Networking charging data synchronization optimization method based on multi-level distributed computing architecture

Abstract

The invention discloses a networking charging data synchronization optimization method based on a multi-level distributed computing architecture, and the method comprises the steps: S1, collecting the real-time flow data of each toll station, and carrying out the preliminary cleaning and preprocessing; s2, performing spatial-temporal feature extraction on the preprocessed flow data by using a spatial-temporal convolutional neural network to generate spatial-temporal feature representation; s3, based on a Neural Prophet model, performing short-term and long-term traffic prediction, and generating a traffic prediction value; s4, constructing a dynamic data synchronization strategy based on the traffic prediction value; s5, the optimized data synchronization strategy is transmitted to a regional center layer, and specific execution details of data synchronization are adjusted; s6, the cloud layer generates a global optimization scheme according to the optimization information fed back by the regional center layer; and S7, continuously evaluating the execution effect of the synchronization strategy. According to the method, an efficient and scientific optimization scheme can be provided in networking charging data synchronization optimization, and remarkable technical values and economic benefits are brought to practical application.

Description

Technical Field

[0001] The present invention relates to the technical field of data synchronization optimization, and particularly to a method for optimizing network toll data synchronization based on a multi-level distributed computing architecture. Background Art

[0002] In modern network toll systems, with the continuous growth of traffic flow and the increasing complexity of toll systems, the real-time synchronization and accurate prediction of toll data have become important issues for improving system efficiency and optimizing operations. Traditional network toll data synchronization methods mostly rely on simple time-based synchronization strategies, and usually use prediction models based on time series analysis to predict the traffic flow at toll stations. However, these traditional methods often ignore the mutual influence of traffic flow between toll stations, that is, spatio-temporal dependence, resulting in the inability to effectively capture the traffic flow fluctuation rules between toll stations in practical applications, thereby affecting the accuracy and efficiency of data synchronization.

[0003] Existing traffic prediction technologies are mainly based on time series models, such as ARIMA and Exponential Smoothing. These models can perform time series prediction on the traffic flow of a single toll station, but cannot handle the spatial correlation between stations. In traditional methods, since the traffic flow prediction of each toll station is carried out independently, there are significant errors in traffic flow synchronization between stations. For example, the traffic flow fluctuations of some toll stations may be affected by the traffic flow fluctuations of neighboring stations, and traditional methods are difficult to capture this spatio-temporal correlation, resulting in the inability to fully consider the mutual dependence relationship between stations during data synchronization, thereby affecting the synchronization accuracy.

[0004] In addition, the current data synchronization strategy also has limitations. Most traditional synchronization methods rely on a fixed synchronization period, usually based on a fixed time window for data synchronization. This method ignores the dynamic variability of traffic flow. For example, during peak periods or special holidays, the traffic flow changes drastically, and the fixed synchronization period cannot effectively cope with these changes, resulting in synchronization delays or over-synchronization, and being unable to flexibly respond to various traffic flow change scenarios.

[0005] To address these problems, in recent years, more and more research has begun to explore spatio-temporal data modeling technologies based on deep learning, such as spatio-temporal convolutional neural networks and neural network models. These methods can consider both the time and space dimensions simultaneously, thereby more accurately capturing the spatio-temporal dependence of traffic flow data. For example, ST-CNN can better understand the interaction and traffic flow fluctuation rules between different toll stations by performing convolutional processing on traffic flow data in both the time and space dimensions, thereby improving the prediction accuracy. At the same time, the Neural Prophet model can combine historical data and real-time data to perform short-term and long-term traffic flow predictions, and dynamically adjust model parameters to better adapt to traffic flow changes.

[0006] Although these emerging methods have made certain progress in dealing with spatio-temporal dependencies and prediction accuracy, there are still some deficiencies in the existing technologies. First, although ST-CNN can capture spatio-temporal dependencies, in large-scale networked toll collection systems, the traffic data volume of each toll station is huge, and how to efficiently process and store these data remains a challenge. The computational complexity in existing methods is relatively high, especially during the data transmission process between the cloud and edge computing, which may cause delays and affect the efficiency of real-time data synchronization. Second, although the Neural Prophet model has made breakthroughs in traffic prediction, in practical applications, due to the traffic changes between toll stations being affected by multiple factors

[0007] Therefore, there are still deficiencies in the real-time, accuracy, and adaptability of data synchronization in the existing technologies for networked toll collection systems. In order to better cope with the spatio-temporal dependencies, dynamic variability of traffic, and challenges of system scale, there is an urgent need for a new method that can make full use of advanced deep learning technologies to improve the overall performance and data synchronization efficiency of networked toll collection systems through spatio-temporal feature extraction, traffic prediction, and optimization of dynamic synchronization strategies. Summary of the Invention

[0008] An object of the present invention is to propose an optimized method for networked toll collection data synchronization based on a multi-level distributed computing architecture. The present invention can provide an efficient and scientific optimization scheme for networked toll collection data synchronization, bringing significant technical value and economic benefits to practical applications.

