A digital safety monitoring and early warning system and method based on big data

Abstract

The present invention discloses a digital safety monitoring and early warning system and method based on big data, which relates to the field of digital monitoring technology. By using big data technology, historical traffic data is obtained and the traffic flow range is divided. By calculating the chaos index Hzs of each current monitoring period, abnormal fluctuations in traffic flow can be quickly detected. When the chaos index Hzs exceeds the set danger threshold Q, the system automatically issues a hidden state recognition instruction. This function helps to discover traffic hazards and abnormal conditions at an early stage, and timely measures can be taken to reduce traffic congestion and accident risks. After receiving the hidden state recognition instruction, the transformation probability matrix Bgz and the actual probability matrix Sgz are obtained respectively by using big data technology and statistical analysis means, and combined with forward variables, the optimal path probability of each monitoring period can be analyzed.

Description

Technical Field

[0001] The present invention relates to the field of digital monitoring technology, and in particular to a digital safety monitoring and early warning system and method based on big data. Background Art

[0002] With the rapid development of big data and digital technology, intelligent transportation systems (ITS) have become a key development direction in the field of traffic management and safety. Intelligent transportation systems are committed to improving the safety, efficiency and sustainability of road traffic through advanced data collection and analysis technologies. As part of the intelligent transportation system, digital safety monitoring and early warning systems play an important role in traffic flow analysis and real-time early warning. Especially in the safety management of traffic roads, timely monitoring and early warning can effectively reduce traffic accidents, improve road use efficiency, and ensure driving safety.

[0003] Although the digital safety monitoring and early warning method based on big data has significant advantages in improving traffic safety, the existing system still has certain shortcomings in practical applications. At present, traditional traffic safety monitoring and early warning methods often rely on static data and manual analysis, and have problems such as delayed response, insufficient data processing capabilities, and low early warning accuracy. Secondly, traditional methods are difficult to adapt to the rapidly changing traffic environment, have limited response capabilities to sudden traffic incidents, and are difficult to analyze the current hidden state in a timely manner, especially in the rapidly changing urban traffic environment, which greatly affects the timely maintenance of traffic order. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a digital safety monitoring and early warning system and method based on big data, which solves the problems in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a digital safety monitoring and early warning method based on big data, comprising the following steps:

[0006] S1. Pre-select a section of traffic road and use several sets of monitoring instruments to collect traffic data on the traffic road, convert the traffic data into digital format to obtain an observation set, and upload the observation set to the cloud platform after dimensionless processing;

[0007] S2. Use big data technology to obtain historical data, summarize the traffic flow range based on the historical data and divide the traffic status into three groups. At the same time, according to the observation set, construct the chaos index Hzs in each monitoring period. If the chaos index Hzs exceeds the danger threshold Q, issue a hidden state recognition instruction to the outside;

[0008] S3. After receiving the hidden state recognition instruction, according to the statistical analysis of the observation set and combined with the three groups of traffic states, the transformation probability matrix Bgz and the actual probability matrix Sgz are obtained respectively, and the initial analysis monitoring point is set to obtain the initial probability distribution matrix Cgz by statistics. Based on the initial probability distribution matrix Cgz and the actual probability matrix Sgz, the forward variable is constructed. , and analyze the probability of the optimal path under different traffic conditions during the corresponding monitoring period ;

[0009] S4: Based on the obtained optimal path probability in S3 , through backtracking, the hidden state of each monitoring period is obtained, and a better state sequence is constructed.

[0010] Preferably, S1 specifically includes:

[0011] S11. The pre-selected traffic road is used as the target area, and several groups of monitoring instruments are deployed around the target area to monitor and collect traffic data in the target area in real time, wherein the traffic data includes the number of vehicles running ahead in each monitoring period. , Lane change frequency , Traffic density and the average speed of vehicles ;

[0012] S12, converting the traffic data obtained in S11 from analog form to digital form to generate an observation set, and using dimensionless processing technology to standardize the data in the observation set to eliminate dimensional differences between different data sources.

[0013] Preferably, S2 specifically includes:

[0014] S21. Using big data technology to obtain historical data, the historical data includes the number of vehicles running ahead in each monitoring period during the historical period , Lane change frequency , Traffic density and the average speed of vehicles ;

[0015] S211. Based on historical data, construct traffic flow factors for each monitoring period in the historical period , which is obtained by the following formula:

[0016] ;

[0017] In the formula, is the maximum value of vehicle density on the traffic road;

[0018] S212, the traffic flow factor in each monitoring period of the historical period Statistics are conducted to summarize the traffic flow range and sort them according to the numerical value. The specific sorting method is: sort the values ​​in ascending order to generate a monitoring sequence, and divide the monitoring sequence into three equal parts, and obtain the No. 1 sub-monitoring sequence, No. 2 sub-monitoring sequence and No. 3 sub-monitoring sequence in turn. At the same time, the No. 1 sub-monitoring sequence, No. 2 sub-monitoring sequence and No. 3 sub-monitoring sequence are respectively marked as "low flow", "medium flow" and "high flow" to obtain three groups of traffic conditions.

