A base station special-purpose zero-carbon intelligent dispatching management system

Abstract

The present application relates to the technical field of base station scheduling management, and particularly relates to a base station special zero-carbon intelligent scheduling management system, which comprises: analyzing base station load risks caused by frequent switching of vehicle flow connection based on operation data information and energy environment data in the operation process of the base station; analyzing connection competition risks of the base station in traffic congestion in combination with scheduling configuration information and real-time traffic data; constructing a base station operation risk quantification model based on the analysis results of the base station load risks caused by frequent switching of vehicle flow connection and the analysis results of the connection competition risks of the base station in traffic congestion, and evaluating operation risks in the operation process of each base station in the traffic network; and adjusting the scheduling management scheme of the base station according to the evaluation results of the operation risks in the operation process of each base station in the traffic network, so that early warning of potential overload risks is realized, and the reliability of base station operation is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of base station scheduling management, and particularly relates to a base station special zero-carbon intelligent scheduling management system. BACKGROUND

[0002] As the key infrastructure supporting modern digital society and ensuring smooth mobile communication, the stable operation of communication base stations is directly related to the efficiency of regional economic development and the reliability of public communication services. In the process of continuous operation, base stations not only face the increasing pressure of data traffic, leading to the continuous rise of energy consumption, which seriously affects the operation cost and environmental sustainability, but also face the dynamic changes of traffic flow, which will cause the load of base stations to fluctuate dramatically, from the initial increase of signaling pressure, to the intensification of resource competition in the middle stage, to the decline of service quality in the later stage. If these operation risks cannot be timely warned and scheduled, it may cause network congestion, connection interruption and other communication accidents, which not only affects the user experience, but also hinders the development of key applications such as smart city and vehicle networking. Traditional base station operation and maintenance management relies on static thresholds and manual strategy adjustment, which not only has a lagging response, but also is difficult to accurately adapt to the scheduling decision in the face of complex traffic environment and dynamic business load. Using big data analysis technology for operation and maintenance management can analyze the running state of base stations in real time, including load fluctuation, energy consumption change and other indicators that are difficult for humans to analyze in real time, get rid of the limitations of experience-based decision-making, and improve network management efficiency.

[0003] However, the prior art does not deeply correlate and cooperatively analyze the signaling load risk of the base station and the vehicle connection competition risk caused by traffic congestion when performing operation and maintenance management on the base station. For example, when the load of the base station increases due to frequent switching, if the pressure of connection request caused by the congestion of the adjacent road is not considered, it is difficult to accurately evaluate the comprehensive operation risk and the potential service interruption probability of the base station, thereby leading to one-sidedness of the scheduling decision. At the same time, the prior art also lacks accurate quantification of the dynamic correlation between base station energy consumption and business load, which makes it difficult for the scheduling strategy to achieve fine energy saving while ensuring service quality, and cannot provide comprehensive and optimized decision basis for intelligent scheduling under the zero-carbon goal.

[0004] In order to solve these problems, the present application designs a base station special zero-carbon intelligent scheduling management system. SUMMARY

[0005] The purpose of the present application is to provide a base station special zero-carbon intelligent scheduling management system, which acquires base station operation, scheduling configuration, energy environment and real-time traffic data, respectively analyzes the load risk caused by switching and the connection competition risk caused by congestion, constructs an operation risk quantification model for evaluation, and dynamically adjusts the base station scheduling scheme based on the evaluation results, thereby realizing the zero-carbon intelligent scheduling management of the base station.

[0006] The present application is implemented as follows:

[0007] In a first aspect, the present application provides a base station dedicated zero-carbon intelligent scheduling management system, comprising a multi-source information acquisition module, a connection switching detection module, a congestion influence detection module, a base station operation evaluation module, and a scheduling adjustment module.

[0008] The multi-source information acquisition module is configured to acquire operation data information and scheduling configuration information during base station operation, as well as energy environment data, and simultaneously acquire real-time traffic data of a traffic network in which the base station is located.

[0009] The connection switching detection module is configured to analyze base station load risks caused by frequent switching of vehicle-mounted traffic connections based on operation data information and energy environment data during base station operation.

[0010] The congestion influence detection module is configured to analyze connection competition risks of the base station during traffic congestion by combining scheduling configuration information and real-time traffic data.

[0011] The base station operation evaluation module is configured to construct a base station operation risk quantification model based on analysis results of base station load risks caused by frequent switching of vehicle-mounted traffic connections and analysis results of connection competition risks of the base station during traffic congestion, and evaluate operation risks of each base station in the traffic network during operation.

[0012] The scheduling adjustment module is configured to adjust a scheduling management scheme of the base station according to evaluation results of operation risks of each base station in the traffic network during operation.

[0013] As a preferred embodiment of the present application, the analysis of base station load risks caused by frequent switching of vehicle-mounted traffic connections based on operation data information and energy environment data during base station operation comprises the following specific steps:

[0014] S21. Extract switching event time sequence and performance monitoring data of each base station in the traffic network in a current monitoring period from operation data information during base station operation, and simultaneously extract real-time power consumption data of each base station in the current monitoring period from energy environment data; and perform time sequence alignment on the switching event time sequence and performance monitoring data of each base station in the current monitoring period and the real-time power consumption data.

[0015] S22, set a fixed time length sliding time window for the current monitoring period, and scan the above various types of data aligned in time sequence at a preset step; based on the switching event time sequence of each base station in the current monitoring period, in each time window, calculate the ratio of the total number of switching events occurring in the window to the window length, to obtain the switching frequency per unit time of each base station in each time window; divide the switching frequency per unit time of each base station by the average switching frequency per unit time of all base stations to obtain the signaling relative pressure state of each base station in each time window;

[0016] S23, using the sliding standard deviation algorithm, based on the performance monitoring data of each base station in the current monitoring period, extracting the central processor utilization rate sequence of each base station; calculate the standard deviation of the central processor utilization rate sequence of each base station in each time window, and calculate the Gini coefficient of the central processor utilization rate sequence of each base station based on the principle of Lorenz curve, and the standard deviation and Gini coefficient of the central processor utilization rate sequence of each base station are weighted and summed according to the preset weight to generate the load fluctuation entropy value of each base station in each time window; divide the load fluctuation entropy value of each base station in each time window by the average load fluctuation entropy value of all base stations to obtain the load relative fluctuation state of each base station in each time window;

[0017] S24, based on the operation data information and energy environment data in the running process of the base station, a base station energy consumption analysis model is constructed to quantify the switching related energy consumption increment of each base station in each time window;

[0018] S25, the signaling relative pressure state, the load relative fluctuation state and the switching related energy consumption increment of each base station in each time window are respectively processed by minimum-maximum standardization, and the signaling relative pressure state, the load relative fluctuation state and the switching related energy consumption increment of each base station in each time window after standardization are weighted and summed to obtain the base station load risk caused by frequent switching of vehicle flow connection in the running process of each base station in each time window.