[0009] The optimized method for networked toll collection data synchronization based on a multi-level distributed computing architecture according to an embodiment of the present invention includes the following steps:

[0010] S1. Collect the real-time traffic data of each toll station, and perform preliminary cleaning and preprocessing on the collected traffic data at the edge computing layer;

[0011] S2. Use a spatio-temporal convolutional neural network to extract spatio-temporal features from the preprocessed traffic data, and perform convolutional operations on both the time and space dimensions using a multi-dimensional convolutional kernel to generate spatio-temporal feature representations;

[0012] S3. Based on the Neural Prophet model, combine historical traffic data and spatio-temporal feature representations to perform short-term and long-term traffic predictions, and at the same time, through error correction and seasonal adjustment, generate traffic prediction values;

[0013] S4. Based on the traffic prediction values, construct a dynamic data synchronization strategy, optimize the time window and frequency of data synchronization according to the traffic change trends of each toll station, and dynamically adjust the data synchronization strategies of different stations;

[0014] S5. Transmit the optimized data synchronization policy to the regional center layer, and the regional center layer adjusts the specific implementation details of data synchronization according to the traffic conditions of each toll station and the synchronization policy;

[0015] S6. The cloud layer generates a global optimization plan based on the optimization information fed back by the regional center layer and executes the global data synchronization scheduling;

[0016] S7. Through the real-time data monitoring and feedback mechanism, continuously evaluate the execution effect of the synchronization policy, dynamically adjust the data synchronization policy and optimize it.

[0017] Optionally, S1 includes the following steps:

[0018] S11. In the networked toll collection system, collect the real-time traffic data of each toll station, and the traffic data includes the number of vehicles passing through each toll station, the toll amount, the traffic time period, the geographical location information of the toll station and the timestamp;

[0019] S12. Conduct a preliminary cleaning of the collected original traffic data, remove outliers and noise data, and supplement the missing data through interpolation methods;

[0020] S13. According to the geographical location information of the toll station, calibrate the traffic data of each toll station with spatial coordinates, obtain the accurate location of each toll station, and generate a traffic data set with spatial location information;

[0021] S14. Mark the timestamp for the preprocessed traffic data to form a multi-dimensional time series data set.

[0022] Optionally, S2 includes the following steps:

[0023] S21. Construct a spatio-temporal data matrix. Let the preprocessed traffic data set be where T represents the number of time steps, N represents the number of toll stations. For each toll station i and time step t, the traffic data is represented as M(t, i), and M(t, i) represents the traffic data of toll station i at time t;

[0024] S22. Define a multi-dimensional convolution kernel. Let the convolution window size in the time dimension be k t , and the convolution window size in the spatial dimension be k n , and the convolution kernel is Perform spatio-temporal convolution operation to obtain the convolution result

[0025]

[0026] where F(t ′ ,i′ ) represents the time step t ′ and site i ′ of the convolution result, is the temporal convolution weight, is the spatial convolution weight, b is the bias term, t ′ = 1, 2, …, T - k t + 1, i ′ = 1, 2, …, N - k n + 1;

[0027] S23. Use the ReLU activation function to perform non - linear activation processing on the convolution result F to obtain the activation output F ′ ;

[0028] S24. Perform a max - pooling operation on the activation output F ′ with a pooling window size of p t × p n , and perform normalization processing to obtain the normalized spatio - temporal feature representation G norm .

[0029] Optionally, the S3 includes the following steps:

[0030] S31. Combine the spatio - temporal feature representation historical traffic data real - time traffic data as the input sequence X of the Neural Prophet model t :

[0031] X t = {M historical (t - k), M real-time (t - k + 1), …, M real-time (t), G norm};

[0032] S32. Use the Neural Prophet model to predict the traffic at future times, and the prediction output is the traffic predicted for time step t + 1;

[0033] The Neural Prophet model updates its parameters by optimizing the following loss function:

[0034]

[0035] where, M true (t + 1) is the actual traffic at time step t + 1, is the traffic predicted for time step t + 1, λ is the regularization coefficient, θ is the learning parameter of the model, ‖.‖2 is the L2 regularization term;

[0036] S33. Calculate the flow prediction error

[0037]

[0038] where M true (t + 1) is the actual flow at time t + 1, is the predicted flow at time t + 1;

[0039] Input the error e(t + 1) into the error correction module for error correction. The corrected flow prediction value is:

[0040]

[0041] where R(e(t + 1), α) is a correction function based on the error e(t + 1) and the correction coefficient α;

[0042] S34. Perform seasonal adjustment on the corrected flow prediction value as follows:

[0043]

[0044] where γ is the seasonal adjustment factor, T season is the length of the seasonal cycle, used to capture seasonal fluctuations, is the flow prediction value after seasonal adjustment.

