Intelligent cloud check-in management system based on multi-data fusion processing

Abstract

The present invention discloses an intelligent cloud check-in management system based on multi-data fusion processing, which relates to the technical field of cloud check-in management, in order to solve the problem that the behavior of vehicle data on highways cannot be better judged and processed. The present invention can integrate different types of data such as basic data, driving data, charging data and video data through a unified data preprocessing process to form a comprehensive and multi-dimensional vehicle data view, and use the most suitable algorithm for model construction for different types of data, which can give full play to the advantages of each algorithm and improve the accuracy and adaptability of the model. In the process of abnormal data processing, different abnormal processing departments need to cooperate and coordinate with each other. This promotes communication and information sharing between departments and improves the collaborative combat capability of the entire system.

Description

Technical Field

[0001] The present invention relates to the technical field of cloud check-in management, and in particular to an intelligent cloud check-in management system based on multi-data fusion processing. Background Art

[0002] Check-in management, also known as cloud check-in management, is a self-service check-in service management model based on cloud computing technology.

[0003] The patent application with publication number CN109859337A discloses a highway intelligent check-in cloud platform system. It mainly sets up a cross-regional check-in dispatching system to realize the centralized management of all check-in personnel in the jurisdiction area, and can remotely check in online, and can quickly allocate check-in personnel online in real time according to the set priority rules. The check-in personnel can quickly handle and feedback the special situation processing results and judgment basis to the exit toll administrator in the background, avoiding the traditional situation of each toll station setting up a separate check-in room and personnel dispersion, and being unable to handle special situations in time when encountering busy business time periods during holiday peaks, effectively improving the efficiency of check-in management. Although the above patent solves the management problem, there are still the following problems in actual operation:

[0004] 1. After acquiring the vehicle data, the data is not further processed and integrated, resulting in poor quality of vehicle data, which makes it impossible to build a model later.

[0005] 2. When transmitting vehicle data, the transmission channel is not selected, resulting in slow data transmission speed, affecting subsequent event processing.

[0006] 3. Failure to select the corresponding exception handling department based on the vehicle’s abnormality type, resulting in the inability to better resolve abnormal events. Summary of the invention

[0007] The purpose of the present invention is to provide an intelligent cloud check-in management system based on multi-data fusion processing. Through a unified data preprocessing process, different types of data such as basic data, driving data, charging data and video data can be integrated together to form a comprehensive and multi-dimensional vehicle data view. The most suitable algorithm is used for model construction for different types of data, which can give full play to the advantages of each algorithm and improve the accuracy and adaptability of the model. In the process of abnormal data processing, different abnormal processing departments need to cooperate and coordinate with each other. This promotes communication and information sharing between departments, improves the collaborative combat capability of the entire system, and can solve the problems in the prior art.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] Intelligent cloud check-in management system based on multi-data fusion processing, including:

[0010] The vehicle data acquisition unit is used to:

[0011] Retrieve vehicle data from the database, classify the data according to the type of vehicle data, uniquely encode and label the classified data, and obtain the target vehicle data after uniquely encoding and labeling;

[0012] Vehicle data processing and fusion unit for:

[0013] The target vehicle data is preprocessed, and after the data preprocessing is completed, data fusion is performed to obtain target fusion data;

[0014] Data transmission processing unit, used for:

[0015] The target fusion data is transmitted to the cloud check-in control terminal, and the transmission channel with the fastest transmission speed is automatically selected during data transmission. After the target fusion data is transmitted to the cloud check-in control terminal, the model is constructed using the deep learning model. After the model is constructed, the vehicle fusion model data is obtained;

[0016] Abnormal data dispatch unit, used for:

[0017] Perform abnormal data analysis on the vehicle fusion model data, determine the abnormal area and abnormal type in the vehicle fusion model data based on the analysis results, select the abnormality handling department based on the abnormality type, and send the abnormal area data to the abnormality handling department for abnormality feedback;

[0018] Abnormal feedback operation unit, used for:

[0019] After the exception handling department has completed handling of the exception, it will describe the handling situation in the cloud check-in background, mark the handled exception as completed, and mark the unhandled exception as unhandled.

[0020] Preferably, the vehicle data acquisition unit is further used for:

[0021] Vehicle data includes basic vehicle data, video data, driving data and charging data;

[0022] Among them, basic data includes vehicle model, license plate number, vehicle identification number and owner information; video data includes videos of vehicles entering toll stations and driving on highways; driving data includes vehicle driving trajectory, driving speed and driving time; toll data includes payment data of vehicles from toll station A to toll station B;

[0023] Classify basic data, video data, driving data and charging data respectively;

[0024] After the classification is completed, the basic data, video data, driving data and charging data of the vehicle are uniquely coded and labeled;

[0025] The vehicle's license plate number and the time it entered the toll station are used as the prefix of the code, and then a hash function is used to generate the suffix of the code, and finally the vehicle's unique code number is obtained;

[0026] The vehicle data with the unique coding number is marked as the target vehicle data.

[0027] Preferably, the vehicle data processing and fusion unit comprises:

[0028] Vehicle data preprocessing module, used for:

[0029] Perform data preprocessing on target vehicle data;

[0030] The data preprocessing process is to clean the basic data, driving data and charging data, unify the data format and convert the data in turn;

[0031] Denoising the video data, dividing the video data into several segments after the noise removal, and extracting key frames from the several segments;

[0032] Image recognition technology is used to perform feature recognition of license plate numbers, vehicles, and driver faces in key frames, and data compression is performed on video data after feature recognition to obtain pre-processed video data after data compression.

[0033] Preferably, the vehicle data processing and fusion unit further includes:

[0034] Vehicle data fusion module for:

[0035] Perform data fusion on the target vehicle data after data preprocessing;

[0036] Among them, the basic data, video data, driving data and charging data in the target vehicle data are firstly associated with each other, and the data association includes timestamp association, position association and identifier association;

[0037] After data association, feature data in basic data, video data, driving data and toll data are obtained;

[0038] The feature data includes the physical attributes and configuration information of the vehicle in the basic data; the feature recognition data in the video data; the parameter data of the vehicle during operation in the driving data; the fee and toll station data in the toll data;

[0039] The feature data is fused into feature vectors through the feature layer fusion method, and the target fused data is obtained after the feature vectors are fused.

[0040] Preferably, the data transmission processing unit includes:

[0041] Converged data transmission module for:

[0042] Transmit the target fusion data to the cloud check-in control terminal for data analysis and control;

[0043] Before data transmission, the target fused data is divided into data segments, wherein the target fused data is divided into a plurality of data segments of the same length;

[0044] Determine the data transmission volume of the target fusion data according to the number of data segments;

[0045] Confirming the transmission data of the transmission channel, wherein the transmission data includes the remaining capacity of the data transmission channel and the signal strength of the transmission channel;

[0046] Selecting a transmission channel according to the data transmission amount, wherein a transmission channel whose data transmission amount is less than the remaining capacity of the transmission channel is used as a first transmission channel;

[0047] When the number of first transmission channels is greater than one, selecting the transmission channel with the strongest signal strength as the second transmission channel;

[0048] The target fusion data is transmitted to the cloud check-in control terminal through the first transmission channel or the second transmission channel.

