series correlation feature matrix and the energy storage state time series feature matrix. Accordingly, considering that the power generation-load demand time series correlation feature matrix and the energy storage state time series feature matrix respectively express the characteristic correlation relationship between the power generation and load demand of the energy storage device in time series and the characteristic information of the battery energy storage state in time series. Therefore, in order to more comprehensively and accurately predict the energy demand in the short term in the future, in the technical solution of the present application, the energy demand fusion feature matrix is obtained based on the power generation-load demand time series correlation feature matrix and the energy storage state time series feature matrix, so that the characteristic information of the energy demand can be comprehensively characterized to make a more accurate energy demand prediction. In particular, in a specific embodiment of the present application, the energy demand fusion feature matrix is obtained based on the power generation-load demand time series correlation feature matrix and the energy storage state time series feature matrix, including: cascading the power generation-load demand time series correlation feature matrix and the energy storage state time series feature matrix to obtain the energy demand fusion feature matrix.
[0055] In the embodiment of the present application, the feature optimization module 160 is used to perform similarity information correction based on the measurement dimension on the energy demand fusion feature vector expanded from the energy demand fusion feature matrix to obtain an optimized energy demand fusion feature vector.
[0056] In particular, in the technical solution of the present application, it is considered that factors such as sensor accuracy, equipment failure or environmental interference may cause the collected data to contain noise and outliers. The timing encoder may not accurately reflect the long-term dependencies and periodic patterns in the time series data. Changes in the battery's charging, discharging, temperature, voltage and current states may cause dynamic fluctuations in the feature vector. Changes in the external environment and fluctuations in load demand may affect the operation of the energy storage device, thereby affecting the energy demand fusion feature vector finally obtained. As a result, the energy demand fusion feature vector finally obtained has a local abnormal distribution, which in turn causes the decoder to have a category outlier offset relative to the decoder when it is decoded and judged by the decoder, affecting the decoding iteration effect. Specifically, the local abnormal distribution may cause the decoder or decoder to learn the wrong pattern during the training process, resulting in predictions or judgments biased towards outliers. In order to solve this problem, in the technical solution of the present application, the energy demand fusion feature vector is corrected based on the similarity information of the metric dimension.
[0057] Among them, the energy demand fusion feature vector is corrected for similarity information based on the metric dimension to obtain an optimized energy demand fusion feature vector, including: determining a decoding weight matrix of a decoder; performing matrix decoupling on the decoding weight matrix in units of row vectors to obtain a set of decoding weight vectors; calculating an energy demand fusion metric learning loss information item between each decoding weight vector in the set of decoding weight vectors and the energy demand fusion feature vector to obtain an energy demand fusion metric learning compensation representation vector; inputting the energy demand fusion metric learning compensation representation vector into a softmax activation function to obtain a probabilistic energy demand fusion metric learning compensation weight vector; calculating the positional point multiplication of the probabilistic energy demand fusion metric learning compensation weight vector and the energy demand fusion feature vector to obtain a point product feature vector, and calculating the weighted sum of the point product feature vector and the energy demand fusion feature vector to obtain the optimized energy demand fusion feature vector.
[0058] The energy demand fusion metric learning loss information item between each decoding weight vector in the set of decoding weight vectors and the energy demand fusion feature vector is calculated by the following compensation formula to obtain an energy demand fusion metric learning compensation representation vector;
[0059] Wherein, the compensation formula is:
[0060]
[0061] in, represents the first in the set of decoding weight vectors The decoding weight vector The eigenvalues at the positions, The energy demand fusion feature vector is represented by The eigenvalues at the positions, represents the length of the feature vector, represents the energy demand fusion metric learning compensation representation vector The feature value at each position.
[0062] That is, the energy demand fusion feature vector is corrected based on the similarity information of the metric dimension, which first determines the decoding weight matrix of the decoder to understand the decoder distinguishing different features. Subsequently, the decoding weight matrix is decoupled by row vectors to obtain a set of decoding weight vectors to allow the model to independently analyze the weight of each category and provide more refined control for feature compensation. Next, the energy demand fusion metric learning loss information item is calculated. This process involves quantitative analysis of the interaction between each weight vector and the energy demand fusion feature vector. Through the calculation of a specific loss function, the inconsistency between the feature vector and the weight information is captured, providing a quantitative indicator for compensation. Next, the energy demand fusion metric learning compensation representation vector is input into the activation function, nonlinearly transformed, and converted into a probabilistic form. The application of the activation function introduces nonlinearity, enhances the model's ability to interpret the compensation weight, and makes the compensation weight more consistent with the actual distribution characteristics. Then, the energy demand fusion metric learning compensation representation vector processed by the activation function is used to calculate the probabilistic energy demand fusion metric learning compensation weight vector. This step provides probabilistic guidance for the adjustment of the feature vector in a probabilistic form, making the optimization process more consistent with statistical characteristics, thereby enhancing the accuracy of the model's feature adjustment. Finally, by calculating the point product of the probabilistic energy demand fusion metric learning compensation weight vector and the energy demand fusion feature vector, and then calculating the weighted sum with the original energy demand fusion feature vector, the optimized energy demand fusion feature vector is obtained. This step combines the original feature information and compensation information to generate an adjusted feature vector, reducing the impact of abnormal distribution and improving the accuracy of decoding.
