A traffic travel information recognition and management system and method based on data processing

Abstract

The present invention discloses a traffic travel information recognition and management system and method based on data processing, which relates to the technical field of data analysis. The present invention collects signaling data, identifies the destination end point, forms a travel recognition system, and summarizes the historical recognition records of the personnel numbers; calculates the average stay duration and the threshold value, and determines the feature record set of the feature numbers; collects the occupational information of the feature numbers, obtains the departure time points of the occupations, and determines the occupational feature time periods; sets the training period, calculates the recognition success rate of the feature time periods, determines the monitoring period, and calculates the risk value; monitors the real-time recognition success rate of the feature time periods, and if it is lower than the threshold value, it reminds to divide the grid, and if it is higher than the threshold value, it adjusts the monitoring period, so as to improve the utilization rate of resources.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and specifically to a traffic travel information recognition and management system and method based on data processing. Background Art

[0002] With the development of intelligent transportation systems, the accurate recognition and management of traffic travel information have become key issues for improving traffic efficiency, optimizing urban planning, and enhancing travel services. Traditional traffic travel data mainly rely on traffic monitoring devices or manual surveys. In recent years, with the development of communication technologies, especially the wide application of signaling data, signaling data has become an important data source for analyzing urban travel characteristics, making the collection of traffic travel data more efficient and real-time, and greatly promoting the in-depth development of traffic data analysis;

[0003] Over time, the accuracy of traffic travel information recognition technology based on signaling data will fluctuate. Then, it is necessary to periodically re-partition the original grid records of signaling data more carefully. However, the prerequisite for grid partitioning is to evaluate the recognition accuracy. When to evaluate the recognition accuracy is of utmost importance. If the evaluation frequency is too frequent, it may consume a large amount of resources, while if the evaluation frequency is too sparse, it may lead to a decrease in recognition accuracy. Therefore, it is necessary to find an appropriate time to evaluate, so as to refine the original grid, improve the recognition accuracy, and effectively improve the utilization rate of resources. Summary of the Invention

[0004] The purpose of the present invention is to provide a traffic travel information recognition and management system and method based on data processing to solve the problems raised in the prior art.

[0005] To solve the above technical problems, the present invention provides the following technical solution: A traffic travel information recognition and management method based on data processing, the method comprising:

[0006] Step S100: By collecting signaling data, identifying the destination end of a person, forming a travel recognition system, collecting the person numbers in the historical recognition records, and summarizing the historical recognition record set of a certain person number;

[0007] Step S200: In the historical recognition record set of a certain person number, calculate the average daily stay duration of the person number, calculate the average stay duration threshold, and determine the characteristic recognition record set of the characteristic number;

[0008] Step S300: Collect the occupations corresponding to the characteristic numbers, obtain the characteristic recognition record set of a certain occupation, collect the departure time points of the characteristic recognition records, determine the time periods corresponding to the characteristic recognition records, and determine the characteristic time periods of a certain occupation;

[0009] Step S400: Set the training period, calculate the recognition success rate of the feature time period, determine the monitoring period of the feature time period, and calculate the risk value of the recognition success rate;

[0010] Step S500: Calculate the real-time recognition success rate of the feature time period within the monitoring period of the feature time period. If the real-time recognition success rate is lower than the recognition success rate threshold, remind the staff to re-divide the grid. If the real-time recognition success rate is higher than the recognition success rate threshold, calculate the dynamically adjusted monitoring period.

[0011] Further, step S100 includes:

[0012] Step S101: Divide the urban area into several grids. By arranging the base stations of communication devices in the grids, collect the signaling data of any person within a certain grid range, number the collected persons, and input the signaling data into the travel recognition model to identify the destination end point of the person, and form a travel recognition system;

[0013] Step S102: In the historical recognition record set, collect the person numbers in a certain historical recognition record, and summarize the historical recognition records with the same person number to obtain the historical recognition record set of a certain person number;

[0014] The signaling data includes the specific timestamp of collecting the signaling data, the staying time at each location, the destination end point, the actual end point, etc.;

[0015] The specific working process of the travel recognition model is to obtain the location information of the person through communication devices such as base stations and Wi-Fi access points, clean, denoise and structure the collected original data, extract the trajectory data of the person, and establish model recognition rules based on historical trajectories and travel characteristics, such as common paths, destination classification, etc. Use machine learning or deep learning algorithms such as neural networks and spatio-temporal analysis models to identify the destination end point of the person under the current conditions.