[0019] Preferably, S2 also includes:

[0020] S22, based on the traffic flow factors in each monitoring period in the historical period The calculation method is used to obtain the traffic flow factor in each monitoring period. At the same time, according to the obtained observation set, the number of vehicles running ahead is and lane change frequency All are dimensionless and the corresponding data values ​​are mapped to the interval To obtain the chaos index Hzs in each monitoring period, the specific method is as follows:

[0021] ;

[0022] In the formula, , and Represented as the number of vehicles running ahead , Lane change frequency and traffic flow factor The weight value of , C is represented as the correction constant.

[0023] Preferably, S2 also includes:

[0024] S23. Preset a danger threshold Q, and compare and analyze the chaos index Hzs with the danger threshold Q to determine whether the traffic road in the corresponding monitoring period is in an abnormal state. The specific determination content is as follows:

[0025] S231, if the chaos index Hzs exceeds the danger threshold Q, the traffic road in the corresponding monitoring period will be judged to be in an abnormal state, and a hidden state recognition instruction will be issued externally;

[0026] S232. If the chaos index Hzs does not exceed the danger threshold Q, it will be determined that the traffic road in the corresponding monitoring period is not in an abnormal state, and no hidden state recognition instruction will be issued to the outside for the time being.

[0027] Preferably, S3 specifically includes:

[0028] S31. Preset the initial analysis monitoring points in a day, and calculate the probabilities of different traffic conditions at the initial analysis monitoring points according to the observation set to obtain the initial probability distribution matrix Cgz. The specific contents are as follows:

[0029] ;

[0030] in, , and They represent the probability of the initial monitoring point being at low flow, medium flow, and high flow respectively;

[0031] S32, after receiving the hidden state recognition instruction issued in S23, according to the statistical analysis of the observation set and combined with the three groups of traffic states, respectively obtain the transformation probability matrix Bgz and the actual probability matrix Sgz, the specific contents are as follows:

[0032] ; ;

[0033] in, Expressed as the transition probability from “low flow” to “low flow”, It is expressed as the transition probability from “low flow” to “medium flow”;

[0034] It is expressed as the actual probability of observing the traffic flow factor in the first monitoring period under the "low flow" state, It is expressed as the actual probability of observing the traffic flow factor in the first monitoring period under the "medium flow" state.

[0035] Preferably, S3 also includes:

[0036] S33, based on the initial probability distribution matrix Cgz and the actual probability matrix Sgz, analyze the forward variables of each traffic state at t=1 , the specific contents are as follows:

[0037] ;

[0038] In the formula, Expressed as the probability of being in state i at the initial analysis monitoring point, It is expressed as the probability of observing the traffic flow factor at the initial analysis monitoring point under state i.

[0039] Preferably, S3 also includes:

[0040] S34, based on the traffic flow factors in each monitoring period , analyze the probability of a better path under each traffic condition during the corresponding monitoring period The specific acquisition method is as follows:

[0041]

[0042] In the formula, It is expressed as the probability of the better path being in state j during the t-1th monitoring period; is the transition probability, which indicates the probability of transitioning from state j to state i; is the actual probability, which means the actual probability of observing the traffic flow factor in the tth monitoring period under state i.

[0043] Preferably, S4 specifically includes:

[0044] S41, according to the probability of the optimal path under each traffic state in the corresponding monitoring period obtained in S34 , compare the probability of the optimal path under each traffic condition in the same monitoring period , from which the probability of a better path is selected The maximum value of the better path probability The traffic state corresponding to the maximum value of is taken as the hidden state of the corresponding monitoring period;

[0045] S42. According to the method of obtaining the hidden state of the corresponding monitoring period in S41, the hidden states of different monitoring periods are obtained respectively to construct a better state sequence, and based on the obtained better state sequence, a line graph or a bar graph is used to show the change trend of the traffic state over time.