[0045] Optionally, step S4 includes the following steps:

[0046] S41. Construct a dynamic data synchronization strategy based on the generated flow prediction value The size W of the time window and the synchronization frequency F t in the synchronization strategy are determined by the following optimization function: t where

[0047]

[0048] where is the synchronized flow data, F t is the synchronization frequency, α is an adjustment factor for balancing the synchronization error and the frequency, and the optimization objective is to minimize the difference between the synchronization error and the synchronization frequency

[0049] S42. Calculate the synchronization timing and synchronization frequency for each toll station according to the output of the synchronization strategy:

[0050]

[0051] Among them, Δt sync is the synchronization delay time, τ i (t + 1) is the synchronization timing of site i, F i (t + 1) is the synchronization frequency of site i;

[0052] S43. Dynamically adjust the synchronization strategy, update the parameters of the synchronization strategy based on real-time monitoring feedback, and the real-time monitoring error is ∈(t + 1):

[0053]

[0054] Among them, is the real-time traffic data at time t + 1;

[0055] If the monitoring error ∈(t + 1) exceeds the preset threshold ∈ threshold , then optimize the synchronization strategy by adjusting the synchronization timing and frequency parameters:

[0056]

[0057] Among them, γ is the adjustment factor, is the adjusted synchronization traffic data.

[0058] Optionally, the S5 includes the following steps:

[0059] S51. Transmit the optimized data synchronization strategy to the regional center layer. The regional center layer analyzes the execution parameters of the synchronization strategy and further adjusts them according to the traffic conditions of each regional toll station, and obtains the synchronization strategy for each region as

[0060] S52. Calculate the synchronization timing τ for each toll station i within the region and the synchronization frequency F i (t + 1) according to the synchronization strategy i (t + 1);

[0061] S53. According to the global optimization plan of the regional center layer, apply the updated synchronization strategy to all toll stations within the region, and transmit the synchronization operation parameters of each station to the edge computing layer. The edge computing layer realizes the synchronization operation according to the synchronization timing τ i (t + 1) and the synchronization frequency F i (t + 1);

[0062] S54. During the synchronization execution process, if abnormal fluctuations are found in the traffic data of some regions or toll stations, the regional center layer will adjust the synchronization strategy through real-time monitoring;

[0063] During the execution of the S55 synchronization strategy, it will be dynamically updated to cope with changes in factors such as traffic fluctuations and network delays. The regional center layer adjusts the synchronization timing and frequency parameters in real time according to the feedback information.

[0064] Optionally, the S6 includes the following steps:

[0065] S61. Transmit the synchronization strategy to the cloud layer for global data synchronization scheduling. The cloud layer aggregates the synchronization strategy parameters of each region and coordinates them through a global optimization algorithm to generate a global synchronization plan.

[0066] S62. Apply the global synchronization plan to the global data synchronization scheduling system of the cloud layer. The scheduling system issues synchronization instructions to different regions and sites according to the global synchronization strategy.

[0067] S63. Continuously evaluate the effect of the global synchronization strategy through the real-time monitoring and feedback mechanism of the cloud layer.

[0068] S64. Dynamically adjust the parameters of the global synchronization strategy according to the global synchronization plan and real-time feedback.

[0069] The beneficial effects of the present invention are as follows:

[0070] (1) By introducing an optimized method for networked toll data synchronization based on a multi-level distributed computing architecture, the present invention fully solves the problems of insufficient spatio-temporal dependence processing, low traffic prediction accuracy, and fixed data synchronization strategy in the prior art, significantly improving the overall performance and data synchronization efficiency of the networked toll system. By using a spatio-temporal convolutional neural network and a Neural Prophet model, the present invention can accurately capture the spatio-temporal dependence relationship between different toll stations, enabling traffic prediction to not only depend on time series but also consider the mutual influence of traffic between stations, thereby improving the accuracy of traffic prediction.

[0071] (2) By combining a multi-level distributed computing architecture, the system can coordinate and optimize data synchronization between different levels (such as the edge computing layer, regional center layer, and cloud layer). The edge computing layer dynamically adjusts the synchronization timing and synchronization frequency according to local real-time traffic data and the global synchronization strategy, thereby achieving precise real-time data synchronization. The regional center layer and the cloud layer optimize the synchronization strategy according to the traffic characteristics and synchronization requirements of different regions to ensure flexible response in the case of large traffic fluctuations or special situations, avoiding problems such as over-synchronization or delayed synchronization.