[0049] Preferably, the target fused data is divided into data segments before data transmission, including:

[0050] Before data transmission, extracting the channel remaining capacity corresponding to all transmission channels corresponding to the data transmission;

[0051] According to the channel residual capacity corresponding to all transmission channels corresponding to the data transmission, the transmission channel comprehensive coefficient is obtained; wherein the transmission channel comprehensive coefficient is obtained by the following formula:

[0052] ;

[0053] Where G represents the comprehensive coefficient of the transmission channel; n represents the total number of transmission channels; P i represents the priority parameter corresponding to the i-th transmission channel, and the priority parameter is 1 and 1.5. When the priority is high, the priority parameter is 1.5, and when the priority is low, the priority parameter is 1; C i represents the remaining channel capacity of the i-th transmission channel; C ti represents the total channel capacity of the i-th transmission channel; η i Indicates the percentage value corresponding to the saturation of the i-th transmission channel; R irepresents the ratio between the amount of redundant data in the i-th transmission channel and the remaining channel capacity; r i represents the redundant data impact coefficient in the i-th transmission channel, and its value range is 0-1; U i represents the usage intensity corresponding to the i-th transmission channel, that is, the ratio between the usage frequency per unit time and the total usage frequency per unit time of all transmission channels; α i represents the weight coefficient corresponding to the i-th transmission channel;

[0054] Comparing the transmission channel comprehensive coefficient with a preset coefficient threshold;

[0055] When the transmission channel comprehensive coefficient is lower than the preset coefficient threshold, it indicates that the current transmission channel load is too large, and an overload warning is issued;

[0056] When the transmission channel comprehensive coefficient is not lower than a preset coefficient threshold, the target fusion data is divided into data segments.

[0057] Preferably, when the transmission channel comprehensive coefficient is not lower than a preset coefficient threshold, the target fusion data is divided into data segments, including:

[0058] When the transmission channel comprehensive coefficient is not lower than a preset coefficient threshold, extracting the channel remaining capacity corresponding to all current transmission channels;

[0059] Extract the total data volume corresponding to the target fusion data;

[0060] Obtaining a data volume setting range of a single data segment corresponding to the data segment division by using the channel remaining capacity corresponding to all current transmission channels and the total data volume corresponding to the target fused data;

[0061] The upper limit and lower limit of the data volume setting range are obtained by the following formula:

[0062] ;

[0063] Among them, C max Indicates the upper limit value corresponding to the data volume setting range; D t represents the total amount of data corresponding to the target fusion data; n represents the total number of transmission channels; S t represents the sum of the remaining capacities corresponding to all transmission channels; a represents a first adjustment factor, and the value range of the first adjustment factor is 0.5-2.0; S i represents the remaining channel capacity corresponding to the i-th transmission channel;

[0064] ;

[0065] Among them, Cmin Indicates the lower limit of the data volume setting range; D t represents the total amount of data corresponding to the target fusion data; n represents the total number of transmission channels; S t represents the sum of the remaining capacities corresponding to all transmission channels; β represents the second adjustment factor, and the value range of the second adjustment factor is 0.6-2.3; S i represents the remaining channel capacity corresponding to the i-th transmission channel; γ represents the third adjustment factor, and the value range of the second adjustment factor is 0.4-2.2;

[0066] According to the data volume setting range of the single data segment, the data volume of the single data segment corresponding to the data segment division is set as the target data volume;

[0067] The target fusion data is divided according to the target data volume to obtain multiple data segments corresponding to the target fusion data.

[0068] Preferably, after the transmission channel for data transmission is determined, determining the number of data segments for a single data transmission according to the remaining channel capacity of the transmission channel currently used for data transmission includes:

[0069] After the transmission channel for data transmission is determined, extracting the current remaining channel capacity of the transmission channel for data transmission; wherein the transmission channel for data transmission is the second transmission channel;

[0070] Extracting the remaining channel capacity corresponding to each unit time of the history of the transmission channel used for data transmission;

[0071] The remaining channel capacity fluctuation coefficient corresponding to the transmission channel for data transmission is obtained according to the current remaining channel capacity of the transmission channel for data transmission combined with the historical remaining channel capacity corresponding to each unit time of the transmission channel for data transmission; wherein the remaining channel capacity fluctuation coefficient is obtained by the following formula:

[0072] ;

[0073] Among them, w S represents the residual channel capacity fluctuation coefficient; k represents the first adjustment coefficient, and the value range of the first adjustment coefficient is 0.8-2.7; g represents the second adjustment coefficient, and the value range of the second adjustment coefficient is 1.5-3.2; S c Indicates the current remaining channel capacity of the transmission channel used for data transmission; S a represents the average value of the remaining channel capacity per unit time of the transmission channel used for data transmission; S brepresents the standard deviation of the remaining channel capacity corresponding to each unit time of the history of the transmission channel for data transmission; ε represents the minimum constant used to prevent the denominator from being 0; U represents the usage intensity corresponding to the maximum value of the usage frequency of the transmission channel for data transmission; P represents the priority parameter corresponding to the transmission channel for data transmission;

[0074] Extract the target data volume corresponding to each data segment of the current target fusion data;

[0075] The number of data segments for a single data transmission is obtained according to the remaining channel capacity fluctuation coefficient combined with the target data volume corresponding to each data segment of the target fused data and the current remaining channel capacity of the transmission channel for data transmission, wherein the number of data segments for a single data transmission is obtained by the following formula:

[0076] ;

[0077] Where N represents the number of data segments transmitted in a single data transmission and is rounded down; S c Indicates the current remaining channel capacity of the transmission channel used for data transmission; w S represents the remaining channel capacity fluctuation coefficient; T c Indicates the target data size of each data segment; M indicates the maximum number of data segments allowed for a single data transmission.

[0078] Preferably, the data transmission processing unit further includes:

[0079] Fusion data model building blocks for:

[0080] The target fusion data is transmitted to the cloud check-in control terminal for model building;

[0081] Among them, the basic data in the target fusion data is modeled using the decision tree method; the video data is modeled using the convolutional neural network method; the driving data is modeled using the recurrent neural network method; and the toll data is modeled using the clustering algorithm.

[0082] After the model is built, the feature data in the target fusion data is used to perform model integration, and an integrated model is obtained after integration;

[0083] Divide the integrated model into data sets and test sets;

[0084] Use the data set to train the integrated model, use the test set to evaluate the integrated model, and adjust the parameters and structure of the integrated model based on the training and evaluation results;

[0085] After the parameter and structure adjustment is completed, the vehicle fusion model data is obtained.