[0046] A digital safety monitoring and early warning system based on big data, including a data module, an instruction issuing module, a hidden state analysis module and a backtracking module;

[0047] The data module is used to pre-select a section of traffic road, and use a number of monitoring instruments to collect traffic data on the traffic road, and convert the traffic data into a digital format to obtain an observation set, and after dimensionless processing, upload the observation set to the cloud platform;

[0048] The instruction issuing module is used to use big data technology to obtain historical data, summarize the traffic flow range based on the historical data and divide the traffic status into three groups, and at the same time, construct the chaos index Hzs in each monitoring period according to the observation set. If the chaos index Hzs exceeds the danger threshold Q, a hidden state recognition instruction is issued externally;

[0049] The hidden state analysis module is used to obtain the transformation probability matrix Bgz and the actual probability matrix Sgz respectively according to the statistical analysis of historical data and the three groups of traffic states after receiving the hidden state recognition instruction, and set the initial monitoring point to obtain the initial probability distribution matrix Cgz by statistics. Based on the initial probability distribution matrix Cgz and the actual probability matrix Sgz, the forward variable According to the chaos index Hzs in each monitoring period, the probability of the optimal path under different traffic conditions in the corresponding monitoring period is analyzed. ;

[0050] The backtracking module is based on the obtained better path probability in S3 , through backtracking, the hidden state of each monitoring period is obtained, and a better state sequence is constructed.

[0051] The present invention provides a digital safety monitoring and early warning system and method based on big data, which has the following beneficial effects:

[0052] (1) By collecting traffic data in real time on the road and uploading it to the cloud platform for processing, this method can quickly identify the real-time status of road traffic, and the dimensionless observation set ensures data standardization, reduces the errors caused by different data types, and improves the reliability of data and the accuracy of analysis. By using big data technology, historical traffic data is obtained and the traffic flow range is divided, and the three groups of traffic states are accurately divided, which effectively combines historical and current traffic conditions and provides a more reliable reference standard for subsequent state identification and monitoring. This analysis method based on historical data can better cope with complex traffic changes and improve the accuracy of early warning. By calculating the chaos index Hzs of each current monitoring period, abnormal fluctuations in traffic flow can be quickly detected. When the chaos index Hzs exceeds the set danger threshold Q, the system automatically issues a hidden state recognition instruction. This function helps to discover traffic hazards and abnormal conditions at an early stage, and can take timely measures to reduce traffic congestion and accident risks. After receiving the hidden state recognition instruction, the transformation probability matrix Bgz and the actual probability matrix Sgz are obtained respectively by using big data technology and statistical analysis methods, and combined with the forward variables, the optimal path probability of each monitoring period can be analyzed. Through this method, the optimal state sequence of traffic in the corresponding monitoring period can be further inferred, making the entire traffic monitoring system more efficient and able to reasonably allocate traffic resources, thereby facilitating the optimization of traffic flow. Based on the optimal path probability obtained in the early stage, the backtracking method can accurately infer the hidden state of each monitoring period, thereby forming a better state sequence. This method not only improves the accuracy of traffic condition analysis, but also helps traffic management departments predict future traffic trends, formulate more scientific traffic management plans, and reduce traffic congestion and accidents. In general, this method can quickly and accurately monitor the real-time status of road traffic, predict future traffic flow changes, improve the intelligence level and response speed of the traffic management system, effectively prevent traffic accidents and improve road traffic efficiency through the digitization of traffic data and big data analysis.

[0053] (2) By using big data technology to obtain historical data, the system can conduct in-depth analysis of traffic from multiple dimensions and time periods, covering important traffic parameters such as the number of vehicles running ahead, lane change frequency, traffic density, and average vehicle speed. This large-scale analysis based on historical data can more accurately reflect the long-term trend of road operation, thereby providing more comprehensive basic information for the traffic management system. The calculation formula introduces the key parameter of the maximum vehicle density of traffic roads, which can effectively quantify the vehicle density of the road when it reaches the most congested state. Combined with the flow performance of each traffic period in the historical data, the system can construct a traffic flow factor that conforms to the actual situation, providing a reliable data basis for subsequent flow analysis and state classification. The system can classify different traffic conditions more accurately by sorting and dividing the sorted monitoring sequence into three groups of states: "low flow", "medium flow" and "high flow". This grouping method is not only convenient for managing and analyzing road operation conditions under different traffic flows, but also provides a clear basis for traffic status for further hidden state identification and early warning. By introducing a quantitative historical traffic flow range division model, the system provides specific numerical indicators for the decision-making process in traffic management. The division of different flow intervals makes traffic flow management more scientific and standardized, thus providing solid data support for daily traffic regulation, congestion relief and accident early warning. In summary, S2 makes traffic monitoring more accurate and efficient by applying big data technology and using historical data to calculate traffic flow factors and classify flow states. At the same time, the constructed flow state classification system can provide a more reliable basis for judgment for traffic management, help traffic managers better deal with different traffic conditions, and improve the overall operation efficiency of road traffic.

[0054] (3) By comparing the chaos index Hzs in different monitoring periods in real time, it is possible to quickly determine whether the traffic road is in an abnormal state, and automatically issue hidden state recognition instructions to provide early warnings for traffic management personnel. This function can effectively respond to emergencies or traffic congestion and take timely measures to avoid the deterioration of abnormal traffic conditions.