[0016] Further, step S200 includes:

[0017] Step S201: In the historical recognition record set of a certain person number, obtain the system collection date and the person's staying duration in a certain historical recognition record, summarize the historical recognition records of the same month, and calculate the average daily staying duration of the person according to the following formula:

[0018] ;

[0019] where T represents the average daily staying duration of the person, t a represents the staying duration of the person on the a-th day, and b represents the total number of days;

[0020] Step S202: Aggregate the average daily stay duration of all personnel with numbers, and calculate the threshold of the average stay duration of personnel according to the following formula:

[0021] ;

[0022] where T' represents the threshold of the average stay duration of personnel, and T c represents the average daily stay duration of the personnel with the c-th personnel number, and d represents the total number of personnel numbers;

[0023] Step S203: Compare the average daily stay duration of the personnel with a certain personnel number with the threshold of the average stay duration of personnel. If the average daily stay duration of the personnel exceeds the threshold of the average stay duration of personnel, mark the personnel number as a characteristic number, and set the historical recognition record set of the characteristic number as the characteristic recognition record set;

[0024] By analyzing the collection date of the historical recognition record and the stay duration of the personnel, aggregating the data and calculating the average daily stay duration, it can more accurately reflect the behavioral characteristics of each personnel. By comparing the average daily stay duration of the personnel with the average stay duration threshold of all personnel, it is possible to effectively screen out those who live or work in the grid for a long time.

[0025] Further, step S300 includes:

[0026] Step S301: Collect the occupations corresponding to a certain characteristic number, aggregate the characteristic numbers of the same occupation to obtain a characteristic number set of a certain occupation, and aggregate the characteristic recognition record sets of all characteristic numbers in the characteristic number set to obtain a characteristic recognition record set of a certain occupation;

[0027] Step S302: Divide the time of a day into several time periods, obtain the time range corresponding to any time period. In the characteristic recognition record set of a certain occupation, collect the departure time point in a certain characteristic recognition record to obtain the time period corresponding to the characteristic recognition record, aggregate the time periods corresponding to all characteristic recognition records, collect the number of characteristic recognition records in a certain time period, and calculate the occurrence frequency of the characteristic recognition records in the time period according to the following formula:

[0028] ;

[0029] where A represents the occurrence frequency of the characteristic recognition records in the time period, B represents the number of the characteristic recognition records in the time period, and C represents the total number of the characteristic recognition records;

[0030] Step S303: Set a frequency threshold, set the time periods during which the occurrence frequency of the feature recognition records exceeds the frequency threshold as feature time periods, and summarize them to obtain a set of feature time periods for a certain occupation;

[0031] Through the summary and analysis of the feature numbers and feature recognition records of the same occupation, it is possible to effectively identify and classify occupation-specific behavior patterns, and the behavior characteristics of personnel in different occupations will be different;

[0032] Divide the time of a day into multiple time periods, and calculate the occurrence frequency of the feature recognition records in different time periods, so that the behavior characteristics of personnel can be analyzed more precisely. For example, the activity frequency of a certain occupation in a specific time period may be regular, and identifying these "high-frequency" time periods helps to understand the work patterns, behavior concentration periods, etc. of the personnel.

[0033] Furthermore, step S400 includes:

[0034] Step S401: Select several consecutive days as the training period. In the set of feature time periods of a certain occupation, obtain the set of recognition records in a certain feature time period on a certain day, collect the speculated destination end point and the actual destination end point of a certain recognition record. If the speculated destination end point is the same as the actual destination end point, then the recognition record result is successful; if the speculated destination end point is different from the actual destination end point, then the recognition record result is failed;

[0035] Step S402: Respectively collect the numbers of successful and failed recognition record results in a certain feature time period on a certain day, and calculate the recognition success rate in a certain feature time period on a certain day according to the following formula:

[0036] ;

[0037] Among them, X represents the recognition success rate in a certain feature time period on a certain day, Y represents the number of successful record results in a certain feature time period on a certain day, and Z represents the number of failed record results in a certain feature time period on a certain day;

[0038] Step S403: Summarize the recognition success rates of a certain feature time period within the training period in chronological order, and draw a recognition success rate curve graph of a certain feature time period with the date as the X-axis and the recognition success rate as the Y-axis;

[0039] Step S404: Set a recognition success rate threshold, mark the recognition success rate threshold on the recognition success rate curve graph of the certain feature time period, set an abnormal number of days threshold, collect the number of consecutive days below the recognition success rate threshold. If the number of days reaches the abnormal number of days threshold, then collect the date of the last day in the number of days, and set it as the monitoring period of a certain feature time period;

[0040] Step S405: Collect the dates within the monitoring period that are lower than the recognition success rate threshold and set them as abnormal dates, obtain the recognition success rate of the abnormal dates and set it as the abnormal recognition success rate, summarize the abnormal dates, calculate the interval duration between each abnormal date, and calculate the risk value of the recognition success rate according to the following formula:

[0041] ;

[0042] where H represents the risk value of the recognition success rate, N represents the number of days lower than the recognition success rate threshold, E e represents the interval duration of the e-th segment, f represents the total number of segments of the interval duration, V represents the recognition success rate threshold, S k represents the k-th abnormal recognition success rate, h represents the total number of abnormal recognition success rates, and U represents the weight value of the risk value;

[0043] By setting the recognition success rate threshold and the abnormal number of days threshold, automatically identify abnormal situations that continuously fall below the success rate standard, determine the monitoring period. This function realizes early warning of long-term performance decline, provides a basis for timely adjustment of strategies, and thus improves system stability;

[0044] Through the analysis of the monitoring period and the risk value, it is possible to accurately locate the problem time period and its risk level, which helps to reasonably allocate resources, optimize the scheduling plan, and improve management efficiency.