[0019] S26, the base station load risk caused by frequent switching of vehicle flow connection in the running process of each base station in each time window is imported into the pre-constructed base station load risk time sequence prediction model, and the base station load risk caused by frequent switching of vehicle flow connection in the running process of each base station in the next time window is output.

[0020] As a preferred embodiment of the present application, the base station energy consumption analysis model is constructed to quantify the switching related energy consumption increment of each base station in each time window, which specifically includes the following steps:

[0021] S241, acquire operation data information and energy environment data in the running process of the base station, take the unit time switching frequency in each time window in the current monitoring period, the average user session time, the number of activated users at the starting time of the time window in the operation data information, and the absolute increment of real-time power consumption in each time window in the current monitoring period in the energy environment data as the regression analysis data set, and divide the regression analysis data set into a regression analysis training set and a regression analysis verification set;

[0022] S242, construct a gradient boosting decision tree regression model, take the unit time switching frequency, the average user session time, and the number of activated users at the starting time of the window in each time window in the regression analysis training set as the input features of the gradient boosting decision tree regression model, take the absolute increment of real-time power consumption in each time window in the regression analysis training set as the output target of the gradient boosting decision tree regression model, train the gradient boosting decision tree regression model, and obtain an initial regression analysis model;

[0023] S243, perform model verification on the initial regression analysis model through the regression analysis verification set, and output the initial regression analysis model with a preset model accuracy as the base station energy consumption analysis model;

[0024] S244, based on steps S24-S26, acquire the base station energy consumption analysis model of each base station; adopt a Shapley value attribution analysis method to quantify the contribution degree of each input feature to the absolute increment of real-time power consumption in the base station energy consumption analysis model of each base station, take the contribution degree of the unit time switching frequency as the switching-related weight of each base station, and multiply the switching-related weight of each base station by the absolute increment of real-time power consumption of each base station in each time window to obtain the switching-related energy consumption increment of each base station in each time window.

[0025] As a preferred embodiment of the present application, the connection competition risk of the base station in the traffic jam is analyzed in combination with the scheduling configuration information and the real-time traffic data, including the following specific steps:

[0026] S31, acquire the coverage range geographic information of each base station from the scheduling configuration information in the running process of the base station, and acquire the geographic information and road section grade of each road section in the traffic network where each base station is located, and the average traffic flow data of each road section as the spatio-temporal correlation data set from the real-time traffic data;

[0027] S32, the coverage range geographic information of each base station in the space-time correlation data set is spatially superimposed and analyzed with the geographic information of each road section in the traffic road network, all traffic road sections in the coverage range of each base station are identified; according to the road section level, the weight of each level of road section is allocated, and the weight of all road sections in the coverage range is normalized to obtain the normalized weight coefficient of each road section relative to the corresponding service base station; wherein the road section level includes main road section and secondary road section; based on the normalized weight coefficient, the real-time average traffic flow data of all road sections in each time window in the coverage range of each base station is weighted and fused to calculate the average traffic flow density in the coverage range of each base station in each time window;

[0028] S33, based on the time window length in the current monitoring period, the historical time window is divided in the historical monitoring period, the number of users, the average traffic flow density, the time segment feature, the week type feature and the holiday flag of each base station in each historical time window in the historical monitoring period are obtained; the number of users, the average traffic flow density, the time segment feature, the week type feature and the holiday flag of each base station in each historical time window are used as the user prediction data set, and the user prediction data set is divided into user prediction training set and user prediction verification set.

[0029] As a preferred embodiment of the application, the connection competition risk of the base station in the traffic jam is analyzed in combination with the scheduling configuration information and the real-time traffic data, and the following specific steps are further included:

[0030] S34, a long short-term memory network model is constructed, the number of users, the average traffic flow density, the time segment feature, the week type feature and the holiday flag of each base station in each historical time window are used as the input features of the long short-term memory network model based on the user prediction training set, the number of users of each base station in the next historical time window in the user prediction training set is used as the output target of the long short-term memory network, the long short-term memory network model is trained to obtain an initial user number prediction model; the initial user number prediction model is verified through the user prediction verification set, and the initial user number prediction model with a preset second model accuracy is output as the user number prediction model;

[0031] S35, the average traffic flow density, the time segment feature, the week type feature and the holiday flag of each base station in each time window in the current monitoring period are input into the user number prediction model, and the predicted number of users of each base station in the next time window is output; the maximum connection user capacity of each base station is obtained from the scheduling configuration information, and the ratio of the predicted number of users of each base station in the next time window to the maximum connection user capacity is used as the connection competition intensity of each base station in the next time window;

[0032] S36, divide the connection competition strength of each base station in the next time window by the average of the connection competition strength of all base stations in the next time window, to obtain the connection competition risk of each base station in the next time window.

[0033] As a preferred embodiment of the present application, a base station operation risk quantification model is constructed to evaluate the operation risk of each base station in the traffic network during operation, including the following specific steps:

[0034] S41, extract the base station load risk and connection competition risk caused by frequent switching of vehicle flow connection during operation of each base station in the next time window;

[0035] S42, weight and sum the base station load risk and connection competition risk caused by frequent switching of vehicle flow connection during operation of each base station in the next time window, to obtain the operation risk of each base station in the traffic network during operation in the next time window.

[0036] As a preferred embodiment of the present application, according to the evaluation result of the operation risk of each base station in the traffic network during operation, the scheduling management scheme of the base station is adjusted, including the following specific contents:

[0037] Obtain the operation risk of all base stations in the same coverage area in the traffic network during operation in the next time window, and preferentially schedule the corresponding base station receiving the newly established connection request in the coverage area with the smallest operation risk.

[0038] In a second aspect, the present application provides a base station special zero-carbon intelligent scheduling management method, including the following specific steps:

[0039] Obtain operation data information and scheduling configuration information of the base station during operation, and energy environment data, and obtain real-time traffic data of the traffic network where the base station is located;

[0040] Based on the operation data information and energy environment data of the base station during operation, analyze the base station load risk caused by frequent switching of vehicle flow connection;

[0041] Combine the scheduling configuration information and real-time traffic data to analyze the connection competition risk of the base station during traffic congestion;

[0042] Based on the analysis result of the base station load risk caused by frequent switching of vehicle flow connection and the analysis result of the connection competition risk of the base station during traffic congestion, a base station operation risk quantification model is constructed to evaluate the operation risk of each base station in the traffic network during operation;

[0043] According to the evaluation result of the operation risk of each base station in the traffic network during operation, the scheduling management scheme of the base station is adjusted.