[0072] (3) The dynamic data synchronization strategy proposed by the present invention can adjust the synchronization timing and frequency in real time in the face of traffic fluctuations compared with the fixed synchronization period of traditional methods, greatly improving the adaptability and real-time performance of synchronization. The system can dynamically adjust the synchronization strategy according to the actual situation. During peak periods or special holidays when traffic fluctuations are large, through precise optimization of synchronization timing and frequency, waste of bandwidth resources and synchronization delays are avoided, ensuring the synchronization efficiency and accuracy of data. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] The accompanying drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0074] Figure 1 is a flowchart of an optimized method for synchronized network toll data based on a multi-level distributed computing architecture proposed by the present invention;

[0075] Figure 2 is a flowchart of generating traffic prediction values by the Neural Prophet model in the optimized method for synchronized network toll data based on a multi-level distributed computing architecture proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0076] Now, the present invention will be further described in detail with reference to the accompanying drawings. These drawings are all simplified schematic diagrams, only illustrating the basic structure of the present invention in a schematic manner, so they only show the components related to the present invention.

[0077] Refer to Figure 1 - Figure 2 , an optimized method for synchronized network toll data based on a multi-level distributed computing architecture includes the following steps:

[0078] S1. Collect real-time traffic data of each toll station, and perform preliminary cleaning and preprocessing on the collected traffic data at the edge computing layer;

[0079] S2. Use a spatio-temporal convolutional neural network to extract spatio-temporal features from the preprocessed traffic data, and perform convolutional operations on both the time and space dimensions using a multi-dimensional convolutional kernel to generate a spatio-temporal feature representation;

[0080] S3. Based on the Neural Prophet model, combine historical traffic data and spatio-temporal feature representation to perform short-term and long-term traffic predictions, and generate traffic prediction values after error correction and seasonal adjustment;

[0081] S4. Based on the traffic prediction values, construct a dynamic data synchronization strategy, optimize the time window and frequency of data synchronization according to the traffic change trend of each toll station, and dynamically adjust the data synchronization strategies of different stations;

[0082] S5. Transmit the optimized data synchronization policy to the regional center layer. The regional center layer adjusts the specific execution details of data synchronization according to the traffic conditions and synchronization policy of each toll station.

[0083] S6. The cloud layer generates a global optimization plan based on the optimization information fed back by the regional center layer and executes the global data synchronization scheduling.

[0084] S7. Through the real-time data monitoring and feedback mechanism, continuously evaluate the execution effect of the synchronization policy, dynamically adjust the data synchronization policy and optimize it.

[0085] By collecting the real-time traffic data of each toll station and performing preliminary cleaning and preprocessing, the present invention can effectively remove noise data and fill in missing values, thereby ensuring the quality of the traffic data input into subsequent processing. Data preprocessing not only helps improve the reliability of the data, but also reduces the prediction errors caused by data quality problems. In addition, the addition of timestamp markers and geographical location information enables the traffic data to be combined with the dimensions of time and space, providing more comprehensive background information for subsequent spatio-temporal feature extraction and analysis.

[0086] In this embodiment, S1 includes the following steps:

[0087] S11. In the networked toll collection system, collect the real-time traffic data of each toll station. The traffic data includes the number of vehicles passing through each toll station, the toll amount, the traffic time period, the geographical location information of the toll station, and the timestamp.

[0088] S12. Perform preliminary cleaning on the collected original traffic data, remove outliers and noise data, and supplement the missing data by interpolation method.

[0089] S13. According to the geographical location information of the toll station, calibrate the spatial coordinates of the traffic data of each toll station, obtain the accurate location of each toll station, and generate a traffic data set with spatial location information.

[0090] S14. Perform timestamp marking on the preprocessed traffic data to form a multi-dimensional time series data set.

[0091] By using a spatio-temporal convolutional neural network to extract the spatio-temporal features of traffic data, the present invention can effectively capture the spatio-temporal dependence relationships between different toll stations. This process solves the defect that traditional methods cannot handle spatio-temporal correlations, enabling traffic prediction to consider not only the changes in the time dimension but also the spatial relationships between toll stations, thereby improving the accuracy and precision of prediction. The extraction of spatio-temporal features provides richer information support for subsequent traffic prediction and data synchronization policy optimization.