[0045] Further, step S500 includes:

[0046] Step S501: During the characteristic time period of a certain occupation, respectively collect the numbers of successful and failed real-time results of the recognition records, calculate the real-time recognition success rate of the characteristic time period within the monitoring period of the characteristic time period. If the real-time recognition success rate is lower than the recognition success rate threshold, remind the staff to re-divide the grid. If the real-time recognition success rate is higher than the recognition success rate threshold, then execute step S502;

[0047] Step S502: If the real-time recognition success rate is lower than the risk value of the recognition success rate, increase the risk count by one, and calculate the dynamically adjusted monitoring period according to the following formula:

[0048] ;

[0049] where Q’ represents the dynamically adjusted monitoring period, Q represents the monitoring period of the characteristic time period, W represents the risk count, j represents the real-time recognition success rate of the characteristic time period, and calculate the recognition success rate of the characteristic time period according to the dynamically adjusted monitoring period;

[0050] By combining the number of risk occurrences and the real-time success rate, the monitoring period is redefined through a dynamic adjustment formula, enabling the system to adapt to fluctuations in the success rate and changes in risks. This dynamic adjustment mechanism avoids resource waste or unnecessary delays that may be caused by a fixed monitoring period, improving the accuracy and flexibility of monitoring.

[0051] By dynamically adjusting the monitoring period, the system can automatically allocate resources according to the risk level, avoiding over-monitoring during low-risk time periods, and thus making more efficient use of human and equipment resources.

[0052] To better implement the above method, a traffic travel information recognition and management system based on data processing is also proposed. The system includes a travel recognition system module, a feature number module, a feature time period module, a training module, and a dynamic adjustment module.

[0053] Travel recognition system module: By collecting signaling data, identify the destination end of the person, form a travel recognition system, collect the person numbers in the historical recognition records, and summarize the historical recognition record set of a certain person number.

[0054] Feature number module: In the historical recognition record set of a certain person number, calculate the average daily stay duration of the person number, calculate the average stay duration threshold, and determine the feature recognition record set of the feature number.

[0055] Feature time period module: Collect the occupations corresponding to the feature numbers, obtain the feature recognition record set of a certain occupation, collect the departure time points of the feature recognition records, determine the time period corresponding to the feature recognition records, and determine the feature time period of a certain occupation.

[0056] Training module: Set the training period, calculate the recognition success rate of the feature time period, determine the monitoring period of the feature time period, and calculate the risk value of the recognition success rate.

[0057] Dynamic adjustment module: Calculate the real-time recognition success rate of the feature time period within the monitoring period of the feature time period. If the real-time recognition success rate is lower than the recognition success rate threshold, remind the staff to re-divide the grid. If the real-time recognition success rate is higher than the recognition success rate threshold, calculate the dynamically adjusted monitoring period.

[0058] Furthermore, the feature number module includes a unit for calculating the average stay duration threshold of the person and a unit for determining the feature number:

[0059] Calculation unit for the average residence duration threshold of personnel: In the historical identification record set of a certain personnel number, obtain the system collection date and the personnel residence duration in a certain historical identification record, summarize the historical identification records of the same month, calculate the average daily residence duration of the personnel, summarize the average daily residence duration of all personnel numbers, and calculate the average residence duration threshold of the personnel;

[0060] Feature number determination unit: Compare the average daily residence duration of a certain personnel number with the average residence duration threshold of the personnel. If the average daily residence duration of the personnel exceeds the average residence duration threshold of the personnel, mark the personnel number as the feature number.

[0061] Furthermore, the training module includes a monitoring cycle calculation unit and a recognition success rate risk value calculation unit:

[0062] Monitoring cycle calculation unit: Select a continuous number of days as the training cycle. In the set of characteristic time periods of a certain occupation, obtain the set of identification records in a certain characteristic time period on a certain day, collect the predicted destination end point and the actual destination end point of a certain identification record, respectively collect the numbers of successful and failed identification records in a certain characteristic time period on a certain day, calculate the recognition success rate in a certain characteristic time period on a certain day, summarize the recognition success rates of a certain characteristic time period within the training cycle in chronological order, and draw a recognition success rate curve graph of a certain characteristic time period with the date as the X-axis and the recognition success rate as the Y-axis; Set the recognition success rate threshold, mark the recognition success rate threshold on the recognition success rate curve graph of the certain characteristic time period, set the abnormal number of days threshold, collect the number of consecutive days below the recognition success rate threshold, and if the number of days reaches the abnormal number of days threshold, collect the date of the last day in the number of days and set it as the monitoring cycle of a certain characteristic time period;

[0063] Recognition success rate risk value calculation unit: Collect the dates below the recognition success rate threshold within the monitoring cycle and set them as abnormal dates, obtain the recognition success rates of the abnormal dates and set them as abnormal recognition success rates, summarize the abnormal dates, calculate the interval duration between each abnormal date, and calculate the risk value of the recognition success rate.