[0044] In a third aspect, the present application provides an electronic device, comprising a processor and a memory, wherein the memory stores a computer program that can be invoked by the processor, and the processor executes a base station special zero-carbon intelligent scheduling management method by invoking the computer program stored in the memory.

[0045] Compared with the prior art, the present application has the following advantages and beneficial effects:

[0046] 1. The present application analyzes the base station load risk caused by frequent switching of vehicle-mounted traffic connection, realizes early warning of potential overload risk of the base station, and improves the reliability of base station operation;

[0047] 2. The present application combines scheduling configuration and real-time traffic data analysis to connect competition risk, dynamically associates traffic state with base station connection capacity, realizes accurate prediction of network congestion risk, and guarantees service quality and user experience under high business load;

[0048] 3. The present application constructs a base station operation risk quantification model and dynamically adjusts the scheduling scheme based on the risk evaluation result, realizes adaptive optimization allocation of network resources, effectively reduces the overall energy consumption while guaranteeing the service performance, and effectively supports the zero-carbon operation target of the communication network. BRIEF DESCRIPTION OF DRAWINGS

[0049] Other features, objects and advantages of the present application will become more apparent from the following detailed description of non-limiting embodiments, made with reference to the accompanying drawings:

[0050] Figure 1 FIG. 1 is a schematic diagram of the overall flow of a base station special zero-carbon intelligent scheduling management method of the present application;

[0051] Figure 2 FIG. 2 is a structural schematic diagram of a base station special zero-carbon intelligent scheduling management system of the present application;

[0052] Figure 3 FIG. 3 is an analysis flowchart of step S2 of a base station special zero-carbon intelligent scheduling management method of the present application;

[0053] Figure 4 FIG. 4 is an analysis flowchart of step S3 of a base station special zero-carbon intelligent scheduling management method of the present application. DETAILED DESCRIPTION

[0054] The technical solutions of the present application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solutions of the present application, and are not limitations of the technical solutions of the present application. In the case of no conflict, the technical features in the embodiments of the present application and the embodiments can be combined with each other.

[0055] Embodiment 1

[0056] As Figure 1 shown, the embodiment provides a base station dedicated zero-carbon intelligent scheduling management method, comprising the following specific steps:

[0057] S1, obtaining operation data information and scheduling configuration information in the base station running process and energy environment data, and simultaneously obtaining real-time traffic data of the traffic network where the base station is located;

[0058] S2, based on the operation data information and the energy environment data in the base station running process, analyzing the base station load risk caused by frequent switching of vehicle flow connection;

[0059] S3, combining the scheduling configuration information and the real-time traffic data, analyzing the connection competition risk of the base station in traffic congestion;

[0060] S4, based on the analysis results of the base station load risk caused by frequent switching of vehicle flow connection and the analysis results of the connection competition risk of the base station in traffic congestion, constructing a base station operation risk quantification model to evaluate the operation risk of each base station in the traffic network in the running process;

[0061] S5, according to the operation risk evaluation results of each base station in the traffic network in the running process, adjusting the scheduling management scheme of the base station.

[0062] In this embodiment, as Figure 3 shown, in step S2, based on the operation data information and the energy environment data in the base station running process, the base station load risk caused by frequent switching of vehicle flow connection is analyzed, comprising the following specific steps:

[0063] S21, from the operation data information in the base station running process, extracting the switching event time sequence and performance monitoring data of each base station in the traffic network in the current monitoring period; simultaneously extracting the real-time power consumption data of each base station in the current monitoring period from the energy environment data; time aligning the switching event time sequence and performance monitoring data and real-time power consumption data of each base station in the current monitoring period;

[0064] In this embodiment, the handover event timing sequence refers to the sequence of time points recorded by the base station when the user equipment switches from one base station coverage area to another during movement. This is collected in real time by the base station's wireless resource management module and includes the handover request time, handover completion time, handover type, and related user identifier. The performance monitoring data includes the central processor utilization rate of the base station, which is obtained by periodic sampling in the base station's performance management system. The sampling frequency is usually set to once per second or dynamically adjusted according to network load to ensure real-time and accuracy of the data. In this embodiment, real-time power consumption data is sourced from the base station's energy monitoring system, collected by smart meters or embedded sensors, recording the power consumption values of the base station at different time points, usually in kilowatt-hours, for analyzing the energy efficiency and usage patterns of the base station. In this embodiment, different data sources may have different timestamp accuracy or collection intervals. Through time alignment, all data is aligned on the same time axis to accurately correlate the relationship between different variables for subsequent analysis. Specifically, the specific operation of time alignment includes standardizing all data timestamps using a time synchronization protocol, and then uniformly aligning the data to the same sampling frequency using interpolation methods, such as aligning all data to one data point per second. This embodiment provides a high-quality, consistent data foundation for subsequent sliding window analysis and risk quantification, ensuring that the analysis results accurately reflect the running state of the base station. At the same time, through alignment processing, the dynamic relationship between vehicle-mounted traffic handover events and base station load can be more accurately captured, providing data support for intelligent scheduling.

[0065] S22, set a fixed-length sliding time window for the current monitoring period, and scan the above time-aligned data of various types with a preset step length; based on the time sequence of the handover events of each base station in the current monitoring period, in each time window, calculate the ratio of the total number of handover events occurring in the window to the window length, to obtain the unit time handover frequency of each base station in each time window; divide the unit time handover frequency of each base station by the average unit time handover frequency of all base stations to obtain the signaling relative pressure state of each base station in each time window; it should be noted that the length of the sliding time window is dynamically set according to the business characteristics of the base station and the network environment, for example, the length of the sliding time window is set to 10 minutes in this embodiment to adapt to the rapid change of vehicle flow, and the preset step length determines the interval of window movement, such as moving every 1 minute, so that continuous coverage monitoring period can be realized, and key events can be avoided. Further, this embodiment quantifies the unit time handover frequency of each base station in each time window, which reflects the intensity of the base station in processing handover events in unit time, and is an important indicator for measuring the signaling load of the base station. The calculation method of the unit time handover frequency is simple and direct, but it can effectively capture the signaling pressure fluctuation of the base station, for example, in a high flow area, frequent vehicle handover will cause this value to rise sharply, prompting potential load risk. Further, this embodiment analyzes the signaling relative pressure state of each base station in each time window, eliminates the influence of network overall fluctuation, and highlights the relative load situation of a single base station in the entire road network. For example, if the unit time handover frequency of a base station is much higher than the average, its signaling relative pressure state value will be greater than 1, indicating that the base station is bearing an abnormally high signaling burden, and may need to be scheduled first. This embodiment makes the risk analysis more comparable through relative processing, and can identify abnormal base stations relative to the overall network, without being disturbed by global flow changes. At the same time, through sliding window scanning, this embodiment can track the change trend of the signaling pressure in real time, provide time sequence data basis for predicting future risks, and also lay a foundation for subsequent load fluctuation and energy consumption analysis, and enhance the adaptability of the scheduling system to dynamic network environment.