[0092] In this embodiment, S2 includes the following steps:

[0093] S21. Construct a spatio-temporal data matrix. Let the preprocessed traffic data set be where T represents the number of time steps, N represents the number of toll stations. For each toll station i and time step t, the traffic data is denoted as M(t, i), and M(t, i) represents the traffic data of toll station i at time t;

[0094] S22. Define a multi-dimensional convolution kernel. Let the convolution window size in the time dimension be k t , and the convolution window size in the space dimension be k n . The convolution kernel is Perform a spatio-temporal convolution operation to obtain a convolution result

[0095]

[0096] where F(t ′ , i ′ ) represents the convolution result at time step t ′ and station i ′ . is the time convolution weight, is the space convolution weight, b is the bias term, t ′ = 1, 2, …, T - k t + 1, i ′ = 1, 2, …, N - k n + 1;

[0097] S23. Use the ReLU activation function to perform non-linear activation processing on the convolution result F to obtain an activation output F ′ ;

[0098] S24. Perform a max pooling operation on the activation output F ′ , with a pooling window size of p t × p n , and perform normalization processing to obtain a normalized spatio-temporal feature representation G norm .

[0099] The present invention further enhances the prediction ability of the system through traffic prediction based on the Neural Prophet model. Combining historical data and real-time data, Neural Prophet can perform short-term and long-term traffic predictions and handle complex factors such as seasonal variations and holiday effects. This process significantly improves the flexibility and adaptability of traffic prediction, enabling it to more accurately reflect the traffic change trends in different time periods. In addition, the adaptive error correction function during the prediction process can reduce prediction errors, making the future traffic prediction results more reliable, thereby providing an accurate basis for optimizing the synchronization strategy.

[0100] In this embodiment, S3 includes the following steps:

[0101] S31. Combine spatio-temporal feature representations Historical traffic data Real-time traffic data As the input sequence X of the Neural Prophet model t :

[0102] X t = {M historical (t - k), M real-time (t - k + 1), …, M real-time (t), G norm};

[0103] S32. Use the Neural Prophet model to predict the traffic at future times, and the prediction output is The traffic predicted for time t + 1;

[0104] The Neural Prophet model updates its parameters by optimizing the following loss function:

[0105]

[0106] where, M true (t + 1) is the actual traffic at time t + 1, is the traffic predicted for time t + 1, λ is the regularization coefficient, θ is the learning parameter of the model, and ‖.‖2 is the L2 regularization term;

[0107] S33. Calculate the traffic prediction error

[0108]

[0109] where, M true (t + 1) is the actual traffic at time t + 1, is the traffic predicted for time t + 1;

[0110] The error e(t + 1) is input into the error correction module for error correction, and the corrected flow prediction value is:

[0111]

[0112] where R(e(t + 1), α) is a correction function based on the error e(t + 1) and the correction coefficient α;

[0113] S34. Perform seasonal adjustment on the corrected flow prediction value as follows:

[0114]

[0115] where γ is the seasonal adjustment factor, T season is the length of the seasonal cycle, which is used to capture seasonal fluctuations, and is the flow prediction value after seasonal adjustment.

[0116] The dynamic data synchronization strategy constructed based on the flow prediction results of the present invention can dynamically adjust the synchronization strategy according to the flow changes of different toll stations by optimizing the synchronization timing and frequency. This method overcomes the deficiencies of the traditional fixed-cycle synchronization strategy, can flexibly adapt to flow fluctuations and real-time changes, avoids resource waste or delay problems caused by the fixed synchronization frequency, and can more accurately reflect the interdependent relationship between stations by combining spatio-temporal characteristics, ensuring efficient synchronization during high-flow periods and special time periods.

[0117] In this embodiment, S4 includes the following steps:

[0118] S41. Construct a dynamic data synchronization strategy based on the generated flow prediction value The size W of the time window in the synchronization strategy t and the synchronization frequency F t are determined by the following optimization function:

[0119]

[0120] where is the synchronized flow data, F t is the synchronization frequency, α is the adjustment factor for balancing the synchronization error and the frequency, and the optimization objective is to minimize the difference between the synchronization error and the synchronization frequency

[0121] S42. Calculate the synchronization timing and synchronization frequency of each toll station according to the output of the synchronization strategy:

[0122]

[0123] Among them, Δt sync is the synchronization delay time, τ i (t + 1) is the synchronization opportunity of site i, F i (t + 1) is the synchronization frequency of site i;

[0124] S43. Dynamically adjust the synchronization strategy, update the parameters of the synchronization strategy based on real-time monitoring feedback, and the real-time monitoring error is ∈(t + 1):

[0125]

[0126] Among them, is the real-time traffic data at time t + 1;

[0127] If the monitoring error ∈(t + 1) exceeds the preset threshold ∈ threshold , then optimize the synchronization strategy by adjusting the synchronization opportunity and frequency parameters:

[0128]

[0129] Among them, γ is the adjustment factor, is the adjusted synchronization traffic data.