[0064] Furthermore, the dynamic adjustment module includes a real-time evaluation unit and a monitoring cycle calculation unit after dynamic adjustment:

[0065] Real-time evaluation unit: In the characteristic time period of a certain occupation, respectively collect the numbers of successful and failed real-time results of the identification records, calculate the real-time recognition success rate of the characteristic time period within the monitoring cycle of the characteristic time period. If the real-time recognition success rate is lower than the recognition success rate threshold, remind the staff to re-divide the grid;

[0066] Calculate the dynamically adjusted monitoring period unit: If the real-time recognition success rate is lower than the risk value of the recognition success rate, increase the risk count by one, calculate the dynamically adjusted monitoring period, and calculate the recognition success rate of the characteristic time period according to the dynamically adjusted monitoring period.

[0067] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0068] 1. By setting the real-time recognition success rate threshold and the risk value, and combining the mechanism of dynamically adjusting the monitoring period, it is possible to remind and adjust the grid division when abnormal travel behaviors occur;

[0069] 2. By calculating the stay duration threshold and the real-time recognition success rate, this method can accurately identify the characteristic grids and high-frequency behavior time periods, optimize resource allocation, avoid over-monitoring of low-demand areas or time periods, and improve the utilization efficiency of system resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 It is a schematic flowchart of a traffic travel information recognition and management method based on data processing according to the present invention;

[0071] Figure 2 It is a schematic structural diagram of a traffic travel information recognition and management system based on data processing according to the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0072] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0073] Please refer to Figure 1 and Figure 2 The present invention provides a technical solution: A traffic travel information recognition and management method based on data processing, the method includes:

[0074] Step S100: By collecting signaling data, identify the destination end of the person, form a travel recognition system, collect the person numbers in the historical recognition records, and summarize the historical recognition record set of a certain person number;

[0075] Among them, step S100 includes:

[0076] Step S101: Divide the urban area into several grids. By deploying base stations of communication devices in the grids, collect the signaling data of any person within a certain grid range, number the persons being collected, input the signaling data into the travel recognition model to identify the destination end point of the person, and form a travel recognition system.

[0077] Step S102: In the historical recognition record set, collect the person numbers in a certain historical recognition record, and summarize the historical recognition records with the same person number to obtain the historical recognition record set of a certain person number.

[0078] Step S200: In the historical recognition record set of a certain person number, calculate the average daily stay duration of the person number, calculate the average stay duration threshold, and determine the feature recognition record set of the feature number.

[0079] Among them, Step S200 includes:

[0080] Step S201: In the historical recognition record set of a certain person number, obtain the system collection date and the person's stay duration in a certain historical recognition record, summarize the historical recognition records of the same month, and calculate the average daily stay duration of the person according to the following formula:

[0081] ;

[0082] Among them, T represents the average daily stay duration of the person, t a represents the stay duration of the person on the a-th day, and b represents the total number of days;

[0083] Step S202: Summarize the average daily stay durations of all person numbers, and calculate the average stay duration threshold according to the following formula:

[0084] ;

[0085] Among them, T' represents the average stay duration threshold, T c represents the average daily stay duration of the c-th person number, and d represents the total number of person numbers;

[0086] Step S203: Compare the average daily stay duration of a certain person number with the average stay duration threshold. If the average daily stay duration of the person exceeds the average stay duration threshold, mark the person number as the feature number, and set the historical recognition record set of the feature number as the feature recognition record set.

[0087] Step S300: Collect the occupations corresponding to the feature numbers to obtain a set of feature recognition records for a certain occupation. Collect the starting time points of the feature recognition records, determine the time periods corresponding to the feature recognition records, and determine the feature time periods for a certain occupation.

[0088] Among them, step S300 includes:

[0089] Step S301: Collect the occupations corresponding to a certain feature number, summarize the feature numbers of the same occupation to obtain a set of feature numbers for a certain occupation, and summarize the sets of feature recognition records of all feature numbers in the set of feature numbers to obtain a set of feature recognition records for a certain occupation.

[0090] Step S302: Divide the time of a day into several time periods, obtain the time range corresponding to any time period. In the set of feature recognition records for a certain occupation, collect the starting time point in a certain feature recognition record to obtain the time period corresponding to the feature recognition record. Summarize the time periods corresponding to all feature recognition records, collect the number of feature recognition records in a certain time period, and calculate the occurrence frequency of feature recognition records in the time period according to the following formula:

[0091] ;

[0092] Among them, A represents the occurrence frequency of feature recognition records in the time period, B represents the number of feature recognition records in the time period, and C represents the total number of feature recognition records.

[0093] Step S303: Set a frequency threshold, set the time periods with the occurrence frequency of feature recognition records exceeding the frequency threshold as feature time periods, and summarize them to obtain a set of feature time periods for a certain occupation.