[0066] S23, based on the performance monitoring data of each base station in the current monitoring period, the central processor utilization rate sequence of each base station is extracted by using the sliding standard deviation algorithm; the standard deviation of the central processor utilization rate sequence of each base station in each time window is calculated, and the Gini coefficient of the central processor utilization rate sequence of each base station is calculated based on the principle of Lorenz curve; the standard deviation and the Gini coefficient of the central processor utilization rate sequence of each base station are weighted and summed according to the preset weight to generate the load fluctuation entropy value of each base station in each time window; the load relative fluctuation state of each base station in each time window is obtained by dividing the load fluctuation entropy value of each base station in each time window by the average value of the load fluctuation entropy values of all base stations; it should be noted that the central processor utilization rate sequence is a key indicator extracted from the performance monitoring data, which is collected by the monitoring system of the base station at a fixed interval (the default fixed interval is 1 second in this embodiment), and reflects the load condition of the processing capacity of the base station. The sliding standard deviation algorithm is used to calculate the fluctuation degree of the utilization rate sequence in each time window, and the larger the standard deviation, the more unstable the load, which may be caused by frequent switching events. The calculation of the Gini coefficient is based on the inequality measurement method in economics, which evaluates the distribution uniformity of the utilization rate sequence by constructing the Lorenz curve. Specifically, the area ratio of the cumulative distribution curve and the absolute equality line after sorting the utilization rate value is taken as the Gini coefficient, and the higher the Gini coefficient, the more uneven the load distribution, for example, the processor utilization rate peak is prominent at some time, while it is lower at other time, which may exacerbate the risk of the base station. The standard deviation and the Gini coefficient are weighted and summed to generate the load fluctuation entropy value, wherein the preset weight is usually set according to historical data or expert experience, for example, the standard deviation weight can be set to 0.6 and the Gini coefficient weight can be set to 0.4 in this embodiment, to balance the influence of fluctuation amplitude and distribution uniformity, and the higher the load fluctuation entropy value, the more intense the load fluctuation and the more uneven the distribution, indicating that the running state of the base station is unstable. Further, the embodiment comprehensively evaluates the load fluctuation characteristics of the base station by the composite index, not only focusing on the fluctuation amplitude, but also considering the imbalance of time distribution, so as to more comprehensively capture the potential risks brought by the vehicle flow switching, for example, in traffic congestion, the load of the base station may instantaneously soar, and the entropy value can timely reflect such mutation. Further, the embodiment can also dynamically adjust the weight through the machine learning model, so as to adapt to different network scenes and improve the accuracy of risk assessment.

[0067] S24, based on the operation data information and energy environment data in the running process of the base station, a base station energy consumption analysis model is constructed to quantify the switching related energy consumption increment of each base station in each time window;

[0068] In this embodiment, a base station energy consumption analysis model is constructed to quantify the switching related energy consumption increment of each base station in each time window, which specifically includes the following steps:

[0069] S241, obtaining operation data information and energy environment data in the running process of the base station, taking the unit time switching frequency in each time window in the current monitoring period, the average user session duration, the number of activated users at the starting time of the time window in the operation data information, and the absolute increment of real-time power consumption in each time window in the current monitoring period in the energy environment data as a regression analysis data set, and dividing the regression analysis data set into a regression analysis training set and a regression analysis verification set;

[0070] S242, constructing a gradient boosting decision tree regression model, taking the unit time switching frequency, the average user session duration, and the number of activated users at the starting time of the window in each time window in the regression analysis training set as input features of the gradient boosting decision tree regression model, taking the absolute increment of real-time power consumption in each time window in the regression analysis training set as the output target of the gradient boosting decision tree regression model, training the gradient boosting decision tree regression model, and obtaining an initial regression analysis model;

[0071] S243, performing model verification on the initial regression analysis model through the regression analysis verification set, and outputting the initial regression analysis model with a preset model accuracy as the base station energy consumption analysis model;

[0072] S244, based on steps S24-S26, obtaining the base station energy consumption analysis model of each base station; using a Shapley value attribution analysis method to quantify the contribution of each input feature to the absolute increment of real-time power consumption in the base station energy consumption analysis model of each base station, taking the contribution of the unit time switching frequency as the switching related weight of each base station, and multiplying the switching related weight of each base station with the absolute increment of real-time power consumption of each base station in each time window to obtain the switching related energy consumption increment of each base station in each time window.

[0073] Specifically, in this embodiment, the unit time switching frequency is calculated from step S22, which reflects the signaling load intensity; the average user session duration is extracted from the base station connection log, which represents the average connection duration of the user, and is calculated by the difference between the session start and end times; the number of active users refers to the number of users in the connected state at the start of the time window, which is obtained from the base station user management module in real time; and the absolute increment of real-time power consumption is obtained by comparing the power consumption values of adjacent time windows, which represents the energy consumption change. These data together constitute the regression analysis dataset, which is divided into a regression analysis training set and a regression analysis verification set, and the division ratio in this embodiment is defaulted to 7:3 to ensure the independence of model training and evaluation. In this embodiment, the gradient boosting decision tree regression model is an ensemble learning algorithm that fits complex nonlinear relationships by iteratively training multiple decision trees, and takes the unit time switching frequency, the average user session duration, and the number of active users at the start of the window in each time window in the regression analysis training set as input features, and takes the absolute increment of real-time power consumption in each time window in the regression analysis training set as the output target, and trains the model. In this embodiment, the model training process includes setting the number of trees, learning rate and other hyperparameters, and optimizing the performance through cross-validation to obtain an initial regression analysis model. Then, the initial regression analysis model is verified by the regression analysis verification set, and the accuracy is evaluated using indicators such as mean square error or R square value, and the model with an accuracy greater than or equal to the preset model accuracy (the preset model accuracy in this embodiment is defaulted to 90%) is output as the base station energy consumption analysis model. Finally, the Shapley value attribution analysis method is used to quantify the contribution of each input feature to the absolute increment of real-time power consumption in the base station energy consumption analysis model of each base station. Specifically, in this embodiment, the Shapley value is based on cooperative game theory, and the average marginal contribution is calculated by permutation and combination of features, and the contribution of the unit time switching frequency as an input feature is taken as the switching-related weight of each base station. Then, the switching-related weight of each base station is multiplied by the absolute increment of real-time power consumption of each base station in each time window to obtain the switching-related energy consumption increment of each base station in each time window. Based on the above, this embodiment accurately quantifies the impact of switching events on energy consumption through machine learning models, avoiding the bias that may be caused by simple correlation analysis.