[0055] (4) The transformation probability matrix and the actual probability matrix can describe the transition between different traffic states and the probability of the actually observed traffic flow factors as accurately as possible. Through such a matrix, the system can effectively capture the dynamic laws of traffic state changes in order to better predict future traffic states, thereby providing traffic management personnel with more accurate traffic flow prediction information, which helps to make timely and effective decisions in actual operations and optimize traffic management and scheduling strategies. Through the comprehensive use of the initial probability distribution matrix, the transformation probability matrix, and the actual probability matrix, the system can better adapt to changes in different traffic conditions. This method ensures that the model can dynamically adjust parameters according to actual traffic data, improves the accuracy of traffic flow prediction and the adaptability of the system, and reduces the errors that may be caused by changes in traffic states.

[0056] (5) By comparing the probability of the optimal path under each traffic state in each monitoring period, the optimal hidden state in each period can be effectively screened out. This process ensures that the system can select a traffic state that is relatively consistent with the actual situation in each monitoring period, thereby improving the ability to accurately identify and predict changes in traffic flow states. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a flow chart of a digital safety monitoring and early warning method based on big data according to the present invention;

[0058] Figure 2 This is a block diagram of a digital safety monitoring and early warning system based on big data in the present invention. DETAILED DESCRIPTION

[0059] 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.

[0060] Example 1

[0061] See also Figure 1 The present invention provides a digital safety monitoring and early warning method based on big data, comprising the following steps:

[0062] S1. Pre-select a section of traffic road and use several sets of monitoring instruments to collect traffic data on the traffic road, convert the traffic data into digital format to obtain an observation set, and upload the observation set to the cloud platform after dimensionless processing;

[0063] S2. Use big data technology to obtain historical data, summarize the traffic flow range based on the historical data and divide the traffic status into three groups. At the same time, according to the observation set, construct the chaos index Hzs in each monitoring period. If the chaos index Hzs exceeds the danger threshold Q, issue a hidden state recognition instruction to the outside;

[0064] S3. After receiving the hidden state recognition instruction, according to the statistical analysis of the observation set and combined with the three groups of traffic states, the transformation probability matrix Bgz and the actual probability matrix Sgz are obtained respectively, and the initial analysis monitoring point is set to obtain the initial probability distribution matrix Cgz by statistics. Based on the initial probability distribution matrix Cgz and the actual probability matrix Sgz, the forward variable is constructed. , and analyze the probability of the optimal path under different traffic conditions during the corresponding monitoring period ;

[0065] S4: Based on the obtained optimal path probability in S3 , through backtracking, the hidden state of each monitoring period is obtained, and a better state sequence is constructed.

[0066] During the operation of this system, multiple groups of monitoring instruments in step S1 collect traffic road data in real time and upload the data to the cloud platform, further realizing real-time monitoring of traffic flow. Combined with the historical data analysis of big data technology, the system can dynamically reflect the current traffic conditions, especially through the observation set after dimensionless processing, which can ensure the uniformity of data format and provide an efficient basis for subsequent analysis. In step S2, the degree of chaos of traffic conditions is quantified by constructing the chaos index Hzs of each monitoring period to ensure that when the chaos index exceeds the set danger threshold Q, an early warning can be triggered in time. This early warning mechanism further enhances the foresight of traffic management and helps to take regulatory measures in time to prevent further deterioration of traffic conditions. In step S3, through statistical analysis, the transformation probability matrix Bgz, the actual probability matrix Sgz and the initial probability distribution matrix Cgz are used to construct forward variables and analyze the probability of the optimal path in the monitoring period. This analysis method based on probability model can effectively identify the hidden state of traffic and help predict the changes in traffic flow in subsequent periods. This feature can provide high-quality decision-making basis for traffic management departments and accurately grasp the traffic trends of roads. Step S4 traces back the optimal path, obtains the hidden state of each monitoring period, and constructs the optimal state sequence. This method helps to obtain the optimal traffic state evolution path in multiple time periods, thereby providing a quantifiable prediction basis for future changes in traffic flow, and then realizing intelligent management of road traffic. In summary, this method can effectively improve the accuracy, real-time and early warning capabilities of traffic flow monitoring by introducing big data technology, historical data analysis, probability models and chaos index, providing a more comprehensive and effective traffic flow analysis tool for traffic management departments, and providing strong support for traffic planning and emergency management.

[0067] Example 2

[0068] Please refer to Figure 1 , specifically: S1 specifically includes:

[0069] S11. The pre-selected traffic road is used as the target area, and several groups of monitoring instruments are deployed around the target area to monitor and collect traffic data in the target area in real time, wherein the traffic data includes the number of vehicles running ahead in each monitoring period. , Lane change frequency , Traffic density and the average speed of vehicles ;

[0070] S12. Convert the traffic data obtained in S11 from analog form to digital form to generate an observation set. Through digitization, the information is stored and processed in the form of binary code (0 and 1), so that it can be managed, analyzed and transmitted through computers and other digital devices; and use dimensionless processing technology to standardize the data in the observation set to eliminate the dimensional differences between different data sources, which is conducive to the unified representation of data of different magnitudes in the observation set in the same model, facilitating the analysis of the proportional relationship of the system.