[0094] For example, there are 5 feature recognition records in the first occupation. The starting time point of the first feature recognition record is 7:45, the starting time point of the second feature recognition record is 7:35, the starting time point of the third feature recognition record is 12:15, the starting time point of the fourth feature recognition record is 8:00, and the starting time point of the fifth feature recognition record is 11:00. Divide a day into the following 4 time periods: the morning time period is 06:00 - 10:00, the noon time period is 10:00 - 14:00, the afternoon time period is 14:00 - 18:00, and the evening time period is 18:00 - 22:00.

[0095] It is calculated that the morning frequency is 0.6, the noon frequency is 0.4, and the set frequency threshold is 0.5. Then the feature time period of the first occupation is the morning.

[0096] Step S400: Set the training period, calculate the recognition success rate for the characteristic time period, determine the monitoring period for the characteristic time period, and calculate the risk value of the recognition success rate;

[0097] Among them, step S400 includes:

[0098] Step S401: Select several consecutive days as the training period. In the set of characteristic time periods of a certain occupation, obtain the set of recognition records in a certain characteristic time period on a certain day, collect the predicted destination end point and the actual destination end point of a certain recognition record. If the predicted destination end point is the same as the actual destination end point, then the result of the recognition record is successful; if the predicted destination end point is different from the actual destination end point, then the result of the recognition record is failed;

[0099] Step S402: Respectively collect the number of successful and failed recognition records in a certain characteristic time period on a certain day, and calculate the recognition success rate in a certain characteristic time period on a certain day according to the following formula:

[0100] ;

[0101] Among them, X represents the recognition success rate in a certain characteristic time period on a certain day, Y represents the number of records with successful results in a certain characteristic time period on a certain day, and Z represents the number of records with failed results in a certain characteristic time period on a certain day;

[0102] Step S403: Summarize the recognition success rates of a certain characteristic time period within the training period in chronological order, and draw a recognition success rate curve graph of a certain characteristic time period with the date as the X-axis and the recognition success rate as the Y-axis;

[0103] Step S404: Set the recognition success rate threshold, mark the recognition success rate threshold on the recognition success rate curve graph of the certain characteristic time period, set the abnormal days threshold, collect the number of consecutive days below the recognition success rate threshold. If the number of days reaches the abnormal days threshold, then collect the date of the last day among these days, and set it as the monitoring period of a certain characteristic time period;

[0104] Step S405: Collect the dates below the recognition success rate threshold during the monitoring period and set them as abnormal dates, obtain the recognition success rate of the abnormal dates and set it as the abnormal recognition success rate, summarize the abnormal dates, calculate the interval duration between each abnormal date, and calculate the risk value of the recognition success rate according to the following formula:

[0105] ;

[0106] Among them, H represents the risk value of the recognition success rate, N represents the number of days below the recognition success rate threshold, E eThe interval duration represented as the e-th segment, f represents the total number of segments of the interval duration, V represents the recognition success rate threshold, and S k The k-th abnormal recognition success rate is represented as k , h represents the total number of abnormal recognition success rates, and U represents the weight value of the risk value;

[0107] For example, set the recognition success rate threshold V to 0.6, the abnormal day threshold to 2 days, the success rate on 2024-11-01 is 0.50, the success rate on 2024-11-02 is 0.67, the success rate on 2024-11-03 is 0.55, the success rate on 2024-11-04 is 0.45, 2024-11-01 and 2024-11-03 are abnormal dates, the interval from 2024-11-01 to 2024-11-03 is 2 days, and the calculated risk value is 0.3.

[0108] Step S500: Calculate the real-time recognition success rate of the feature time period within the monitoring cycle of the feature time period. If the real-time recognition success rate is lower than the recognition success rate threshold, remind the staff to re-divide the grid. If the real-time recognition success rate is higher than the recognition success rate threshold, calculate the dynamically adjusted monitoring cycle;

[0109] Among them, step S500 includes:

[0110] Step S501: Within the feature time period of a certain occupation, respectively collect the numbers of successful and failed real-time results of the recognition records, and calculate the real-time recognition success rate of the feature time period within the monitoring cycle of the feature time period. If the real-time recognition success rate is lower than the recognition success rate threshold, remind the staff to re-divide the grid. If the real-time recognition success rate is higher than the recognition success rate threshold, execute step S502;

[0111] Step S502: If the real-time recognition success rate is lower than the risk value of the recognition success rate, increase the risk count by one, and calculate the dynamically adjusted monitoring cycle according to the following formula:

[0112] ;

[0113] Among them, Q’ represents the dynamically adjusted monitoring cycle, Q represents the monitoring cycle of the feature time period, W represents the risk count, j represents the real-time recognition success rate of the feature time period, and calculate the recognition success rate of the feature time period according to the dynamically adjusted monitoring cycle.