[0074] S25, the signaling relative pressure state, the load relative fluctuation state and the switching related energy consumption increment of each base station in each time window are respectively processed by minimum-maximum standardization, the signaling relative pressure state, the load relative fluctuation state and the switching related energy consumption increment of each base station in each time window after the standardization processing are weighted and summed, and the base station load risk caused by frequent switching of vehicle flow connection in the running process of each base station in each time window is obtained; it should be noted that the minimum-maximum standardization in the embodiment is to convert them to the [0, 1] interval, eliminate the dimension influence, and make the weighted sum more comparable. The standardization processing formula is the difference between the original value and the minimum value divided by the difference between the maximum value and the minimum value, wherein the minimum value and the maximum value are calculated based on the data of all base stations in the current monitoring period, and the value after standardization reflects the relative position. Illustratively, the signaling relative pressure state in the embodiment can have an original value between 0.5 and 2, and after standardization, it becomes a value between 0 and 1. Similarly, other indicators are processed. After standardization processing, weighted sum is performed, and the weight is set by regression analysis on historical data. Illustratively, the signaling relative pressure state weight can be set to 0.4, the load relative fluctuation state weight can be set to 0.3, and the switching related energy consumption increment weight can be set to 0.3 according to historical data in the embodiment, to reflect the importance of different factors on the load risk. The higher the base station load risk in the embodiment indicates that the base station faces more serious load problems due to vehicle flow switching. The weight can also be dynamically adjusted according to historical fault data or real-time network state in practical application, for example, the weight of energy consumption increment is increased when energy is tight, to optimize the zero-carbon target.

[0075] S26, import the base station load risk caused by the frequent switching of the vehicle flow connection in the running process of each base station in each time window into the pre-constructed base station load risk time series prediction model, and output the base station load risk caused by the frequent switching of the vehicle flow connection in the running process of each base station in the next time window. It should be noted that the base station load risk time series prediction model is usually constructed based on time series analysis or machine learning methods, for example, using an autoregressive integrated moving average model or a long short-term memory network, to predict future values based on historical load risk data. In this embodiment, first, collect the load risk time series data of each base station in the historical monitoring period, including the risk value of each time window, to form a training data set. Then, pre-process the data, such as removing outliers and smoothing, to improve the stability of the model. During the model training process, the historical risk sequence is used as a feature, and the risk value of the next time window is used as a label to minimize the prediction error through an iterative optimization algorithm. Input the load risk value of the current time window into the model, and output the predicted value of the next time window, so as to identify potential high-risk base stations in advance; realize the forward-looking management of the load risk, and allow the scheduling system to take intervention measures before the problem occurs, such as allocating resources in advance or adjusting the flow, to reduce the probability of service interruption.

[0076] In this embodiment, as shown in Figure 4 The following specific steps are included in step S3 for analyzing the connection competition risk of the base station during traffic congestion in combination with the scheduling configuration information and real-time traffic data:

[0077] S31, obtain the coverage range geographic information of each base station from the scheduling configuration information in the base station operation process; obtain the geographic information and road section level of each road section in the traffic road network where each base station is located, and the average vehicle flow data of each road section as the space-time correlation data set from the real-time traffic data; specifically, the coverage range geographic information of each base station is obtained from the scheduling configuration information in the base station operation process, which is usually stored in the form of polygon coordinates to describe the geographic area covered by the base station signal, and the geographic information and road section level of each road section in the traffic road network where each base station is located, and the average vehicle flow data of each road section are obtained as the space-time correlation data set from the real-time traffic data. The coverage range geographic information is derived by a base station planning tool or a geographic information system, including latitude and longitude boundary and coverage radius, the road section geographic information is obtained from the traffic management department, including road section start and end point coordinates and length, the road section level is divided into main road section and secondary road section, which is based on road function and flow division, for example, expressway is main road and urban road is secondary road; the average vehicle flow data is obtained from the traffic sensor, which represents the number of vehicles passing through the road section per unit time. These data are integrated into the space-time correlation data set to establish the mapping relationship between the base station coverage and the traffic state. Through spatial data fusion, the base station risk assessment in the embodiment can be combined with real-time traffic conditions to improve accuracy. For example, in traffic congestion, vehicle concentration may cause a surge in base station connection requests, and through correlation analysis, early warning can be made.

[0078] S32, spatially superimpose the coverage range geographic information of each base station in the spatiotemporal correlation data set with the geographic information of each road segment in the traffic road network, identify all traffic road segments within the coverage range of each base station; according to the road segment level, assign a weight to each level of road segment, and normalize the weight of all road segments in the coverage range to obtain the normalized weight coefficient of each road segment relative to the corresponding service base station; wherein the road segment level includes main road segments and secondary road segments; based on the normalized weight coefficient, weight fuse the real-time average traffic flow data of all road segments within the coverage range of each base station in each time window to calculate the average traffic flow density within the coverage range of each base station in each time window; it should be noted that the spatial superposition analysis uses geographic information system algorithms such as point-in-polygon or buffer analysis to compare road segment coordinates with base station coverage polygons to determine which road segments are located within the coverage area. For example, for a base station, its coverage range may contain multiple urban roads, and through superposition analysis, these road segments can be accurately listed. According to the road segment level, a weight is assigned to each level of road segment, and the weight is set based on the importance and flow contribution of the road segment, and exemplarily, in this embodiment, the weight of the main road segment is 0.7 by default, and the weight of the secondary road segment is 0.3 by default, to reflect the difference in the influence of different road segments on the load of the base station. The weight of all road segments in the coverage range is normalized to obtain the normalized weight coefficient of each road segment relative to the corresponding service base station, wherein the normalized weight coefficient is the weight of each road segment divided by the sum of all weights, so that the weight coefficient is comparable. Based on the normalized weight coefficient, weight fuse the real-time average traffic flow data of all road segments within the coverage range of each base station in each time window to calculate the average traffic flow density within the coverage range of each base station in each time window. Wherein the higher the average traffic flow density obtained after weight fusion indicates that there are more vehicles in the coverage area and the potential connection request is more intensive. This embodiment converts the dispersed traffic data into a load indicator at the base station level through spatial weight and fusion processing, avoiding the bias that may be caused by simple summation, and normalization ensures fair comparison between different base stations. For example, in the traffic hub area, high traffic flow density may indicate high connection competition risk, and weight fusion can more accurately capture this correlation.