[0071] In this embodiment, by deploying several groups of monitoring instruments in the target area, the traffic data in each monitoring period is collected in real time. These multi-dimensional data provide rich information and help to reflect the traffic status in an all-round way. Through multi-faceted data collection, a more accurate assessment of the traffic situation can be made, further supporting the subsequent traffic status analysis and early warning. The collected traffic data is converted from analog form to digital form, stored and processed in binary code (0 and 1), making data management, analysis and transmission more efficient and accurate. This digital processing not only improves the data processing speed, but also helps to realize automatic management and analysis of traffic data in computer systems, further reducing manual intervention and potential errors. Through dimensionless processing technology, traffic data from different sources and types are standardized, eliminating dimensional differences. Dimensionless processing helps to ensure that all data in the observation set can be analyzed at a unified scale. This data standardization makes the traffic status analysis more consistent and improves the accuracy of the analysis results. By standardizing different data, the rationality of the proportional relationship between the data sources is ensured. This processing step lays the foundation for subsequent big data analysis, making the calculation and analysis of the traffic status model more scientific. In general, the implementation of this method in the S1 stage not only improves the efficiency and comprehensiveness of traffic data collection, but also enhances the uniformity and comparability of data through digitalization and dimensionless processing technology, providing more accurate basic data support for subsequent big data analysis and hidden state identification, and ultimately improving the reliability and intelligence level of traffic safety monitoring and early warning systems.

[0072] Example 3

[0073] Please refer to Figure 1 , specifically: S2 specifically includes:

[0074] S21. Using big data technology to obtain historical data, the historical data includes the number of vehicles running ahead in each monitoring period during the historical period , Lane change frequency , Traffic density and the average speed of vehicles ;

[0075] S211. Based on historical data, construct traffic flow factors for each monitoring period in the historical period , which is obtained by the following formula:

[0076] ;

[0077] In the formula, It is the maximum value of vehicle density on the traffic road, indicating the number of vehicles when the traffic road reaches the most congested state, and the unit is vehicle / km;

[0078] The average driving speed It will affect the speed at which vehicles pass through a certain section of road. When the speed is low, the time that vehicles occupy the road increases, which will reduce the number of vehicles passing through the road per unit time. Therefore, It reflects the negative impact of vehicle speed on traffic flow;

[0079] It is to consider the impact of traffic density on flow and reflects the current traffic density Relative to the maximum density, the increase in flow is suppressed as the density increases. Specifically, It is the maximum value of vehicle density on the traffic road. At this density, the road reaches the most crowded state, the distance between vehicles decreases, the speed decreases, and the traffic flow is restricted;

[0080] Traffic density The higher it is, the more vehicles per unit road length there are, causing the speed to drop, thus affecting the flow.

[0081] S212, the traffic flow factor in each monitoring period of the historical period Statistics are conducted to summarize the traffic flow range and sort them according to the numerical value. The specific sorting method is: sort the values ​​in ascending order to generate a monitoring sequence, and divide the monitoring sequence into three equal parts, and obtain the No. 1 sub-monitoring sequence, No. 2 sub-monitoring sequence and No. 3 sub-monitoring sequence in turn. At the same time, the No. 1 sub-monitoring sequence, No. 2 sub-monitoring sequence and No. 3 sub-monitoring sequence are respectively marked as "low flow", "medium flow" and "high flow" to obtain three groups of traffic conditions.

[0082] In this embodiment, historical data is obtained by using big data technology, covering multiple sets of information in different monitoring periods in the historical period. This method not only provides extensive traffic historical data support, but also ensures the comprehensiveness and accuracy of the analysis results, which helps the traffic management department to better understand the historical changes in traffic flow, thereby providing strong data support for future traffic management. The traffic flow factor in each monitoring period in the historical period calculated by the formula , can accurately quantify the characteristics of historical traffic flow, especially by calculating the maximum value of vehicle density to determine the congestion limit of traffic roads. This quantification process provides an accurate basis for traffic flow analysis, enabling the system to better grasp the key variables under traffic conditions and improve the accuracy of traffic prediction and monitoring. By counting historical traffic flow factors and sorting them by numerical value, the system can divide different traffic flow levels in a scientific way. Specifically, the historical flow factors are sorted and divided into three groups (low flow, medium flow and high flow). This classification method provides a standardized reference basis for further analysis and optimization management of traffic status, and simplifies the complexity of subsequent traffic management decisions. In summary, step S2 collects and analyzes historical traffic flow through big data technology, constructs traffic flow factors, and classifies and summarizes different flow states. This processing method based on historical data not only improves the accuracy of traffic flow analysis, but also provides a clear decision-making basis for traffic management, which ultimately helps safety warning and traffic optimization management of traffic roads.