[0086] Preferably, the abnormal data scheduling unit is further used for:

[0087] Carry out abnormal data analysis on vehicle fusion model data;

[0088] The basic data and driving data in the vehicle fusion model data are subjected to parameter judgment, and the parameter judgment is to compare the parameters of the basic data and driving data in the vehicle fusion model data with the parameters of the standard basic data and driving data, and mark the data whose comparison difference is not within the qualified range as the first abnormal data according to the comparison result;

[0089] The video data in the vehicle fusion model data is tracked by using a tracking algorithm for abnormal behavior, and whether abnormal behavior exists is determined according to the tracking result. If abnormal behavior exists, it is marked as the second abnormal data;

[0090] The toll data in the vehicle fusion model data is used to identify abnormal transactions through toll rules and toll logic. If the vehicle's travel time does not match the business hours of the toll station, or the vehicle's toll amount is lower than expected, it is marked as the third abnormal data;

[0091] The first abnormal data, the second abnormal data and the third abnormal data are respectively processed by different abnormal processing departments for abnormal resolution;

[0092] Finally, the first abnormal data, the second abnormal data and the third abnormal data are sent to the corresponding abnormality processing department.

[0093] Preferably, the abnormal feedback operation unit is further used for:

[0094] The staff of the exception handling department checks and judges the abnormal events on the display terminal;

[0095] And according to the abnormal events, remote personnel are dispatched to solve them;

[0096] After the abnormal event is resolved, the result of the abnormal event resolution is displayed and operated on the display terminal.

[0097] Compared with the prior art, the present invention has the following beneficial effects:

[0098] 1. The intelligent cloud check-in management system based on multi-data fusion processing provided by the present invention can integrate different types of data such as basic data, driving data, charging data and video data through a unified data preprocessing process to form a comprehensive and multi-dimensional vehicle data view. The feature layer fusion method effectively reduces the dimension and complexity of the data by extracting feature data from various types of data and fusing them into feature vectors. The fused target fusion data contains richer and more comprehensive vehicle information, enabling the system to analyze and predict vehicle behavior more intelligently.

[0099] 2. The intelligent cloud check-in management system based on multi-data fusion processing provided by the present invention ensures that the data transmission volume is less than the remaining capacity of the transmission channel, which helps to avoid congestion and delays in the data transmission process and ensure the real-time and integrity of the data. When there are multiple qualified transmission channels, the one with the strongest signal strength is further selected as the second transmission channel, which helps to reduce data loss and errors in the transmission process and improve the reliability of data transmission. The most suitable algorithm is used for model construction for different types of data, which can give full play to the advantages of each algorithm and improve the accuracy and adaptability of the model.

[0100] 3. The intelligent cloud check-in management system based on multi-data fusion processing provided by the present invention adopts a variety of intelligent technologies in the abnormal data analysis process, such as tracking algorithms, charging rules and charging logic. The application of these technologies enables the system to automatically identify and process abnormal data without manual intervention or with little manual intervention. This greatly improves the intelligence and automation level of the system, reduces labor costs and work intensity. In the process of abnormal data processing, different abnormal processing departments need to cooperate and coordinate with each other. This promotes communication and information sharing between departments and improves the collaborative combat capability of the entire system. BRIEF DESCRIPTION OF THE DRAWINGS

[0101] Figure 1 This is a schematic diagram of the intelligent cloud check-in management system module of the present invention;

[0102] Figure 2 It is a schematic diagram of the intelligent cloud check-in management process of the present invention. DETAILED DESCRIPTION

[0103] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0104] In order to solve the problem in the prior art that vehicle data is not further processed and integrated after being acquired, resulting in poor quality of vehicle data and the inability to build a model later, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions:

[0105] Intelligent cloud check-in management system based on multi-data fusion processing, including:

[0106] The vehicle data acquisition unit is used to:

[0107] Retrieve vehicle data from the database, classify the data according to the type of vehicle data, uniquely encode and label the classified data, and obtain the target vehicle data after uniquely encoding and labeling;

[0108] Vehicle data processing and fusion unit for:

[0109] The target vehicle data is preprocessed, and after the data preprocessing is completed, data fusion is performed to obtain target fusion data;

[0110] Data transmission processing unit, used for:

[0111] The target fusion data is transmitted to the cloud check-in control terminal, and the transmission channel with the fastest transmission speed is automatically selected during data transmission. After the target fusion data is transmitted to the cloud check-in control terminal, the model is constructed using the deep learning model. After the model is constructed, the vehicle fusion model data is obtained;

[0112] Abnormal data dispatch unit, used for:

[0113] Perform abnormal data analysis on the vehicle fusion model data, determine the abnormal area and abnormal type in the vehicle fusion model data based on the analysis results, select the abnormality handling department based on the abnormality type, and send the abnormal area data to the abnormality handling department for abnormality feedback;

[0114] Abnormal feedback operation unit, used for:

[0115] After the exception handling department has completed handling of the exception, it will describe the handling situation in the cloud check-in background, mark the handled exception as completed, and mark the unhandled exception as unhandled.

[0116] Specifically, the vehicle data acquisition unit comprehensively classifies and integrates the vehicle's basic data, video data, driving data and charging data, which helps the system to more accurately understand and analyze the vehicle status and behavior. Through the unified data preprocessing process of the vehicle data processing and fusion unit, different types of data such as basic data, driving data, charging data and video data can be integrated together to form a comprehensive and multi-dimensional vehicle data view. The data transmission processing unit confirms the remaining capacity and signal strength of the transmission channel, which can detect and avoid potential transmission problems in advance. The abnormal data scheduling unit not only improves the accuracy and reliability of the data, but also provides a strong guarantee for subsequent data analysis and application. Through the abnormal feedback operation unit, the staff can directly view and judge abnormal events on the display terminal without waiting for data aggregation or report generation, thereby greatly shortening the response time.

[0117] The vehicle data acquisition unit is also used for:

[0118] Vehicle data includes basic vehicle data, video data, driving data and charging data;

[0119] Among them, basic data includes vehicle model, license plate number, vehicle identification number and owner information; video data includes videos of vehicles entering toll stations and driving on highways; driving data includes vehicle driving trajectory, driving speed and driving time; toll data includes payment data of vehicles from toll station A to toll station B;

[0120] Classify basic data, video data, driving data and charging data respectively;

[0121] After the classification is completed, the basic data, video data, driving data and charging data of the vehicle are uniquely coded and labeled;

[0122] The vehicle's license plate number and the time it entered the toll station are used as the prefix of the code, and then a hash function is used to generate the suffix of the code, and finally the vehicle's unique code number is obtained;

[0123] The vehicle data with the unique coding number is marked as the target vehicle data.