[0079] S33, based on the time window length in the current monitoring period, the historical time window is divided in the historical monitoring period, and the number of users, the average traffic density, the time segment feature, the week type feature and the holiday flag of each base station in each historical time window in the historical monitoring period are obtained; the number of users, the average traffic density, the time segment feature, the week type feature and the holiday flag of each base station in each historical time window are taken as a user prediction data set, and the user prediction data set is divided into a user prediction training set and a user prediction verification set. It should be noted that in this embodiment, the historical monitoring period refers to the same time period as the current period, for example, the same morning peak period, so as to ensure data comparability. The time window division is consistent with the current monitoring period, the number of users is extracted from the base station user management log, and the number of connected users is represented; the time segment feature refers to the time period in a day, for example, morning and afternoon, which is used to capture the traffic cycle mode; the week type feature is used to distinguish weekdays from weekends; and the holiday flag is used to mark whether it is a statutory holiday. These features are obtained from calendar data. These data are taken as a user prediction data set, and are divided into a user prediction training set and a user prediction verification set. In this embodiment, the division ratio is defaulted to 7:3 to support model training and evaluation. Further, this embodiment constructs a rich prediction feature set through historical data, which can capture the spatio-temporal regularity of user connection, for example, the traffic density on weekends may be different from that on weekdays, thereby improving the prediction accuracy. At the same time, data division ensures the model generalization ability and avoids overfitting.

[0080] In this embodiment, the connection competition risk of the base station in the traffic jam is analyzed in combination with the scheduling configuration information and the real-time traffic data, and the following specific steps are further included.

[0081] S34, a long short-term memory network model is constructed, based on the user prediction training set, the number of users of each base station, the average traffic density, the time segment feature, the week type feature and the holiday flag in each historical time window are taken as the input features of the long short-term memory network model, the number of users of each base station in the next historical time window in the user prediction training set is taken as the output target of the long short-term memory network, the long short-term memory network model is trained, and an initial user number prediction model is obtained; the initial user number prediction model is verified through the user prediction verification set, and an initial user number prediction model equal to or greater than a preset second model accuracy is output as the user number prediction model; in this embodiment, the long short-term memory network model is a kind of recurrent neural network model, which is good at processing time series data and capturing long-term dependencies through gating mechanism. In this embodiment, model construction includes defining network structure, such as number of hidden layers and number of neurons, setting training parameters such as learning rate and number of iterations, then using training set data to optimize weights through back propagation algorithm, minimizing prediction error, obtaining initial user number prediction model. The initial user number prediction model is verified through the user prediction verification set, the accuracy is evaluated using indicators such as mean absolute percentage error, and the model equal to or greater than the preset second model accuracy (exemplarily, the preset second model accuracy in this embodiment is 85% by default) is output as the user number prediction model. Further, the number of users is directly related to the connection load of the base station, and the deep learning model is used in this embodiment to accurately predict the future number of users, thereby indirectly evaluating the connection competition risk. For example, in traffic congestion, the number of users may increase sharply, and the model can predict this change in advance. At the same time, the long short-term memory network can process nonlinear time series patterns and adapt to complex network environments. This embodiment can provide forward-looking insights for scheduling management through advanced prediction technology, and support efficient allocation of base station resources under the zero-carbon target.

[0082] S35, input the average vehicle flow density of each base station in each time window in the current monitoring period, the time segment feature, the week type feature and the holiday flag into the user number prediction model, and output the predicted user number of each base station in the next time window; obtain the maximum connected user capacity of each base station from the scheduling configuration information, and take the ratio of the predicted user number of each base station in the next time window to the maximum connected user capacity as the connection competition intensity of each base station in the next time window; it should be noted that in the embodiment, the time segment feature, the week type feature and the holiday flag are obtained from the calendar data in real time, and these features are input into the trained user number prediction model after preprocessing, and the model outputs the predicted user number, which represents the number of devices that can be connected in the future time window. The maximum connected user capacity of each base station is obtained from the scheduling configuration information, and the maximum connected user capacity of the base station is determined by the hardware specification of the base station and stored in the configuration database. The ratio of the predicted user number of each base station in the next time window to the maximum connected user capacity is taken as the connection competition intensity of each base station in the next time window, and the higher the value, the more connection resources are nervous and the greater the competition risk is. For example, if the predicted user number is close to the maximum capacity, the connection competition intensity value will be close to 1, prompting that connection failure or service quality degradation may occur. The embodiment converts user prediction into an intuitive competition index, which is convenient for identifying high-risk base stations, and considers the hardware limitation of the base station based on the capacity ratio, so that the evaluation is more practical.

[0083] S36, divide the connection competition intensity of each base station in the next time window by the average value of the connection competition intensity of all base stations in the next time window to obtain the connection competition risk of each base station in the next time window. In the embodiment, the connection competition intensity represents the connection pressure of a single base station, and the average value of the connection competition intensity is calculated based on the values of all base stations in the entire traffic network, reflecting the overall competition level of the network. After division, the obtained connection competition risk is a relative value. If the connection competition risk is greater than 1, it indicates that the competition intensity of the base station is higher than the average level of the network, and the risk is higher. If the connection competition risk is less than 1, it indicates that the risk is lower. The embodiment eliminates the influence of global factors such as overall traffic increase on holidays through relative processing, so that the risk evaluation is more focused on the abnormal situation of individual base stations. For example, when the overall traffic is high, even if the competition intensity of a certain base station is high in absolute value, it may not be outstanding relative to the average value, thereby avoiding false positives.