[0083] Example 4

[0084] Please refer to Figure 1 Specifically, S2 also includes:

[0085] S22, based on the traffic flow factors in each monitoring period in the historical period The calculation method is used to obtain the traffic flow factor in each monitoring period. At the same time, according to the obtained observation set, the number of vehicles running ahead is and lane change frequency All are dimensionless and the corresponding data values ​​are mapped to the interval To obtain the chaos index Hzs in each monitoring period, the specific method is as follows:

[0086] ;

[0087] In the formula, , and Represented as the number of vehicles running ahead , Lane change frequency and traffic flow factor The weight value of C is expressed as a correction constant, where 0 < ≤1,0< ≤1,0< ≤1, and + + =1.

[0088] The number of vehicles mentioned above and lane change frequency All of them can be detected and obtained through the camera;

[0089] S2 also includes:

[0090] S23. Preset a danger threshold Q, and compare and analyze the chaos index Hzs with the danger threshold Q to determine whether the traffic road in the corresponding monitoring period is in an abnormal state. The specific determination content is as follows:

[0091] S231, if the chaos index Hzs exceeds the danger threshold Q, the traffic road in the corresponding monitoring period will be judged to be in an abnormal state, and a hidden state recognition instruction will be issued externally;

[0092] S232. If the chaos index Hzs does not exceed the danger threshold Q, it will be determined that the traffic road in the corresponding monitoring period is not in an abnormal state, and no hidden state recognition instruction will be issued to the outside for the time being.

[0093] It should be noted that the danger threshold Q can be obtained in the following ways:

[0094] The chaos index Hzs in each current monitoring period is counted, and the average chaos index in the monitoring period is obtained by combining statistical algorithms. and its standard deviation , and based on the average chaos index during the monitoring period and its standard deviation , set the danger threshold Q, where ; Where k is a constant, usually taking a value of 1-3, corresponding to different confidence levels, and the specific value is set by the user (based on actual conditions).

[0095] In this embodiment, the chaos index Hzs is constructed by a multi-dimensional and multi-parameter comprehensive calculation method, which can evaluate the traffic conditions in many aspects, integrate multiple key factors affecting traffic flow into a quantitative indicator, and further improve the accuracy and effectiveness of monitoring. In S22, the observed data is dimensionless, so that different types of traffic data can be analyzed uniformly in the same calculation model, effectively eliminating the influence of dimensional differences on data processing, which not only improves the comparability of the analysis results, but also enhances the versatility and flexibility of the model, and can adapt to practical applications under different roads and traffic environments. By calculating the chaos index Hzs, the system can quickly and accurately evaluate the traffic conditions in each monitoring period. Hzs not only reflects the behavioral characteristics of the vehicle, but also combines the changing trend of the overall traffic flow, and can effectively identify potential traffic anomalies or congestion risks. In S23, a dangerous threshold Q is set. By comparing the chaos index Hzs with the threshold Q, it can be judged in real time whether the traffic road is in an abnormal state. This automatic judgment mechanism based on big data and algorithms enables the traffic management system to quickly identify abnormal traffic conditions and issue hidden state recognition instructions, further improving the efficiency of traffic monitoring and response. When the chaos index Hzs exceeds the danger threshold Q, the system automatically identifies abnormal traffic conditions and triggers hidden state recognition instructions. This automated early warning mechanism ensures that traffic managers can quickly take countermeasures, further reducing potential accidents and congestion risks, and improving the intelligent level of traffic safety monitoring. In summary, through the design of S22 and S23, it is possible to further realize the comprehensive analysis of multi-dimensional traffic parameters, accurately reflect traffic conditions, and improve the efficiency and accuracy of traffic monitoring and early warning through the automated abnormal state recognition mechanism. This method not only helps traffic managers to find problems in a timely manner, but also effectively reduces the occurrence of traffic accidents and congestion, and improves the overall safety and operation efficiency of the transportation system.

[0096] Example 5

[0097] Please refer to Figure 1 , specifically: S3 specifically includes:

[0098] S31. Preset the initial analysis monitoring points in a day, and calculate the probabilities of different traffic conditions at the initial analysis monitoring points according to the observation set to obtain the initial probability distribution matrix Cgz. The specific contents are as follows:

[0099] ;

[0100] in, , and They represent the probability of the initial monitoring point being at low flow, medium flow, and high flow respectively;

[0101] S32, after receiving the hidden state recognition instruction issued in S23, according to the statistical analysis of the observation set and combined with the three groups of traffic states, respectively obtain the transformation probability matrix Bgz and the actual probability matrix Sgz, the specific contents are as follows:

[0102] ; ;

[0103] in, Expressed as the transition probability from “low flow” to “low flow”, It is expressed as the transition probability from "low flow" to "medium flow". Similarly, the other parameters mentioned above are also the same;

[0104] It is expressed as the actual probability of observing the traffic flow factor in the first monitoring period under the "low flow" state, It is expressed as the actual probability of observing the traffic flow factor in the first monitoring period under the "medium flow" state. Similarly, the same is true for the other parameters mentioned above.