[0130] Through the further optimization of the synchronization strategy by the regional center layer in the present invention, the synchronization operations of each region are made more coordinated. The regional center layer can effectively adjust the synchronization strategy according to the traffic conditions and mutual dependencies of each toll station, avoiding synchronization errors between different toll stations within the region. This optimization not only improves the overall efficiency of the system but also ensures that the synchronization accuracy and real-time performance between regions are guaranteed during the multi-region coordination process, thereby improving the performance of the overall system.

[0131] In this embodiment, S5 includes the following steps:

[0132] S51. Transmit the optimized data synchronization strategy to the regional center layer. The regional center layer analyzes the execution parameters of the synchronization strategy and further adjusts them according to the traffic conditions of each regional toll station to obtain the synchronization strategy for each region as

[0133] S52. Calculate the synchronization opportunity τ for each toll station i within the region according to the synchronization strategy i (t + 1) and the synchronization frequency F i (t + 1);

[0134] S53. According to the global optimization plan of the regional center layer, transmit the updated synchronization strategy Apply to all toll stations within the region, and transmit the synchronization operation parameters of each station to the edge computing layer. The edge computing layer performs synchronization operations according to the synchronization timing τ i (t + 1) and the synchronization frequency F i (t + 1);

[0135] S54. During the synchronization execution, if abnormal fluctuations are found in the traffic data of certain regions or toll stations, the regional center layer will adjust the synchronization strategy through real-time monitoring;

[0136] S55. The synchronization strategy will be dynamically updated during execution to cope with changes in factors such as traffic fluctuations and network delays. The regional center layer adjusts the synchronization timing and frequency parameters in real time according to the feedback information.

[0137] The global synchronization scheduling of the cloud layer of the present invention performs global optimization according to the feedback information of the regional center layer, enabling the system to achieve efficient synchronization scheduling in a large-scale networked toll system. The scheduling system of the cloud layer not only precisely adjusts the synchronization timing and frequency according to the global optimization scheme, but also can flexibly adjust the synchronization strategy when the system load is high or abnormal conditions occur. Through global optimization, it can better coordinate data synchronization between different regions, improve the overall synchronization efficiency and response speed of the system, and ensure that data synchronization can operate efficiently and stably in a complex traffic and toll environment.

[0138] In this embodiment, S6 includes the following steps:

[0139] S61. Transmit the synchronization strategy to the cloud layer for global data synchronization scheduling. The cloud layer aggregates the synchronization strategy parameters of each region and coordinates them through a global optimization algorithm to generate a global synchronization plan

[0140] S62. Apply the global synchronization plan to the global data synchronization scheduling system of the cloud layer. The scheduling system issues synchronization instructions to different regions and stations according to the global synchronization strategy;

[0141] S63. Continuously evaluate the effect of the global synchronization strategy through the real-time monitoring and feedback mechanism of the cloud layer;

[0142] S64. Dynamically adjust the parameters of the global synchronization strategy according to the global synchronization plan and real-time feedback.

[0143] Example:

[0144] Example: In the networked toll collection system of a large city, the networked toll collection system of this city consists of multiple toll stations. Each toll station is responsible for a different toll collection area, and the traffic flow in each area fluctuates greatly at different time periods. Due to factors such as changes in traffic conditions, the impact of holidays, and bad weather, the traffic flow fluctuations between toll stations are often interdependent, which poses a huge challenge to traditional traffic flow prediction and data synchronization.

[0145] Before implementation, the system adopted a fixed-time window synchronization strategy. Each toll station synchronized data every 30 minutes. However, due to the failure to consider the traffic flow fluctuations of different stations, the system was unable to respond in a timely manner when the traffic flow suddenly increased, resulting in a large number of delays or duplicate synchronization problems. This not only consumed excessive bandwidth resources but also affected the overall toll collection efficiency. Specifically, during some high-traffic periods, the data transmission delay was up to 5 minutes, while during low-traffic periods, the synchronization frequency was too high, leading to a waste of bandwidth resources.

[0146] To solve this problem, the implementer collected real-time traffic flow data from each toll station at the edge computing layer of the toll collection system, including the number of passing vehicles, toll amounts, and timestamp information per minute for each toll station. Then, a spatio-temporal convolutional neural network was used to process these data to extract spatio-temporal features in order to capture the traffic flow dependence relationships between different toll stations. These spatio-temporal feature information was passed as input data to the Neural Prophet model, and the model made traffic flow predictions based on historical traffic flow data and real-time data.

[0147] Through the prediction of the Neural Prophet model, the system can accurately predict the traffic flow change trends of each toll station within the next 30 minutes. This prediction result not only considers the historical traffic flow of each toll station but also combines external factors such as holiday effects and weather changes to ensure the accuracy of the prediction.