[0124] Specifically, the basic data, video data, driving data and toll data of the vehicle are comprehensively classified and integrated. This fusion of multi-source data can provide more comprehensive and richer vehicle information, which helps the system to understand and analyze the vehicle status and behavior more accurately. By classifying and uniquely coding the vehicle data, the system can retrieve, locate and process the data of a specific vehicle more quickly. This efficient data management method helps to reduce data processing time and improve system response speed. The introduction of unique coding labels ensures the uniqueness and accuracy of vehicle data. Each vehicle has a hash code based on the license plate number and the time of entering the toll station, which greatly reduces the possibility of data confusion and errors. Uniquely coded vehicle data makes data sharing between different systems or departments easier and more efficient. By sharing vehicle data, different systems can work together to improve the efficiency and quality of traffic management.

[0125] Vehicle data processing and fusion unit, including:

[0126] Vehicle data preprocessing module, used for:

[0127] Perform data preprocessing on target vehicle data;

[0128] The data preprocessing process is to clean the basic data, driving data and charging data, unify the data format and convert the data in turn;

[0129] Denoising the video data, dividing the video data into several segments after the noise removal, and extracting key frames from the several segments;

[0130] Image recognition technology is used to perform feature recognition of license plate numbers, vehicles, and driver faces in key frames, and data compression is performed on video data after feature recognition to obtain pre-processed video data after data compression.

[0131] Vehicle data fusion module for:

[0132] Perform data fusion on the target vehicle data after data preprocessing;

[0133] Among them, the basic data, video data, driving data and charging data in the target vehicle data are firstly associated with each other, and the data association includes timestamp association, position association and identifier association;

[0134] After data association, feature data in basic data, video data, driving data and toll data are obtained;

[0135] The feature data includes the physical attributes and configuration information of the vehicle in the basic data; the feature recognition data in the video data; the parameter data of the vehicle during operation in the driving data; the fee and toll station data in the toll data;

[0136] The feature data is fused into feature vectors through the feature layer fusion method, and the target fused data is obtained after the feature vectors are fused.

[0137] Specifically, preprocessing the data through the vehicle data preprocessing module can improve the data quality. The data cleaning step can remove or correct errors, duplications, missing or outliers in the original data, thereby improving the accuracy of subsequent analysis and processing. The unified data format ensures that data from different sources can be seamlessly integrated, avoiding data integration problems caused by inconsistent formats. Data conversion makes the data easier to be understood and processed by machine learning models, improving the training efficiency and prediction performance of the model. Noise removal can reduce useless information in video data, such as image noise, light changes, etc., and improve the quality of video data. Video data segmentation and key frame extraction reduce the amount of data that needs to be processed, while retaining important information in the video, such as vehicle behavior, driver actions, etc. Using image recognition technology for feature recognition, key information such as license plate numbers, vehicles and driver faces can be automatically extracted, providing strong support for subsequent vehicle tracking, driver identity verification and other tasks. The preprocessing process not only processes a single type of data, but also considers the fusion of multi-source data. Through a unified data preprocessing process, different types of data such as basic data, driving data, toll data and video data can be integrated together to form a comprehensive and multi-dimensional vehicle data view. The basic data, video data, driving data and toll data can be associated through the vehicle data fusion module to obtain all-round information about the target vehicle. This fusion of multi-source data makes the system understand the vehicle more deeply and accurately, which helps to reduce the information loss or error that may be caused by a single data source. The timestamp association and position association used in the data association process ensure that the system can track and record the status and behavior of the vehicle in real time. This real-time performance is particularly important for application scenarios such as traffic management and security monitoring, which can help the system respond to and handle emergencies in a timely manner. The feature layer fusion method effectively reduces the dimension and complexity of the data by extracting feature data from various types of data and fusing them into feature vectors. The fused target fusion data contains richer and more comprehensive vehicle information, enabling the system to analyze and predict vehicle behavior more intelligently. The system can predict the vehicle's future route based on its driving trajectory and speed, or identify potential toll evasion based on toll payment data. Whether it is for application scenarios such as traffic flow control and accident warning for traffic management departments, or for toll management and toll evasion tracking at toll stations, more scientific and reasonable decisions and business operations can be made based on the integrated data.

[0138] In order to solve the problem in the prior art that when transmitting vehicle data, the transmission channel is not selected, resulting in slow data transmission speed and affecting the subsequent event processing, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions:

[0139] The data transmission processing unit comprises:

[0140] Converged data transmission module for:

[0141] Transmit the target fusion data to the cloud check-in control terminal for data analysis and control;

[0142] Before data transmission, the target fused data is divided into data segments, wherein the target fused data is divided into a plurality of data segments of the same length;

[0143] Determine the data transmission volume of the target fusion data according to the number of data segments;

[0144] Confirming the transmission data of the transmission channel, wherein the transmission data includes the remaining capacity of the data transmission channel and the signal strength of the transmission channel;

[0145] Selecting a transmission channel according to the data transmission amount, wherein a transmission channel whose data transmission amount is less than the remaining capacity of the transmission channel is used as a first transmission channel;

[0146] When the number of first transmission channels is greater than one, selecting the transmission channel with the strongest signal strength as the second transmission channel;

[0147] The target fusion data is transmitted to the cloud check-in control terminal through the first transmission channel or the second transmission channel.

[0148] Specifically, the target fusion data is divided into data segments before data transmission, including:

[0149] Before data transmission, extracting the channel remaining capacity corresponding to all transmission channels corresponding to the data transmission;

[0150] According to the channel residual capacity corresponding to all transmission channels corresponding to the data transmission, the transmission channel comprehensive coefficient is obtained; wherein the transmission channel comprehensive coefficient is obtained by the following formula:

[0151] ;

[0152] Where G represents the comprehensive coefficient of the transmission channel; n represents the total number of transmission channels; P i represents the priority parameter corresponding to the i-th transmission channel, and the priority parameter is 1 and 1.5. When the priority is high, the priority parameter is 1.5, and when the priority is low, the priority parameter is 1; C i represents the remaining channel capacity of the i-th transmission channel; C ti represents the total channel capacity of the i-th transmission channel; η i Indicates the percentage value corresponding to the saturation of the i-th transmission channel; R i represents the ratio between the amount of redundant data in the i-th transmission channel and the remaining channel capacity; ri represents the redundant data impact coefficient in the i-th transmission channel, and its value range is 0-1; U i represents the usage intensity corresponding to the i-th transmission channel, that is, the ratio between the usage frequency per unit time and the total usage frequency per unit time of all transmission channels; α i represents the weight coefficient corresponding to the i-th transmission channel;

[0153] Comparing the transmission channel comprehensive coefficient with a preset coefficient threshold;

[0154] When the transmission channel comprehensive coefficient is lower than the preset coefficient threshold, it indicates that the current transmission channel load is too large, and an overload warning is issued;

[0155] When the transmission channel comprehensive coefficient is not lower than a preset coefficient threshold, the target fusion data is divided into data segments.