[0114] To better implement the above method, a traffic travel information recognition management system based on data processing is also proposed. The system includes a travel recognition system module, a feature numbering module, a feature time period module, a training module, and a dynamic adjustment module;

[0115] Travel recognition system module: By collecting signaling data, identify the destination end point of a person, form a travel recognition system, collect the person numbers in the historical recognition records, and summarize the historical recognition record set of a certain person number;

[0116] Feature number module: In the historical recognition record set of a certain person number, calculate the average daily stay duration of the person number, calculate the average stay duration threshold, and determine the feature recognition record set of the feature number;

[0117] Among them, the feature number module includes an average stay duration threshold calculation unit for personnel and a feature number determination unit:

[0118] Average stay duration threshold calculation unit for personnel: In the historical recognition record set of a certain person number, obtain the system collection date and the person's stay duration in a certain historical recognition record, summarize the historical recognition records of the same month, calculate the average daily stay duration of the person, summarize the average daily stay durations of all person numbers, and calculate the average stay duration threshold for personnel;

[0119] Feature number determination unit: Compare the average daily stay duration of a certain person number with the average stay duration threshold for personnel. If the average daily stay duration of the person exceeds the average stay duration threshold for personnel, mark the person number as a feature number.

[0120] Feature time period module: Collect the occupations corresponding to the feature numbers, obtain the feature recognition record set of a certain occupation, collect the departure time points of the feature recognition records, determine the time period corresponding to the feature recognition records, and determine the feature time period of a certain occupation;

[0121] Training module: Set the training cycle, calculate the recognition success rate of the feature time period, determine the monitoring cycle of the feature time period, and calculate the risk value of the recognition success rate;

[0122] Among them, the training module includes a monitoring cycle calculation unit and a recognition success rate risk value calculation unit:

[0123] Calculation and monitoring period unit: Select several consecutive days as the training period. In the set of characteristic time periods of a certain occupation, obtain the set of recognition records in a certain characteristic time period on a certain day, collect the speculated destination end point and the actual destination end point of a certain recognition record, respectively collect the numbers of successful and failed recognition records in a certain characteristic time period on a certain day, calculate the recognition success rate in a certain characteristic time period on a certain day, summarize the recognition success rates of a certain characteristic time period within the training period in chronological order, and use the date as the X-axis and the recognition success rate as the Y-axis to draw the recognition success rate curve graph of a certain characteristic time period; Set the recognition success rate threshold, and mark the recognition success rate threshold on the recognition success rate curve graph of the certain characteristic time period, set the abnormal days threshold, collect the number of consecutive days below the recognition success rate threshold, and if the number of days reaches the abnormal days threshold, collect the date of the last day in the number of days, and set it as the monitoring period of a certain characteristic time period.

[0124] Calculation of recognition success rate risk value unit: Collect the dates below the recognition success rate threshold within the monitoring period and set them as abnormal dates, obtain the recognition success rates of the abnormal dates and set them as abnormal recognition success rates, summarize the abnormal dates, calculate the interval duration between each abnormal date, and calculate the risk value of the recognition success rate.

[0125] Dynamic adjustment module: Calculate the real-time recognition success rate of the characteristic time period within the monitoring period of the characteristic time period. If the real-time recognition success rate is lower than the recognition success rate threshold, remind the staff to re-divide the grid. If the real-time recognition success rate is higher than the recognition success rate threshold, calculate the dynamically adjusted monitoring period.

[0126] Among them, the dynamic adjustment module includes a real-time evaluation unit and a unit for calculating the dynamically adjusted monitoring period:

[0127] Real-time evaluation unit: In the characteristic time period of a certain occupation, respectively collect the numbers of successful and failed real-time results of the recognition records, calculate the real-time recognition success rate of the characteristic time period within the monitoring period of the characteristic time period. If the real-time recognition success rate is lower than the recognition success rate threshold, remind the staff to re-divide the grid;

[0128] Unit for calculating the dynamically adjusted monitoring period: If the real-time recognition success rate is lower than the risk value of the recognition success rate, increase the risk count by one, calculate the dynamically adjusted monitoring period, and calculate the recognition success rate of the characteristic time period according to the dynamically adjusted monitoring period.

[0129] It is obvious to those skilled in the art that the present invention is not limited to the details of the above-described exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, in any regard, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the appended claims rather than the above description. Accordingly, all changes that fall within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present invention. Any reference signs in the claims should not be construed as limiting the claims involved.