[0148] According to the traffic flow prediction results, the system dynamically adjusts the data synchronization strategy for each toll station. For stations with higher traffic flow, the synchronization frequency will increase, while for stations with lower traffic flow, the synchronization frequency will be appropriately reduced. This synchronization strategy based on dynamic traffic flow prediction significantly reduces bandwidth waste, improves the efficiency of data synchronization, and at the same time avoids problems of over-synchronization or synchronization delay.

[0149] During the implementation process, the system not only performs global scheduling and optimization in the cloud but also performs local optimization at the regional center layer. Each region adjusts its synchronization strategy according to its traffic flow characteristics, and the cloud layer further optimizes the global synchronization scheduling according to the needs of different regions to ensure that the data synchronization of all regions is completed at the most appropriate time.

[0150] The following is a comparison table of data synchronization at a toll station before and after implementation:

[0151] Table 1 Comparison of Synchronization Effects at a Toll Station before and after Implementation

[0152]

[0153] It can be seen from the table that the system synchronization delay after implementation has been significantly reduced. Especially during peak hours, the data synchronization delay has decreased from 5.2 minutes before implementation to 2.1 minutes, a reduction of approximately 59%. This improvement has greatly enhanced the real-time response ability of the system, enabling the toll collection system to process a large amount of traffic data more efficiently.

[0154] Through the implementation of the method of the present invention, the data synchronization efficiency of the toll station has been significantly improved. The system can dynamically adjust the synchronization strategy according to real-time traffic prediction, avoiding delays and resource waste caused by traffic fluctuations in traditional synchronization methods. In addition, accurate traffic prediction and optimized synchronization strategies have greatly improved the response speed and user experience of the system. The above data fully demonstrate the beneficial effects and technical advantages of the present invention in practical applications.

[0155] As mentioned above, the above are only the preferred specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. A method for optimizing online charging data synchronization based on a multi-level distributed computing architecture, characterized in that: The steps include: S1. Collect real-time traffic data from each toll station, and perform preliminary cleaning and preprocessing on the collected traffic data at the edge computing layer; S2. Use the spatiotemporal convolutional neural network to extract spatiotemporal features of the preprocessed traffic data, and use a multi-dimensional convolution kernel to perform convolution operations on the time and space dimensions at the same time to generate spatiotemporal feature representation; S3, based on the Neural Prophet model, combines historical traffic data and spatiotemporal feature representation to make short-term and long-term traffic forecasts, and generates traffic forecast values ​​after error correction and seasonal adjustment; S4. Based on the traffic forecast value, a dynamic data synchronization strategy is constructed. According to the traffic change trend of each toll station, the time window and frequency of data synchronization are optimized, and the data synchronization strategy of different stations is dynamically adjusted; S5. The optimized data synchronization strategy is transmitted to the regional center layer, and the regional center layer adjusts the specific execution details of data synchronization according to the traffic conditions and synchronization strategy of each toll station; S6. The cloud layer generates a global optimization plan based on the optimization information fed back by the regional center layer and performs global data synchronization scheduling; S7. Through real-time data monitoring and feedback mechanism, continuously evaluate the execution effect of the synchronization strategy, dynamically adjust the data synchronization strategy and optimize it.

2. The method for optimizing online charging data synchronization based on a multi-level distributed computing architecture according to claim 1, characterized in that: The S1 comprises the following steps: S11. In the networked toll collection system, real-time traffic data of each toll station is collected, wherein the traffic data includes the number of vehicles passing through each toll station, the toll amount, the traffic time period, the geographical location information of the toll station and the timestamp; S12. Perform preliminary cleaning on the collected raw traffic data, remove outliers and noise data, and supplement the missing data through interpolation methods; S13, according to the geographical location information of the toll station, the flow data of each toll station is spatially calibrated, and the accurate location of each toll station is obtained to generate a flow data set with spatial location information; S14. Timestamp the preprocessed traffic data to form a multidimensional time series data set.

3. The method for optimizing online charging data synchronization based on a multi-level distributed computing architecture according to claim 1, characterized in that: The S2 comprises the following steps: S21, construct a spatiotemporal data matrix, assuming that the preprocessed traffic data set is Where T represents the number of time steps, N represents the number of toll stations, and for each toll station i and time step t, the flow data is expressed as M(t,i), where M(t,i) represents the flow data of toll station i at time t; S22. Define a multidimensional convolution kernel and set the time dimension convolution window size to k t , the spatial dimension convolution window size is k n , the convolution kernel is Perform spatiotemporal convolution operations to obtain convolution results Among them, F(t ′ ,i ′ ) represents the time step t ′ and site i ′ The convolution result is is the temporal convolution weight, is the spatial convolution weight, b is the bias term, t ′ =1,2,…,Tk t +1,i ′ =1,2,…,Nk n +1; S23, use the ReLU activation function to perform nonlinear activation processing on the convolution result F to obtain the activation output F ′ ; S24, to activate output F ′ Perform the maximum pooling operation with a pooling window size of p t ×p n , and normalize it to obtain the normalized spatiotemporal feature representation G norm .