[0156] The technical effect of the above technical solution is: by calculating the comprehensive coefficient of the transmission channel, the solution can evaluate the current state and transmission capacity of each transmission channel. This helps to make reasonable decisions before data transmission, select or allocate the optimal transmission channel, and thus improve the efficiency of data transmission. The solution takes into account multiple factors such as the remaining capacity, priority, saturation, redundant data volume, usage intensity and weight coefficient of the transmission channel, and can more comprehensively reflect the actual status of the channel. By comparing the comprehensive coefficient of the transmission channel with the preset coefficient threshold, the overloaded channel can be discovered and warned in time, thereby avoiding congestion and delay during data transmission and achieving a balanced distribution of the transmission load. The overload warning mechanism can ensure that potential transmission bottlenecks are identified before data transmission, avoiding data loss or damage due to excessive channel load during transmission, thereby improving the reliability and stability of data transmission. By dynamically evaluating and adjusting the selection of the transmission channel, the solution enables the data transmission system to better adapt to changes in the network environment, such as fluctuations in channel capacity, adjustment of priority, etc., thereby enhancing the system's adaptability and robustness.

[0157] In summary, this technical solution optimizes the data transmission process by comprehensively considering multiple factors, aiming to improve transmission efficiency, achieve load balancing, enhance the reliability of data transmission, and improve the system's adaptability.

[0158] Specifically, when the transmission channel comprehensive coefficient is not lower than a preset coefficient threshold, the target fusion data is divided into data segments, including:

[0159] When the transmission channel comprehensive coefficient is not lower than a preset coefficient threshold, extracting the channel remaining capacity corresponding to all current transmission channels;

[0160] Extract the total data volume corresponding to the target fusion data;

[0161] Obtaining a data volume setting range of a single data segment corresponding to the data segment division by using the channel remaining capacity corresponding to all current transmission channels and the total data volume corresponding to the target fused data;

[0162] The upper limit and lower limit of the data volume setting range are obtained by the following formula:

[0163] ;

[0164] Among them, C max Indicates the upper limit value corresponding to the data volume setting range; D t represents the total amount of data corresponding to the target fusion data; n represents the total number of transmission channels; S t represents the sum of the remaining capacities corresponding to all transmission channels; a represents a first adjustment factor, and the value range of the first adjustment factor is 0.5-2.0; S i represents the remaining channel capacity corresponding to the i-th transmission channel;

[0165] ;

[0166] Among them, C min Indicates the lower limit of the data volume setting range; D t represents the total amount of data corresponding to the target fusion data; n represents the total number of transmission channels; S t represents the sum of the remaining capacities corresponding to all transmission channels; β represents the second adjustment factor, and the value range of the second adjustment factor is 0.6-2.3; S i represents the remaining channel capacity corresponding to the i-th transmission channel; γ represents the third adjustment factor, and the value range of the second adjustment factor is 0.4-2.2;

[0167] According to the data volume setting range of the single data segment, the data volume of the single data segment corresponding to the data segment division is set as the target data volume;

[0168] The target fusion data is divided according to the target data volume to obtain multiple data segments corresponding to the target fusion data.

[0169] The technical effect of the above technical solution is: by dividing the target fused data into data segments, and dynamically calculating the data volume setting range (including upper limit and lower limit) of a single data segment based on the remaining capacity of the transmission channel and the total data volume, the technical solution realizes the refined division of data segments. This division method helps to make more reasonable use of transmission channel resources and avoid low transmission efficiency or waste of resources caused by too large or too small data segments. By introducing the first adjustment factor (a), the second adjustment factor (β) and the third adjustment factor (γ), the technical solution provides additional flexibility to adjust the data volume setting range. These adjustment factors can be adjusted according to actual needs and environmental changes to optimize the efficiency and effect of data transmission. For example, when the transmission conditions are poor, the data volume of a single data segment can be reduced by lowering the upper limit or increasing the lower limit, thereby reducing the risk of transmission failure; when the transmission conditions are good, the data volume of a single data segment can be increased accordingly to improve the transmission efficiency.

[0170] This technical solution takes full account of the complexity of the multi-channel transmission environment. By calculating the sum of the remaining capacity of all transmission channels (St) and combining it with the remaining capacity of a single channel (Si), the data volume setting range of the data segment division is determined. This method helps to achieve load balancing of data transmission in a multi-channel environment, avoiding the situation where some channels are overloaded while other channels are idle, thereby improving the overall transmission performance. This technical solution enhances the robustness of the system by judging whether the transmission channel is overloaded by a preset coefficient threshold and issuing an early warning when it is overloaded. At the same time, by dynamically adjusting the data segment division strategy to adapt to changes in the state of the transmission channel, the system can better respond to emergencies and environmental changes, ensuring the stability and reliability of data transmission. For large-scale target fusion data, this technical solution can support efficient and stable data transmission by reasonably dividing the data segments. The size of each data segment is dynamically adjusted according to the remaining capacity of the transmission channel and the total data volume, which not only ensures transmission efficiency, but also avoids the risk of transmission failure caused by excessively large data segments.

[0171] In summary, this technical solution achieves efficient and stable transmission of target fusion data through technical effects such as refined data segment division, improved transmission efficiency and flexibility, adaptation to multi-channel transmission environment, enhanced system robustness, and support for large-scale data transmission.

[0172] Specifically, after the transmission channel for data transmission is determined, the number of data segments for single data transmission is determined according to the remaining channel capacity of the transmission channel for current data transmission, including:

[0173] After the transmission channel for data transmission is determined, extracting the current remaining channel capacity of the transmission channel for data transmission; wherein the transmission channel for data transmission is the second transmission channel;

[0174] Extracting the remaining channel capacity corresponding to each unit time of the history of the transmission channel used for data transmission;

[0175] The remaining channel capacity fluctuation coefficient corresponding to the transmission channel for data transmission is obtained according to the current remaining channel capacity of the transmission channel for data transmission combined with the historical remaining channel capacity corresponding to each unit time of the transmission channel for data transmission; wherein the remaining channel capacity fluctuation coefficient is obtained by the following formula:

[0176] ;

[0177] Among them, w S represents the residual channel capacity fluctuation coefficient; k represents the first adjustment coefficient, and the value range of the first adjustment coefficient is 0.8-2.7; g represents the second adjustment coefficient, and the value range of the second adjustment coefficient is 1.5-3.2; S c Indicates the current remaining channel capacity of the transmission channel used for data transmission; S a represents the average value of the remaining channel capacity per unit time of the transmission channel used for data transmission; S b represents the standard deviation of the remaining channel capacity corresponding to each unit time of the history of the transmission channel for data transmission; ε represents the minimum constant used to prevent the denominator from being 0; U represents the usage intensity corresponding to the maximum value of the usage frequency of the transmission channel for data transmission; P represents the priority parameter corresponding to the transmission channel for data transmission;

[0178] Extract the target data volume corresponding to each data segment of the current target fusion data;

[0179] The number of data segments for a single data transmission is obtained according to the remaining channel capacity fluctuation coefficient combined with the target data volume corresponding to each data segment of the target fused data and the current remaining channel capacity of the transmission channel for data transmission, wherein the number of data segments for a single data transmission is obtained by the following formula:

[0180] ;

[0181] Where N represents the number of data segments transmitted in a single data transmission and is rounded down; S c Indicates the current remaining channel capacity of the transmission channel used for data transmission; w S represents the remaining channel capacity fluctuation coefficient; T c Indicates the target data size of each data segment; M indicates the maximum number of data segments allowed for a single data transmission.