Claims

1. A traffic travel information recognition and management method based on data processing, characterized in that, The method includes: Step S100: By collecting signaling data, identify the destination end point of a person, establish a travel identification system, collect the person numbers in the historical identification records, and summarize the historical identification record set of a certain person number; Step S200: In the historical identification record set of a certain person number, calculate the average daily stay duration of the person number, calculate the average stay duration threshold, and determine the characteristic identification record set of the characteristic number; Step S300: Collect the occupation corresponding to the characteristic number to obtain the characteristic identification record set of a certain occupation, collect the departure time point of the characteristic identification record, determine the time period corresponding to the characteristic identification record, and determine the characteristic time period of a certain occupation; Step S400: Set the training period, calculate the recognition success rate of the characteristic time period, determine the monitoring period of the characteristic time period, and calculate the risk value of the recognition success rate; Step S400 includes the following steps: Step S401: Select a continuous number of days as the training period. In the characteristic time period set of a certain occupation, obtain the recognition record set in a certain characteristic time period on a certain day, collect the speculated destination end point and the actual destination end point of a certain recognition record. If the speculated destination end point is the same as the actual destination end point, the recognition record result is successful; if the speculated destination end point is different from the actual destination end point, the recognition record result is failed; Step S402: Respectively collect the numbers of successful and failed recognition record results in a certain characteristic time period on a certain day, and calculate the recognition success rate in a certain characteristic time period on a certain day according to the following formula: Where, X represents the recognition success rate in a certain characteristic time period on a certain day, Y represents the number of record results that are successful in a certain characteristic time period on a certain day, and Z represents the number of record results that are failed in a certain characteristic time period on a certain day; Step S403: Summarize the recognition success rates of a certain characteristic time period within the training period in chronological order, and draw a recognition success rate curve graph of a certain characteristic time period with the date as the X-axis and the recognition success rate as the Y-axis; Step S404: Set the recognition success rate threshold, mark the recognition success rate threshold on the recognition success rate curve graph of the certain characteristic time period, set the abnormal day threshold, collect the number of consecutive days below the recognition success rate threshold. If the number of days reaches the abnormal day threshold, collect the date of the last day in the number of days and set it as the monitoring period of a certain characteristic time period; Step S405: Collect the dates below the recognition success rate threshold during the monitoring period and set them as abnormal dates, obtain the recognition success rate of the abnormal dates and set it as the abnormal recognition success rate, summarize the abnormal dates, calculate the interval duration between each abnormal date, and calculate the risk value of the recognition success rate according to the following formula: Among them, H represents the risk value of the recognition success rate, N represents the number of days below the recognition success rate threshold, and E e represents the interval duration of the e-th segment, f represents the total number of segments of the interval duration, V represents the recognition success rate threshold, and S k represents the k-th abnormal recognition success rate, h represents the total number of abnormal recognition success rates, and U represents the weight value of the risk value; Step S500: Calculate the real-time recognition success rate of the characteristic time period during the monitoring period of the characteristic time period. If the real-time recognition success rate is lower than the recognition success rate threshold, remind the staff to re-divide the grid. If the real-time recognition success rate is higher than the recognition success rate threshold, calculate the dynamically adjusted monitoring period.

2. The traffic travel information recognition and management method based on data processing according to claim 1, wherein, The step S100 includes the following steps: Step S101: Divide the urban area into several grids. By arranging the base stations of communication devices in the grids, collect the signaling data of any person within the range of a certain grid, number the collected persons, input the signaling data into the travel recognition model to identify the destination end of the person, and form a travel recognition system; Step S102: In the set of historical recognition records, collect the person numbers in a certain historical recognition record, and summarize the historical recognition records with the same person number to obtain a set of historical recognition records for a certain person number.

3. The traffic travel information recognition and management method based on data processing according to claim 2, wherein, The step S200 includes the following steps: Step S201: In the set of historical recognition records for a certain person number, obtain the system collection date and the person's stay duration in a certain historical recognition record, summarize the historical recognition records of the same month, and calculate the average daily stay duration of the person according to the following formula: Among them, T represents the average daily stay duration of personnel, and t a represents the stay duration of personnel on the a-th day, and b represents the total number of days; Step S202: Summarize the average daily stay durations of all person numbers, and calculate the average stay duration threshold of the person according to the following formula: Among them, T’ represents the average residence time threshold for personnel, and T c represents the average daily residence time of the personnel with the c-th personnel number, and d represents the total number of personnel numbers; Step S203: Compare the average daily stay duration of a certain person number with the average stay duration threshold of the person. If the average daily stay duration of the person exceeds the average stay duration threshold of the person, mark the person number as a feature number, and set the set of historical recognition records of the feature number as the set of feature recognition records.

4. A traffic travel information recognition and management method based on data processing according to claim 3, characterized in that The step S300 includes the following steps: Step S301: Collect the occupations corresponding to a certain feature number, summarize the feature numbers of the same occupation to obtain a set of feature numbers of a certain occupation, and summarize the sets of feature recognition records of all feature numbers in the set of feature numbers to obtain a set of feature recognition records of a certain occupation; Step S302: Divide the time of a day into several time periods, obtain the time range corresponding to any time period, in the set of feature recognition records of a certain occupation, collect the departure time point in a certain feature recognition record to obtain the time period corresponding to the feature recognition record, summarize the time periods corresponding to all feature recognition records, collect the number of feature recognition records in a certain time period, and calculate the occurrence frequency of feature recognition records in the time period according to the following formula: Where, A represents the occurrence frequency of feature recognition records in the time period, B represents the number of feature recognition records in the time period, and C represents the total number of feature recognition records; Step S303: Set a frequency threshold, and set the time period with the occurrence frequency of feature recognition records exceeding the frequency threshold as the feature time period, and summarize to obtain a set of feature time periods of a certain occupation.