4. The method for optimizing online charging data synchronization based on a multi-level distributed computing architecture according to claim 1, characterized in that: The S3 comprises the following steps: S31. Combining spatiotemporal features Historical traffic data Real-time traffic data As the input sequence X of the Neural Prophet model t : X t ={M historical (t-k),M real-time (t-k+1),…,M real-time (t),G norm }; S32. Use the Neural Prophet model to predict the traffic flow at future times. The prediction output is: is the traffic volume predicted at time t+1; The Neural Prophet model updates its parameters by optimizing the following loss function: Among them, M true (t+1) is the actual flow rate at time t+1, is the traffic volume predicted at time t+1, λ is the regularization coefficient, θ is the learning parameter of the model, and ‖.‖2 is the L2 regularization term; S33. Calculate flow prediction error Among them, M true (t+1) is the actual flow rate at time t+1, is the traffic volume predicted at time t+1; The error e(t+1) is input into the error correction module for error correction. The corrected flow prediction value is for: Wherein, R(e(t+1),α) is a correction function based on the error e(t+1) and the correction coefficient α; S34, the corrected flow prediction value To make seasonal adjustments: Among them, γ is the seasonal adjustment factor, T season is the length of the seasonal cycle, To capture seasonal fluctuations, is the seasonally adjusted flow forecast value.

5. The method for optimizing online charging data synchronization based on a multi-level distributed computing architecture according to claim 1, characterized in that: The S4 comprises the following steps: S41. Based on the generated traffic prediction value Construct a dynamic data synchronization strategy, the time window size W in the synchronization strategy t and synchronization frequency F t Determined by the following optimization function: in, is the traffic data after synchronization, F t is the synchronization frequency, α is the adjustment factor between the synchronization error and the frequency, and the target S is optimized by minimizing the difference between the synchronization error and the synchronization frequency; S42. Calculate the synchronization timing and frequency of each toll station according to the output of the synchronization strategy: Among them, Δt sync is the synchronization delay time, τ i (t+1) is the synchronization timing of site i, F i (t+1) is the synchronization frequency of site i; S43, dynamically adjust the synchronization strategy, update the parameters of the synchronization strategy based on real-time monitoring feedback, and the real-time monitoring error is ∈(t+1): in, It is the real-time traffic data at time t+1; If the monitoring error ∈(t+1) exceeds the preset threshold ∈ threshold , optimize the synchronization strategy by adjusting the synchronization timing and frequency parameters: Among them, γ is the adjustment factor, It is the adjusted synchronous traffic data.

6. The method for optimizing online charging data synchronization based on a multi-level distributed computing architecture according to claim 1, characterized in that: The S5 comprises the following steps: S51. The optimized data synchronization strategy is transmitted to the regional center layer. The regional center layer analyzes the execution parameters of the synchronization strategy and makes further adjustments according to the traffic conditions of the toll stations in each region. The synchronization strategy for each region is obtained as follows: S52, according to the synchronization strategy For each toll station i in the area, the synchronization timing τ i (t+1) and synchronization frequency F i (t+1) is calculated; S53, according to the global optimization scheme of the regional center layer, the updated synchronization strategy It is applied to all charging stations in the area, and the synchronization operation parameters of each station are transmitted to the edge computing layer, which is based on the synchronization timing τ i (t+1) and synchronization frequency F i (t+1) realizes synchronous operation; S54. During the synchronization process, if abnormal fluctuations are found in the traffic data of certain areas or toll stations, the regional center layer will adjust the synchronization strategy through real-time monitoring; S55. The synchronization strategy will be dynamically updated during the execution process to cope with changes in factors such as traffic fluctuations and network delays. The regional center layer will adjust the synchronization timing and frequency parameters in real time based on feedback information.

7. The method for optimizing online charging data synchronization based on a multi-level distributed computing architecture according to claim 1, characterized in that: The S6 comprises the following steps: S61. Synchronize strategy The data is transmitted to the cloud layer for global data synchronization scheduling. The cloud layer summarizes the synchronization strategy parameters of each region and coordinates them through a global optimization algorithm to generate a global synchronization plan. S62, the global synchronization scheme A global data synchronization scheduling system applied to the cloud layer. The scheduling system issues synchronization instructions to different regions and sites according to the global synchronization strategy. S63. Continuously evaluate the effectiveness of the global synchronization strategy through real-time monitoring and feedback mechanisms at the cloud layer; S64. Dynamically adjust the parameters of the global synchronization strategy according to the global synchronization plan and real-time feedback.

Images