[0182] The technical effect of the above technical solution is that by extracting the remaining channel capacity of the current transmission channel and its historical remaining channel capacity per unit time, the solution can dynamically evaluate the actual transmission capacity of the channel. This helps to make more accurate decisions before data transmission and avoid transmission failure or delay caused by insufficient channel capacity. S ,This scheme takes the fluctuation of channel capacity into account. This helps to arrange data transmission more reasonably and improve the stability and reliability of transmission when the channel capacity is unstable.

[0183] Based on the remaining channel capacity fluctuation coefficient, the target data volume of each data segment of the target fused data, and the remaining channel capacity of the current transmission channel, the solution can calculate the optimal number of data segments (N) for a single data transmission. This helps to maximize the use of channel resources and improve the efficiency of data transmission while ensuring transmission stability. By introducing the maximum number of data segments (M) allowed for a single data transmission as a constraint, the solution can flexibly adjust the number of data segments according to different transmission requirements and channel conditions to meet different application scenarios and performance requirements.

[0184] The value range of the first adjustment coefficient (k) and the second adjustment coefficient (g) is configurable, which provides users with the flexibility to adjust system performance according to specific application scenarios. By properly configuring these adjustment coefficients, users can better balance the relationship between transmission efficiency, stability and reliability. When calculating the remaining channel capacity fluctuation coefficient, the scheme also considers the usage intensity (F) and priority parameter (P) corresponding to the maximum frequency of use of the channel. This helps to more reasonably allocate transmission resources according to the actual usage and priority of the channel in a multi-channel environment, and improve the overall performance and adaptability of the system. The scheme supports efficient and stable data transmission of large-scale target fusion data by reasonably dividing the data segments and determining the number of data segments transmitted in a single time. This is of great significance for application scenarios that need to process large amounts of data.

[0185] In summary, the above technical solution realizes refined management and optimized control of the data transmission process through technical effects such as dynamically evaluating channel capacity, optimizing data transmission efficiency, enhancing the system's adaptability and supporting large-scale data transmission.

[0186] Fusion data model building blocks for:

[0187] The target fusion data is transmitted to the cloud check-in control terminal for model building;

[0188] Among them, the basic data in the target fusion data is modeled using the decision tree method; the video data is modeled using the convolutional neural network method; the driving data is modeled using the recurrent neural network method; and the toll data is modeled using the clustering algorithm.

[0189] After the model is built, the feature data in the target fusion data is used to perform model integration, and an integrated model is obtained after integration;

[0190] Divide the integrated model into data sets and test sets;

[0191] Use the data set to train the integrated model, use the test set to evaluate the integrated model, and adjust the parameters and structure of the integrated model based on the training and evaluation results;

[0192] After the parameter and structure adjustment is completed, the vehicle fusion model data is obtained.

[0193] Specifically, the target fused data is divided into several data segments of the same length through the fused data transmission module, which is conducive to parallel transmission and reduces waiting time, thereby improving the overall efficiency of data transmission. The data transmission volume is judged according to the number of data segments, which can more accurately evaluate the required transmission resources and avoid waste or shortage of resources. When selecting a transmission channel, first ensure that the data transmission volume is less than the remaining capacity of the transmission channel, which helps to avoid congestion and delays during data transmission and ensure the real-time and integrity of the data. When there are multiple qualified transmission channels, the one with the strongest signal strength is further selected as the second transmission channel, which helps to reduce data loss and errors during transmission and improve the reliability of data transmission. The optimal transmission channel can be dynamically selected according to the data transmission volume and the real-time status of the transmission channel, enhancing the system Adaptability to different environments and conditions. At the same time, steps such as data segment division and transmission channel selection are configurable and can be adjusted and optimized according to actual needs, which improves the flexibility and scalability of the system. By confirming the remaining capacity and signal strength of the transmission channel, potential transmission problems such as channel congestion and signal interference can be discovered and avoided in advance, thereby improving the stability and reliability of the system. Through the fusion data model construction module, the most suitable algorithm is used for model construction for different types of data (such as basic data, video data, driving data, and toll data), which can give full play to the advantages of each algorithm and improve the accuracy and adaptability of the model. Through model integration, the prediction results of multiple single models are merged, which can reduce the risk of overfitting that may exist in a single model and improve the generalization ability of the model. Even if a single model performs poorly in a specific situation, the integrated model can still maintain good overall performance. Splitting the target fusion data into different parts and building models separately helps to process data in parallel and improve data processing speed. At the same time, dividing the integrated model into data sets and test sets can more scientifically evaluate the model performance and quickly adjust the model parameters and structure based on the evaluation results. The intelligent cloud check-in management system may face various complex and changing scenarios, such as traffic congestion, bad weather, emergencies, etc. By building and integrating multiple models, the system can respond to these complex scenarios more flexibly and provide more accurate and timely decision support.

[0194] In order to solve the problem in the prior art that the corresponding exception handling department is not selected according to the abnormal type of the vehicle, which leads to the inability to better solve the abnormal event, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions:

[0195] The abnormal data dispatch unit is also used for:

[0196] Carry out abnormal data analysis on vehicle fusion model data;

[0197] The basic data and driving data in the vehicle fusion model data are subjected to parameter judgment, and the parameter judgment is to compare the parameters of the basic data and driving data in the vehicle fusion model data with the parameters of the standard basic data and driving data, and mark the data whose comparison difference is not within the qualified range as the first abnormal data according to the comparison result;

[0198] The video data in the vehicle fusion model data is tracked by using a tracking algorithm for abnormal behavior, and whether abnormal behavior exists is determined according to the tracking result. If abnormal behavior exists, it is marked as the second abnormal data;

[0199] The toll data in the vehicle fusion model data is used to identify abnormal transactions through toll rules and toll logic. If the vehicle's travel time does not match the business hours of the toll station, or the vehicle's toll amount is lower than expected, it is marked as the third abnormal data;

[0200] The first abnormal data, the second abnormal data and the third abnormal data are respectively processed by different abnormal processing departments for abnormal resolution;

[0201] Finally, the first abnormal data, the second abnormal data and the third abnormal data are sent to the corresponding abnormality processing department.