[0105] Example: Assume that we have traffic flow data for the past 5 hours, which is divided into three states: low flow, medium flow and high flow. We record the traffic flow state every hour, specifically: low flow in the first hour, low flow in the second hour, high flow in the third hour, medium flow in the fourth hour and low flow in the fifth hour. After statistics, we know that the number of times from low flow to low flow is 1, the number of times from low flow to medium flow is 0, the number of times from low flow to high flow is 1, and so on.

[0106] From statistics, we know that the number of times from low flow to low flow is 1, so the total number of times from the beginning is 3. , and so on;

[0107] For the actual probability matrix Sgz, for example, It is the actual probability of observing the traffic flow factor during the first monitoring period under the "low flow" state. In simple terms, it is a statistical theory.

[0108] In this embodiment, by setting the initial analysis monitoring point in a day and counting the probabilities under different traffic conditions based on the observation set, the initial probability distribution matrix Cgz can be constructed. This matrix clarifies the probability distribution of different traffic conditions (low flow, medium flow, and high flow) at the initial monitoring point. The accurate initial probability distribution provides a reliable starting point for subsequent state prediction and analysis, ensuring that the entire prediction process starts from an accurate basis. After receiving the hidden state recognition instruction, the statistical data of the observation set and the three sets of traffic state information can be used to construct the transformation probability matrix Bgz and the actual probability matrix Sgz. The transformation probability matrix Bgz reflects the transformation probability between traffic states. This information plays an important role in understanding the trend of traffic flow changes and predicting future state transfers, and can help traffic management personnel to perform more accurate flow control and scheduling. The actual probability matrix Sgz is calculated based on actual observation data, so that the model can more accurately reflect the actual situation of the current traffic flow factor. These two matrices combine the theoretical model with actual data, enhancing the adaptability and accuracy of the model to actual traffic conditions.

[0109] In summary, S3 establishes the initial probability distribution matrix, the transformed probability matrix and the actual probability matrix. This comprehensive analysis can not only prepare for reducing traffic congestion and accident rates, but also help optimize the predictive ability of traffic conditions.

[0110] Example 6

[0111] Please refer to Figure 1 , specifically: S3 also includes:

[0112] S33, based on the initial probability distribution matrix Cgz and the actual probability matrix Sgz, analyze the forward variables of each traffic state at t=1 , the specific contents are as follows:

[0113] ;

[0114] In the formula, Expressed as the probability of being in state i at the initial analysis monitoring point, It is expressed as the probability of observing the traffic flow factor at the initial analysis monitoring point under state i.

[0115] S3 also includes:

[0116] S34, based on the traffic flow factors in each monitoring period , analyze the probability of a better path under each traffic condition during the corresponding monitoring period The specific acquisition method is as follows:

[0117]

[0118] In the formula, It is expressed as the probability of the better path being in state j during the t-1th monitoring period; is the transition probability, which indicates the probability of transitioning from state j to state i; The purpose is to find the optimal path from time t-1 to time t, that is, the maximum probability of all possible paths from a state j at t-1 to a state i at t; is the actual probability, which indicates the actual probability of observing the traffic flow factor in the tth monitoring period under state i. It is used to evaluate whether the current observation value is likely to be observed under state i. Here, the observation value is the traffic flow factor.

[0119] t is the current monitoring period, i.e., the current time step; i and j represent the index of the state, i is the state at time t, and j is the state at time t-1;

[0120] Example: Now calculate the probability of a better path under each traffic condition in the second monitoring period , at this time, the probability of the optimal path under each traffic state in the second monitoring period will be calculated separately , specifically,

[0121]

[0122]

[0123]

[0124] If calculated, we get =0.0756, =0.102, =0.0276;

[0125] At this time, by comparing the probability of the optimal path under each traffic state in the second monitoring period, , from which the probability of a better path is selected The maximum value of the better path probability The traffic state corresponding to the maximum value of is taken as the hidden state of the corresponding monitoring period. Here, the traffic state of medium flow is the hidden state of the monitoring period.

[0126] S4 specifically includes:

[0127] S41, according to the probability of the optimal path under each traffic state in the corresponding monitoring period obtained in S34 , compare the probability of the optimal path under each traffic condition in the same monitoring period , from which the probability of a better path is selected The maximum value of the better path probability The traffic state corresponding to the maximum value of is taken as the hidden state of the corresponding monitoring period;

[0128] S42. According to the method of obtaining the hidden state of the corresponding monitoring period in S41, the hidden states of different monitoring periods are obtained respectively to construct a better state sequence, and based on the obtained better state sequence, a line graph or a bar graph is used to show the change trend of the traffic state over time.