5. A traffic travel information recognition and management method based on data processing according to claim 4, characterized in that, The step S500 includes the following steps: Step S501: During the characteristic time period of a certain occupation, respectively collect the numbers of real-time results of success and failure of the recognition records, calculate the real-time recognition success rate during the monitoring period of the characteristic time period. If the real-time recognition success rate is lower than the recognition success rate threshold, remind the staff to re-divide the grid. If the real-time recognition success rate is higher than the recognition success rate threshold, execute Step S502; Step S502: If the real-time recognition success rate is lower than the risk value of the recognition success rate, increase the risk count by one, and calculate the dynamically adjusted monitoring period according to the following formula: where Q’ represents the dynamically adjusted monitoring period, Q represents the monitoring period of the characteristic time period, W represents the risk count, and j represents the real-time recognition success rate of the characteristic time period. Calculate the recognition success rate of the characteristic time period according to the dynamically adjusted monitoring period.

6. A traffic travel information recognition and management system based on data processing, which is used to implement the traffic travel information recognition and management method according to any one of claims 1-5, and is characterized in that, The system includes a travel recognition system module, a characteristic number module, a characteristic time period module, a training module, and a dynamic adjustment module; The travel recognition system module: By collecting signaling data, identify the destination end of the person, form a travel recognition system, collect the person numbers in the historical recognition records, and summarize the historical recognition record set of a certain person number; The characteristic number module: In the historical recognition record set of a certain person number, calculate the average daily stay duration of the person number, calculate the average stay duration threshold, and determine the characteristic recognition record set of the characteristic number; The characteristic time period module: Collect the occupation corresponding to the characteristic number, obtain the characteristic recognition record set of a certain occupation, collect the departure time points of the characteristic recognition records, determine the time period corresponding to the characteristic recognition records, and determine the characteristic time period of a certain occupation; The training module: Set the training period, calculate the recognition success rate of the characteristic time period, determine the monitoring period of the characteristic time period, and calculate the risk value of the recognition success rate; The dynamic adjustment module: Calculate the real-time recognition success rate of the characteristic time period during the monitoring period of the characteristic time period. If the real-time recognition success rate is lower than the recognition success rate threshold, remind the staff to re-divide the grid. If the real-time recognition success rate is higher than the recognition success rate threshold, calculate the dynamically adjusted monitoring period.

7. The traffic travel information recognition and management system based on data processing according to claim 6, characterized in that, The characteristic number module includes an average stay duration calculation unit for personnel and a characteristic number determination unit: The average stay duration calculation unit for personnel: In the historical recognition record set of a certain person number, obtain the system collection date and the person's stay duration in a certain historical recognition record, summarize the historical recognition records of the same month, calculate the average daily stay duration of the person, summarize the average daily stay durations of all person numbers, and calculate the average stay duration threshold for personnel; The characteristic number determination unit: Compare the average daily stay duration of a certain person number with the average stay duration threshold for personnel. If the average daily stay duration of the person exceeds the average stay duration threshold for personnel, mark the person number as a characteristic number.

8. A traffic travel information recognition and management system based on data processing according to claim 6, characterized in that, The training module includes a calculation monitoring period unit and a calculation recognition success rate risk value unit: The calculation monitoring period unit: Selects several consecutive days as the training period. In the set of characteristic time periods of a certain occupation, obtains the set of recognition records in a certain characteristic time period of a certain day, collects the speculated destination end point and the actual destination end point of a certain recognition record, respectively collects the numbers of successful and failed recognition records in a certain characteristic time period of a certain day, calculates the recognition success rate in a certain characteristic time period of a certain day, summarizes the recognition success rates of a certain characteristic time period within the training period in chronological order, and uses the date as the X-axis and the recognition success rate as the Y-axis to draw the recognition success rate curve of a certain characteristic time period; Sets the recognition success rate threshold, marks the recognition success rate threshold on the recognition success rate curve of the certain characteristic time period, sets the abnormal number of days threshold, collects the number of consecutive days below the recognition success rate threshold, and if the number of days reaches the abnormal number of days threshold, collects the date of the last day among the number of days and sets it as the monitoring period of a certain characteristic time period; The calculation recognition success rate risk value unit: Collects the dates below the recognition success rate threshold within the monitoring period and sets them as abnormal dates, obtains the recognition success rates of the abnormal dates and sets them as abnormal recognition success rates, summarizes the abnormal dates, calculates the interval duration between each abnormal date, and calculates the risk value of the recognition success rate.

9. The traffic travel information recognition and management system based on data processing according to claim 6, characterized in that The dynamic adjustment module includes a real-time evaluation unit and a calculation dynamically adjusted monitoring period unit: The real-time evaluation unit: In the characteristic time period of a certain occupation, respectively collects the numbers of successful and failed real-time results of the recognition records, calculates the real-time recognition success rate of the characteristic time period within the monitoring period of the characteristic time period, and if the real-time recognition success rate is lower than the recognition success rate threshold, reminds the staff to re-divide the grid; The calculation dynamically adjusted monitoring period unit: If the real-time recognition success rate is lower than the risk value of the recognition success rate, increases the risk count by one, calculates the dynamically adjusted monitoring period, and calculates the recognition success rate of the characteristic time period according to the dynamically adjusted monitoring period.

Images