[0202] Specifically, abnormal data analysis can timely discover and process abnormal data in the system to prevent these abnormal data from interfering with or damaging the normal operation of the system. By marking and classifying the first, second, and third abnormal data, the system can more accurately locate the problem, so as to take targeted solutions to ensure the safety and stability of the system. By comparing with standard basic data and driving data parameters, as well as tracking abnormal behaviors of video data and identifying abnormal transactions of charging data, the system can identify and eliminate those data that do not meet the specifications or have obvious errors. This not only improves the accuracy and reliability of the data, but also provides a strong guarantee for subsequent data analysis and application. The abnormal data analysis process adopts a variety of intelligent technologies, such as tracking algorithms, charging rules, and charging logic. The application of these technologies enables the system to automatically identify and process abnormal data without manual intervention or with little manual intervention. This greatly improves the intelligence and automation level of the system, reduces labor costs and work intensity, and in the process of abnormal data processing, different abnormal processing departments need to cooperate and coordinate with each other. This promotes communication and information sharing between departments and improves the collaborative combat capability of the entire system. At the same time, by sending abnormal data to the corresponding abnormal processing department, the system also realizes the accurate transmission and effective use of data.

[0203] The abnormal feedback operation unit is also used for:

[0204] The staff of the exception handling department checks and judges the abnormal events on the display terminal;

[0205] And according to the abnormal events, remote personnel are dispatched to solve them;

[0206] After the abnormal event is resolved, the result of the abnormal event resolution is displayed and operated on the display terminal.

[0207] Specifically, staff can view and judge abnormal events directly on the display terminal without waiting for data aggregation or report generation, which greatly shortens the response time. This immediacy helps to quickly locate problems and initiate solutions, reducing potential losses caused by delays. The handling of abnormal events by remote personnel dispatch breaks geographical restrictions, allowing experts or technicians to intervene quickly and effectively solve problems even if they are not on site. This remote collaboration mode improves resource utilization efficiency and reduces labor costs. Remote personnel can quickly formulate and implement solutions based on the specific circumstances of abnormal events. At the same time, the processing results can be directly confirmed on the display terminal, which simplifies the processing process and improves processing efficiency. The fast abnormal response and processing capabilities help to improve user satisfaction and trust in the system. Users can feel the efficiency and reliability of the system, so they are more willing to use and promote the system. The processing results and operation records of abnormal events provide valuable data support for the system. By analyzing these data, potential problems and improvement points of the system can be identified, so as to continuously optimize the performance and functions of the system.

[0208] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0209] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. Intelligent cloud check-in management system based on multi-data fusion processing, characterized by: include: The vehicle data acquisition unit is used to: Retrieve vehicle data from the database, classify the data according to the type of vehicle data, uniquely encode and label the classified data, and obtain the target vehicle data after uniquely encoding and labeling; Vehicle data processing and fusion unit for: The target vehicle data is preprocessed, and after the data preprocessing is completed, data fusion is performed to obtain target fusion data; Data transmission processing unit, used for: The target fusion data is transmitted to the cloud check-in control terminal, and the transmission channel with the fastest transmission speed is automatically selected during data transmission. After the target fusion data is transmitted to the cloud check-in control terminal, the model is constructed using the deep learning model. After the model is constructed, the vehicle fusion model data is obtained; Abnormal data dispatch unit, used for: Perform abnormal data analysis on the vehicle fusion model data, determine the abnormal area and abnormal type in the vehicle fusion model data based on the analysis results, select the abnormality handling department based on the abnormality type, and send the abnormal area data to the abnormality handling department for abnormality feedback; Abnormal feedback operation unit, used for: After the exception handling department has completed the exception handling, it will describe the handling situation in the cloud check-in background, mark the handled exception as completed, and mark the unhandled exception as unhandled; The data transmission processing unit comprises: Converged data transmission module for: Transmit the target fusion data to the cloud check-in control terminal for data analysis and control; Before data transmission, the target fused data is divided into data segments, wherein the target fused data is divided into a plurality of data segments of the same length; Determine the data transmission volume of the target fusion data according to the number of data segments; Confirming the transmission data of the transmission channel, wherein the transmission data includes the remaining capacity of the data transmission channel and the signal strength of the transmission channel; Selecting a transmission channel according to the data transmission amount, wherein a transmission channel whose data transmission amount is less than the remaining capacity of the transmission channel is used as a first transmission channel; When the number of first transmission channels is greater than one, selecting the transmission channel with the strongest signal strength as the second transmission channel; The target fusion data is transmitted to the cloud check-in control terminal through the first transmission channel or the second transmission channel; Fusion data model building blocks for: The target fusion data is transmitted to the cloud check-in control terminal for model building; Among them, the basic data in the target fusion data is modeled using the decision tree method; the video data is modeled using the convolutional neural network method; the driving data is modeled using the recurrent neural network method; and the toll data is modeled using the clustering algorithm. After the model is built, the feature data in the target fusion data is used to perform model integration, and the integrated model is obtained after integration; Divide the integrated model into data sets and test sets; Use the data set to train the integrated model, use the test set to evaluate the integrated model, and adjust the parameters and structure of the integrated model based on the training and evaluation results; After the parameters and structure are adjusted, the vehicle fusion model data is obtained; The target fusion data is divided into data segments before data transmission, including: Before data transmission, extracting the channel remaining capacity corresponding to all transmission channels corresponding to the data transmission; According to the channel residual capacity corresponding to all transmission channels corresponding to the data transmission, the transmission channel comprehensive coefficient is obtained; wherein the transmission channel comprehensive coefficient is obtained by the following formula: ； Where G represents the comprehensive coefficient of the transmission channel; n represents the total number of transmission channels; P i represents the priority parameter corresponding to the i-th transmission channel, and the priority parameter is 1 and 1.