[0084] In the embodiment, the base station operation risk quantification model is constructed in step S4 to evaluate the operation risk of each base station in the traffic network during operation, including the following specific steps:

[0085] S41, extract the base station load risk and connection competition risk caused by frequent switching of vehicle flow connection during operation of each base station in the next time window;

[0086] S42, weight the base station load risk and connection competition risk caused by frequent switching of vehicle flow connection in the running process of each base station in the next time window to obtain the operation risk of each base station in the traffic network in the running process of the next time window. In this embodiment, the weight is set by regression analysis on historical data. For example, based on historical data, the load risk weight can be set to 0.5 and the connection competition risk weight can be set to 0.5 to balance the influence of different risk sources, or adjusted according to real-time needs, such as increasing the load risk weight when energy is tight. The higher the operation risk is, the greater the overall operation risk of the base station is, which may need immediate intervention.

[0087] In this embodiment, the scheduling management scheme of the base station is adjusted according to the operation risk evaluation result of each base station in the traffic network in the running process in step S5, including the following specific contents:

[0088] The operation risk of all base stations in the same coverage area in the traffic network in the running process in the next time window is obtained, and the base station corresponding to the minimum operation risk is preferentially scheduled to receive the new connection request in the coverage area. In this embodiment, the same coverage area refers to the base station coverage area that overlaps in geography, which is determined by spatial analysis. The smaller the operation risk is, the more stable the base station running state is. In the scheduling, this embodiment monitors the new connection request in real time, and according to the operation risk ranking, preferentially allocates new user connection to the base station with the minimum risk, so as to balance the load and reduce the overall risk. Based on the above, this embodiment realizes dynamic load balancing, avoids overload of high-risk base stations, improves network reliability and user experience, and at the same time, by preferentially scheduling low-risk base stations, the device life can be prolonged, the energy consumption can be reduced, and the zero-carbon goal can be supported.

[0089] Embodiment 2

[0090] As shown in Figure 2 , the embodiment provides a base station special zero-carbon intelligent scheduling management system, which comprises a multi-source information acquisition module, a connection switching detection module, a congestion influence detection module, a base station operation evaluation module, and a scheduling adjustment module.

[0091] The multi-source information acquisition module is used to acquire operation data information and scheduling configuration information in the running process of the base station and energy environment data, and simultaneously acquire real-time traffic data of the traffic network where the base station is located.

[0092] The connection switching detection module is used to analyze the base station load risk caused by frequent switching of vehicle flow connection based on the operation data information and energy environment data in the running process of the base station.

[0093] The congestion influence detection module is configured to analyze the connection competition risk of the base station in combination with the scheduling configuration information and the real-time traffic data;

[0094] The base station operation evaluation module is configured to construct a base station operation risk quantification model based on the base station load risk analysis result caused by the frequent switching of the vehicle-mounted traffic connection and the connection competition risk analysis result of the base station in the traffic congestion, and evaluate the operation risk of each base station in the traffic network.

[0095] The scheduling adjustment module is configured to adjust the scheduling management scheme of the base station according to the operation risk evaluation result of each base station in the traffic network.

[0096] The above-mentioned steps of implementing the functions of the parameters and the unit modules in the base station special zero-carbon intelligent scheduling management system of the present application can refer to the parameters and steps in the embodiments of the base station special zero-carbon intelligent scheduling management method, which will not be repeated here.

[0097] Embodiment 3

[0098] The electronic device of the embodiment of the present application comprises a processor and a memory, wherein the memory stores a computer program that can be called by the processor, and the processor executes a base station special zero-carbon intelligent scheduling management method by calling the computer program stored in the memory. It should be noted that all computer programs of the base station special zero-carbon intelligent scheduling management method are implemented by using C language, wherein the multi-source information acquisition module, the connection switching detection module, the congestion influence detection module, the base station operation evaluation module, and the scheduling adjustment module are all controlled by a remote server.

[0099] In the description of the present application, the description of the terms "one embodiment", "example", "specific example" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present application, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0100] The preferred embodiments of the present application disclosed above are only used to help explain the present application. The preferred embodiments do not describe all the details and limit the present application to the specific embodiments described. Obviously, many modifications and changes can be made according to the content of the present application. The embodiments are selected and described in detail in order to better explain the principles and practical applications of the present application, so that those skilled in the art can well understand and utilize the present application. The present application is limited by the claims and their entire scope and equivalents.

Claims

1. A zero-carbon intelligent dispatch and management system for base stations, characterized in that, It includes a multi-source information acquisition module, a connection handover detection module, a congestion impact detection module, a base station operation assessment module, and a scheduling adjustment module; Among them, the multi-source information acquisition module is used to acquire operational data information, scheduling configuration information, and energy and environmental data during the operation of the base station, and at the same time acquire real-time traffic data of the transportation network where the base station is located. The connection handover detection module is used to analyze the base station load risk caused by frequent handover of vehicle traffic connections based on operational data and energy and environmental data during base station operation. The congestion impact detection module is used to analyze the connection contention risk of base stations during traffic congestion by combining scheduling configuration information and real-time traffic data. The base station operation assessment module is used to construct a quantitative model of base station operation risk based on the risk analysis results of base station load caused by frequent switching of vehicle traffic connections and the risk analysis results of base station connection competition during traffic congestion, and to assess the operation risk of each base station in the traffic network during operation. The scheduling and adjustment module is used to adjust the scheduling and management scheme of base stations based on the operational risk assessment results of each base station in the transportation network. The adjustment of the base station scheduling and management scheme based on the operational risk assessment results of each base station in the transportation network includes the following specific contents: The system acquires the operational risks of all base stations within the same coverage area in the transportation network during the next time window, and prioritizes scheduling the base station with the lowest operational risk to receive new connection requests within its coverage area.

2. The base station-specific zero-carbon intelligent dispatch management system according to claim 1, characterized in that, The analysis of base station load risks caused by frequent switching of vehicle traffic connections, based on operational data and energy and environmental data during base station operation, includes the following specific steps: S21. Extract the handover event time sequence and performance monitoring data of each base station in the traffic network during the current monitoring period from the operational data information during the base station operation process; at the same time, extract the real-time power consumption data of each base station during the current monitoring period from the energy and environmental data. The handover event timing sequence, performance monitoring data, and real-time power consumption data of each base station within the current monitoring period are time-aligned. S22. Set a fixed-duration sliding time window for the current monitoring cycle, and scan and process various types of data that have undergone time-series alignment with a preset step size; Based on the handover event time sequence of each base station in the current monitoring period, the relative signaling pressure status of each base station in each time window is analyzed. S23. Using the sliding standard deviation algorithm, extract the CPU utilization sequence of each base station from the performance monitoring data of each base station in the current monitoring period; based on the CPU utilization sequence of each base station, analyze the relative load fluctuation of each base station in each time window. S24. Based on the operational data and energy and environmental data during the operation of the base station, construct a base station energy consumption analysis model to quantify the energy consumption increment related to handover of each base station in each time window. S25. Perform minimum-maximum standardization on the relative signaling pressure state, relative load fluctuation state, and handover-related energy consumption increment of each base station in each time window. Then, perform weighted summation on the relative signaling pressure state, relative load fluctuation state, and handover-related energy consumption increment of each base station in each time window to obtain the base station load risk caused by frequent switching of vehicle traffic connections during the operation of each base station in each time window. S26. Import the base station load risk caused by frequent switching of vehicle traffic connections during the operation of each base station in each time window into the pre-built base station load risk time series prediction model, and output the base station load risk caused by frequent switching of vehicle traffic connections during the operation of each base station in the next time window.