[0063] In an embodiment of the present application, the prediction result generation module 170 is used to obtain a prediction result based on the optimized energy demand fusion feature vector. Specifically, in an embodiment of the present application, the prediction result generation module is used to: input the optimized energy demand fusion feature vector into a decoder-based energy predictor to obtain a prediction result, and the prediction result is used to represent the predicted value of the energy demand in the future short time. That is, the prediction result of the predicted value of the energy demand in the future short time is obtained by decoding the optimized energy demand fusion feature vector. In this way, the energy demand in the short term can be accurately predicted, so as to intelligently adjust the battery charging and discharging strategy, extend the battery life, optimize the use of energy, reduce energy waste, and reduce the fluctuation of the power grid load, thereby enhancing the stability of the power grid.
[0064] In one embodiment of the present application, the optimized energy demand fusion feature vector is input into a decoder-based energy predictor to obtain a prediction result, including: using a decoder to decode and regress the optimized energy demand fusion feature vector using the following decoding formula to obtain a predicted value of energy demand in the future short term, wherein the decoding formula is ,in, is the optimized energy demand fusion feature vector, is the predicted value of energy demand in the short term in the future, is the decoding weight matrix, Represents matrix multiplication.
[0065] In summary, the communication base station energy storage management system 100 based on the embodiment of the present application is explained, which collects the time series of the power generation of the energy storage device, the load demand and the energy storage state of the battery through sensors and intelligent meters, wherein the energy storage state includes the charging state, discharging state, temperature, voltage and current of the battery, and uses artificial intelligence to perform time series feature processing and mutual correlation on the power generation of the device, the load demand and the energy storage state of the battery, so as to obtain the prediction result of the predicted value of the energy demand in the future short time. In this way, the energy demand in the short term can be accurately predicted, so as to intelligently adjust the battery charging and discharging strategy, extend the battery life, optimize the use of energy, reduce energy waste, and reduce the fluctuation of the power grid load, thereby enhancing the stability of the power grid.
[0066] Figure 4 Flow chart of the communication base station energy storage management method according to an embodiment of the present application. Figure 4As shown, according to the communication base station energy storage management method of the embodiment of the present application, it includes: S110, collecting the time series of the power generation of the energy storage device, the load demand and the energy storage state of the battery through sensors and intelligent meters, wherein the energy storage state includes the charging state, discharging state, temperature, voltage and current of the battery; S120, arranging the time series of the power generation of the energy storage device and the time series of the load demand of the energy storage device according to the time dimension into a power generation timing input vector and a load demand timing input vector; S130, performing power generation timing coding and load demand timing coding on the power generation timing input vector and the load demand timing input vector to obtain the power generation timing feature vector and the load demand timing feature vector, and then encoding the power generation timing feature vector and the load demand timing feature vector. The vector and the load demand timing feature vector are feature-correlated to obtain a power generation-load demand timing correlation feature matrix; S140, the time series of the energy storage state of the energy storage device is arranged as an energy storage state timing input matrix, and then the energy storage state timing feature is extracted to obtain an energy storage state timing feature matrix; S160, based on the power generation-load demand timing correlation feature matrix and the energy storage state timing feature matrix, an energy demand fusion feature matrix is obtained; S160, the energy demand fusion feature vector expanded by the energy demand fusion feature matrix is corrected based on the similarity information of the measurement dimension to obtain an optimized energy demand fusion feature vector; and, S170, based on the optimized energy demand fusion feature vector, a prediction result is obtained.
[0067] Here, those skilled in the art can understand that the specific operations of each step in the above communication base station energy storage management method have been referred to above. Figures 1 to 3 The description of the communication base station energy storage management system has been introduced in detail, and therefore, its repeated description will be omitted.
[0068] Although the subject matter has been described in language specific to structural features and / or methodological logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Instead, the specific features and actions described above are merely example forms of implementing the claims.
Claims
1. A communication base station energy storage management system, characterized in that: include: The communication base station parameter acquisition module is used to collect the time series of the power generation of the energy storage device, the load demand and the energy storage state of the battery through sensors and intelligent meters, wherein the energy storage state includes the charging state, discharging state, temperature, voltage and current of the battery; A parameter timing arrangement module, used for arranging the time series of the power generation of the energy storage device and the time series of the load demand of the energy storage device into a power generation timing input vector and a load demand timing input vector according to the time dimension; A feature extraction and association module is used to perform power generation time series coding and load demand time series coding on the power generation time series input vector and the load demand time series input vector to obtain a power generation time series feature vector and a load demand time series feature vector, and then perform feature association on the power generation time series feature vector and the load demand time series feature vector to obtain a power generation-load demand time series association feature matrix; An energy storage state feature extraction module, used for arranging the time series of energy storage states of the energy storage device into an energy storage state time series input matrix and then extracting energy storage state time series features to obtain an energy storage state time series feature matrix; A feature fusion module, used to obtain an energy demand fusion feature matrix based on the power generation-load demand time series correlation feature matrix and the energy storage state time series feature matrix; A feature optimization module, used for performing similarity information correction based on the measurement dimension on the energy demand fusion feature vector expanded by the energy demand fusion feature matrix to obtain an optimized energy demand fusion feature vector; The prediction result generation module is used to obtain the prediction result based on the optimized energy demand fusion feature vector.