5. When the priority is high, the priority parameter is 1.5, and when the priority is low, the priority parameter is 1; C i represents the remaining channel capacity of the i-th transmission channel; C ti represents the total channel capacity of the i-th transmission channel; η i Indicates the percentage value corresponding to the saturation of the i-th transmission channel; R i represents the ratio between the amount of redundant data in the i-th transmission channel and the remaining channel capacity; r i represents the redundant data impact coefficient in the i-th transmission channel, and its value range is 0-1; U i represents the usage intensity corresponding to the i-th transmission channel, that is, the ratio between the usage frequency per unit time and the total usage frequency per unit time of all transmission channels; α i represents the weight coefficient corresponding to the i-th transmission channel; Comparing the transmission channel comprehensive coefficient with a preset coefficient threshold; When the transmission channel comprehensive coefficient is lower than the preset coefficient threshold, it indicates that the current transmission channel load is too large, and an overload warning is issued; When the transmission channel comprehensive coefficient is not lower than a preset coefficient threshold, the target fusion data is divided into data segments; When the transmission channel comprehensive coefficient is not lower than a preset coefficient threshold, the target fusion data is divided into data segments, including: When the transmission channel comprehensive coefficient is not lower than a preset coefficient threshold, extracting the channel remaining capacity corresponding to all current transmission channels; Extract the total data volume corresponding to the target fusion data; Obtaining a data volume setting range of a single data segment corresponding to the data segment division by using the channel remaining capacity corresponding to all current transmission channels and the total data volume corresponding to the target fused data; The upper limit and lower limit of the data volume setting range are obtained by the following formula: ； Among them, C max Indicates the upper limit value corresponding to the data volume setting range; D t represents the total amount of data corresponding to the target fusion data; n represents the total number of transmission channels; S t represents the sum of the remaining capacities corresponding to all transmission channels; a represents the first adjustment factor, and the value range of the first adjustment factor is 0.5-2.0; S i represents the remaining channel capacity corresponding to the i-th transmission channel; ； Among them, C min Indicates the lower limit of the data volume setting range; D t represents the total amount of data corresponding to the target fusion data; n represents the total number of transmission channels; S t represents the sum of the remaining capacities corresponding to all transmission channels; β represents the second adjustment factor, and the value range of the second adjustment factor is 0.6-2.3; S i represents the remaining channel capacity corresponding to the i-th transmission channel; γ represents the third adjustment factor, and the value range of the second adjustment factor is 0.4-2.2; According to the data volume setting range of the single data segment, the data volume of the single data segment corresponding to the data segment division is set as the target data volume; Dividing the target fused data according to the target data volume to obtain multiple data segments corresponding to the target fused data; After the transmission channel for data transmission is determined, the number of data segments for a single data transmission is determined according to the remaining channel capacity of the transmission channel for current data transmission, including: After the transmission channel for data transmission is determined, extracting the current remaining channel capacity of the transmission channel for data transmission; wherein the transmission channel for data transmission is the second transmission channel; Extracting the remaining channel capacity corresponding to each unit time of the history of the transmission channel used for data transmission; The remaining channel capacity fluctuation coefficient corresponding to the transmission channel for data transmission is obtained according to the current remaining channel capacity of the transmission channel for data transmission combined with the historical remaining channel capacity corresponding to each unit time of the transmission channel for data transmission; wherein the remaining channel capacity fluctuation coefficient is obtained by the following formula: ； Among them, w S represents the residual channel capacity fluctuation coefficient; k represents the first adjustment coefficient, and the value range of the first adjustment coefficient is 0.8-2.7; g represents the second adjustment coefficient, and the value range of the second adjustment coefficient is 1.5-3.2; S c Indicates the current remaining channel capacity of the transmission channel used for data transmission; S a represents the average value of the remaining channel capacity per unit time of the transmission channel used for data transmission; S b represents the standard deviation of the remaining channel capacity corresponding to each unit time of the history of the transmission channel for data transmission; ε represents the minimum constant used to prevent the denominator from being 0; U represents the usage intensity corresponding to the maximum value of the usage frequency of the transmission channel for data transmission; P represents the priority parameter corresponding to the transmission channel for data transmission; Extract the target data volume corresponding to each data segment of the current target fusion data; The number of data segments for a single data transmission is obtained according to the remaining channel capacity fluctuation coefficient combined with the target data volume corresponding to each data segment of the target fused data and the current remaining channel capacity of the transmission channel for data transmission, wherein the number of data segments for a single data transmission is obtained by the following formula: ； Where N represents the number of data segments transmitted in a single data transmission and is rounded down; S c Indicates the current remaining channel capacity of the transmission channel used for data transmission; w S represents the remaining channel capacity fluctuation coefficient; T c Indicates the target data size of each data segment; M indicates the maximum number of data segments allowed for a single data transmission.

2. The intelligent cloud check-in management system based on multi-data fusion processing according to claim 1 is characterized in that: The vehicle data acquisition unit is further used for: Vehicle data includes basic vehicle data, video data, driving data and charging data; Among them, basic data includes vehicle model, license plate number, vehicle identification number and owner information; video data includes videos of vehicles entering toll stations and driving on highways; driving data includes vehicle driving trajectory, driving speed and driving time; toll data includes payment data of vehicles from toll station A to toll station B; Classify basic data, video data, driving data and charging data respectively; After the classification is completed, the basic data, video data, driving data and charging data of the vehicle are uniquely coded and labeled; The vehicle's license plate number and the time it entered the toll station are used as the prefix of the code, and then a hash function is used to generate the suffix of the code, and finally the vehicle's unique code number is obtained; The vehicle data with the unique coding number is marked as the target vehicle data.

3. The intelligent cloud check-in management system based on multi-data fusion processing according to claim 2 is characterized in that: The vehicle data processing and fusion unit comprises: Vehicle data preprocessing module, used for: Perform data preprocessing on target vehicle data; The data preprocessing process is to clean the basic data, driving data and charging data, unify the data format and convert the data in turn; Denoising the video data, dividing the video data into several segments after the noise removal, and extracting key frames from the several segments; Image recognition technology is used to perform feature recognition of license plate numbers, vehicles, and driver faces in key frames, and data compression is performed on video data after feature recognition to obtain pre-processed video data after data compression.

4. The intelligent cloud check-in management system based on multi-data fusion processing according to claim 3 is characterized in that: The vehicle data processing and fusion unit further includes: Vehicle data fusion module for: Perform data fusion on the target vehicle data after data preprocessing; Among them, the basic data, video data, driving data and charging data in the target vehicle data are firstly associated with each other, and the data association includes timestamp association, position association and identifier association; After data association, feature data in basic data, video data, driving data and toll data are obtained; The feature data includes the physical attributes and configuration information of the vehicle in the basic data; the feature recognition data in the video data; the parameter data of the vehicle during operation in the driving data; the fee and toll station data in the toll data; The feature data is fused into feature vectors through the feature layer fusion method, and the target fused data is obtained after the feature vectors are fused.

5. The intelligent cloud check-in management system based on multi-data fusion processing according to claim 4 is characterized in that: The abnormal data scheduling unit is further used for: Carry out abnormal data analysis on vehicle fusion model data; The basic data and driving data in the vehicle fusion model data are subjected to parameter judgment, and the parameter judgment is to compare the parameters of the basic data and driving data in the vehicle fusion model data with the parameters of the standard basic data and driving data, and mark the data whose comparison difference is not within the qualified range as the first abnormal data according to the comparison result; The video data in the vehicle fusion model data is tracked by using a tracking algorithm for abnormal behavior, and whether abnormal behavior exists is determined according to the tracking result. If abnormal behavior exists, it is marked as the second abnormal data; The toll data in the vehicle fusion model data is used to identify abnormal transactions through toll rules and toll logic. If the vehicle's travel time does not match the business hours of the toll station, or the vehicle's toll amount is lower than the expected value, it is marked as the third abnormal data; The first abnormal data, the second abnormal data and the third abnormal data are respectively processed by different abnormal processing departments for abnormal resolution; Finally, the first abnormal data, the second abnormal data and the third abnormal data are sent to the corresponding abnormality processing department.

6. The intelligent cloud check-in management system based on multi-data fusion processing according to claim 5 is characterized in that: The abnormal feedback operation unit is further used for: The staff of the exception handling department checks and judges the abnormal events on the display terminal; And according to the abnormal events, remote personnel are dispatched to solve them; After the abnormal event is resolved, the result of the abnormal event resolution is displayed and operated on the display terminal.

Images