3. The base station-specific zero-carbon intelligent dispatch management system according to claim 2, characterized in that, The construction of the base station energy consumption analysis model quantifies the handover-related energy consumption increment of each base station within each time window, specifically including the following steps: S241. Obtain operational data information and energy and environmental data during the operation of the base station. Take the unit time switching frequency, average user session duration, number of active users at the start of the time window in the current monitoring period of the operational data information, and the absolute increment of real-time power consumption in the current monitoring period of the energy and environmental data as the regression analysis dataset, and divide the regression analysis dataset into regression analysis training set and regression analysis validation set. S242. Construct a gradient boosting decision tree regression model. Use the unit time switching frequency, average user session duration, and number of active users at the start of the window in each time window of the regression analysis training set as input features of the gradient boosting decision tree regression model. Use the absolute increment of real-time power consumption in each time window of the regression analysis training set as the output target of the gradient boosting decision tree regression model. Train the gradient boosting decision tree regression model to obtain the initial regression analysis model. S243. The initial regression analysis model is validated using the regression analysis validation set, and the initial regression analysis model with an accuracy greater than or equal to the preset model is output as the base station energy consumption analysis model. S244. Based on steps S241-S243, obtain the base station energy consumption analysis model for each base station; use the Shapley value attribution analysis method to quantify the contribution of each input feature in the base station energy consumption analysis model to the absolute increment of real-time power consumption, take the contribution of the input feature as the handover frequency per unit time as the handover-related weight of each base station, and multiply the handover-related weight of each base station with the absolute increment of real-time power consumption of each base station in each time window to obtain the handover-related energy consumption increment of each base station in each time window.

4. The base station-specific zero-carbon intelligent dispatch management system according to claim 3, characterized in that, The method of combining scheduling configuration information and real-time traffic data to analyze the connection contention risk of base stations during traffic congestion includes the following specific steps: S31. Obtain the geographical coverage information of each base station from the scheduling configuration information during the base station operation process; Geographic information and road segment level of each road segment in the traffic network where each base station is located, as well as the average traffic flow data of each road segment, are obtained from real-time traffic data to form a spatiotemporal correlation dataset. S32. Spatial overlay analysis is performed on the geographical information of the coverage area of ​​each base station in the spatiotemporal correlation dataset and the geographical information of each road segment in the traffic network to identify all traffic segments within the coverage area of ​​each base station; the average traffic density within the coverage area of ​​each base station in each time window is quantified. S33. Based on the duration of the time window within the current monitoring period, divide the historical monitoring period into historical time windows, and obtain the number of users, average traffic density, time segment characteristics, weekday type characteristics, and holiday markers of each base station in each historical time window within the same historical monitoring period; use the number of users, average traffic density, time segment characteristics, weekday type characteristics, and holiday markers of each base station in each historical time window as the user prediction dataset, and divide the user prediction dataset into a user prediction training set and a user prediction validation set.

5. The base station-specific zero-carbon intelligent dispatch management system according to claim 4, characterized in that, The method of combining scheduling configuration information and real-time traffic data to analyze the connection contention risk of base stations during traffic congestion also includes the following specific steps: S34. Construct a Long Short-Term Memory (LSTM) network model. Based on the user prediction training set, use the number of users, average traffic density, time segment features, weekday type features, and holiday markers of each base station in each historical time window as input features of the LTM network model. Use the number of users of each base station in the next historical time window in the user prediction training set as the output target of the LTM network. Train the LTM network model to obtain an initial user number prediction model. Validate the initial user number prediction model through the user prediction validation set. Output an initial user number prediction model with an accuracy greater than or equal to the preset second model as the user number prediction model. S35. Input the average traffic density, time segment characteristics, weekday type characteristics and holiday signs of each base station in each time window within the current monitoring period into the user number prediction model, and output the predicted number of users of each base station in the next time window; obtain the maximum connected user capacity of each base station from the scheduling configuration information, and use the ratio of the predicted number of users of each base station in the next time window to the maximum connected user capacity as the connection competition intensity of each base station in the next time window. S36. Divide the connection competition intensity of each base station in the next time window by the average connection competition intensity of all base stations in the next time window to obtain the connection competition risk of each base station in the next time window.

6. The base station-specific zero-carbon intelligent dispatch management system according to claim 5, characterized in that, The construction of a base station operation risk quantification model to assess the operational risks of each base station in the transportation network includes the following specific steps: S41. Extract the base station load risk and connection contention risk caused by frequent switching of vehicle traffic connections during the operation of each base station in the next time window; S42. The base station load risk and connection competition risk caused by frequent switching of vehicle traffic connections during the operation of each base station in the next time window are weighted and summed to obtain the operational risk of each base station in the transportation network during the operation of each base station in the next time window.

7. A base station-specific zero-carbon intelligent scheduling management method, implemented based on any one of claims 1-6, characterized in that, The specific steps include the following: It acquires operational data, scheduling and configuration information, and energy and environmental data during the operation of the base station, and also acquires real-time traffic data of the transportation network where the base station is located. Based on operational data and energy and environmental data during base station operation, the risk of base station load caused by frequent switching of vehicle traffic connections is analyzed. By combining scheduling configuration information and real-time traffic data, the connection contention risk of base stations during traffic congestion is analyzed. Based on the risk analysis results of base station load caused by frequent switching of vehicle traffic connections and the risk analysis results of base station connection competition during traffic congestion, a quantitative model of base station operation risk is constructed to assess the operation risk of each base station in the traffic network. Based on the operational risk assessment results of each base station in the transportation network, the scheduling and management plan for the base stations is adjusted.

8. An electronic device, comprising: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; characterized in that the processor executes a base station-specific zero-carbon intelligent scheduling and management method as described in claim 7 by calling the computer program stored in the memory.

Images