2. The communication base station energy storage management system according to claim 1, characterized in that: The feature extraction and association module comprises: A timing encoding unit, used for inputting the power generation timing input vector and the load demand timing input vector into a power generation timing encoder and a load demand timing encoder respectively to obtain the power generation timing characteristic vector and the load demand timing characteristic vector; A time series association unit is used to perform time series association on the power generation time series characteristic vector and the load demand time series characteristic vector to obtain the power generation-load demand time series association characteristic matrix.
3. The communication base station energy storage management system according to claim 2, characterized in that: The timing association unit is used for: The power generation time series feature vector and the load demand time series feature vector are input into a feature time series association coding module to obtain the power generation-load demand time series association feature matrix.
4. The communication base station energy storage management system according to claim 3, characterized in that: The timing association unit is used for: The transposed vector of the power generation time series characteristic vector and the load demand time series characteristic vector is calculated to obtain the power generation-load demand time series correlation characteristic matrix.
5. The communication base station energy storage management system according to claim 4, characterized in that: The energy storage state feature extraction module is used to: The time series of the energy storage state of the energy storage device is arranged according to the time dimension and the sample dimension into the energy storage state time series input matrix, and then input into the energy storage state time series feature extractor based on the convolutional neural network model to obtain the energy storage state time series feature matrix.
6. The communication base station energy storage management system according to claim 5, characterized in that: The feature optimization module comprises: Determine a decoding weight matrix of a decoder; Decoupling the decoding weight matrix in units of row vectors to obtain a set of decoding weight vectors; Calculating an energy demand fusion metric learning loss information item between each decoding weight vector in the set of decoding weight vectors and the energy demand fusion feature vector to obtain an energy demand fusion metric learning compensation representation vector; Inputting the energy demand fusion metric learning compensation representation vector into a softmax activation function to obtain a probabilistic energy demand fusion metric learning compensation weight vector; The positional point multiplication of the probabilistic energy demand fusion metric learning compensation weight vector and the energy demand fusion feature vector is calculated to obtain a point product feature vector, and a weighted sum of the point product feature vector and the energy demand fusion feature vector is calculated to obtain the optimized energy demand fusion feature vector.
7. The communication base station energy storage management system according to claim 6, characterized in that: The prediction result generating module is used for: The optimized energy demand fusion feature vector is input into a decoder-based energy predictor to obtain a prediction result, and the prediction result is used to represent the predicted value of energy demand in the future short period of time.
8. A communication base station energy storage management method, characterized in that: include: Collect the time series of energy storage equipment power generation, load demand and battery energy storage status through sensors and smart meters, wherein the energy storage status includes battery charge state, discharge state, temperature, voltage and current; Arrange the time series of the power generation of the energy storage device and the time series of the load demand of the energy storage device according to the time dimension into a power generation time series input vector and a load demand time series input vector; Performing power generation time series coding and load demand time series coding on the power generation time series input vector and the load demand time series input vector to obtain a power generation time series feature vector and a load demand time series feature vector, and then performing feature association on the power generation time series feature vector and the load demand time series feature vector to obtain a power generation-load demand time series association feature matrix; Arranging the time series of the energy storage state of the energy storage device into an energy storage state time series input matrix and then extracting the energy storage state time series features to obtain an energy storage state time series feature matrix; Based on the power generation-load demand time series correlation characteristic matrix and the energy storage state time series characteristic matrix, an energy demand fusion characteristic matrix is obtained; Performing similarity information correction based on the measurement dimension on the energy demand fusion feature vector expanded from the energy demand fusion feature matrix to obtain an optimized energy demand fusion feature vector; Based on the optimized energy demand fusion feature vector, a prediction result is obtained.
9. The communication base station energy storage management method according to claim 8, characterized in that: The power generation time series input vector and the load demand time series input vector are subjected to power generation time series coding and load demand time series coding to obtain a power generation time series feature vector and a load demand time series feature vector, and then the power generation time series feature vector and the load demand time series feature vector are subjected to feature association to obtain a power generation-load demand time series association feature matrix, including: Inputting the power generation time series input vector and the load demand time series input vector into a power generation time series encoder and a load demand time series encoder respectively to obtain the power generation time series feature vector and the load demand time series feature vector; The power generation time series characteristic vector and the load demand time series characteristic vector are time-series associated to obtain the power generation-load demand time series association characteristic matrix.