[0129] In this embodiment, by using the initial probability distribution matrix and the actual probability matrix, the forward variables under each traffic state can be accurately analyzed. This analysis provides the optimal path probability under each state in the current monitoring period t=1, providing a scientific basis for subsequent state prediction. By calculating the probability of the better path under each traffic state in each monitoring period, the optimal path from one time step to the next time step can be clearly identified, which helps to identify the most likely state sequence from all possible paths, thereby improving the accuracy of traffic state prediction. In each monitoring period, by comparing the probability of the better path, the hidden state that is relatively consistent with the actual observation can be accurately identified. This process ensures that the identification of traffic states at different time points is more accurate, thereby providing reliable data support for traffic management. By obtaining the hidden states of different monitoring periods and constructing a better state sequence, the system can form a comprehensive understanding of the traffic state during the entire monitoring period. This sequence helps traffic management personnel identify the regularity of overall traffic flow changes and make more accurate predictions of future traffic flow changes. Using line graphs or bar graphs to display the changing trends of traffic conditions over time can intuitively present the changing patterns of traffic conditions. This visualization method enables traffic management personnel to quickly identify the fluctuation trends of traffic flow and take corresponding management measures, thereby improving the efficiency and accuracy of decision-making.

[0130] By analyzing the hidden state of each monitoring period and constructing a better state sequence, the system can monitor the changes in traffic flow in real time. This real-time monitoring capability enables traffic management personnel to understand the changes in traffic status in a timely manner and respond quickly when abnormal situations occur, thereby optimizing traffic management strategies and reducing the risk of traffic congestion and accidents. Combined with the analysis of forward variables and the probability of the better path, the system can predict future traffic status based on a data-driven approach. This method provides a data-supported decision-making basis, enhances the scientificity and rationality of traffic management, and helps to formulate more accurate traffic management measures and strategies. In summary, steps S3 and S4 not only improve the accuracy of traffic status prediction through precise probability analysis and state sequence construction, but also enhance the system's real-time monitoring and response capabilities. This method provides scientific decision-making support, helps to optimize traffic management, improve traffic flow, and improve the overall efficiency and safety of the transportation system.

[0131] Example 7

[0132] Please refer to Figure 2 ,Specifically: A digitized security monitoring and early warning system based on big data includes a data module, an instruction issuing module, a hidden state analysis module and a backtracking module;

[0133] The data module is used to pre-select a section of traffic road, and use a number of monitoring instruments to collect traffic data on the traffic road, and convert the traffic data into a digital format to obtain an observation set, and after dimensionless processing, upload the observation set to the cloud platform;

[0134] The instruction issuing module is used to use big data technology to obtain historical data, summarize the traffic flow range based on the historical data and divide the traffic status into three groups, and at the same time, construct the chaos index Hzs in each monitoring period according to the observation set. If the chaos index Hzs exceeds the danger threshold Q, a hidden state recognition instruction is issued externally;

[0135] The hidden state analysis module is used to obtain the transformation probability matrix Bgz and the actual probability matrix Sgz respectively according to the statistical analysis of historical data and the three groups of traffic states after receiving the hidden state recognition instruction, and set the initial monitoring point to obtain the initial probability distribution matrix Cgz by statistics. Based on the initial probability distribution matrix Cgz and the actual probability matrix Sgz, the forward variable According to the chaos index Hzs in each monitoring period, the probability of the optimal path under different traffic conditions in the corresponding monitoring period is analyzed. ;

[0136] The backtracking module is based on the obtained better path probability in S3 , through backtracking, the hidden state of each monitoring period is obtained, and a better state sequence is constructed.

[0137] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A digital safety monitoring and early warning method based on big data, characterized by: The following steps are included: S1. Pre-select a section of traffic road and use several sets of monitoring instruments to collect traffic data on the traffic road, convert the traffic data into digital format to obtain an observation set, and upload the observation set to the cloud platform after dimensionless processing; S1 specifically includes: S11. The pre-selected traffic road is used as the target area, and several groups of monitoring instruments are deployed around the target area to monitor and collect traffic data in the target area in real time, wherein the traffic data includes the number of vehicles running ahead in each monitoring period. , Lane change frequency , Traffic density and the average speed of vehicles ; S12, converting the traffic data obtained in S11 from analog form to digital form to generate an observation set, and using dimensionless processing technology to standardize the data in the observation set to eliminate dimensional differences between different data sources; S2. Use big data technology to obtain historical data, summarize the traffic flow range based on the historical data and divide the traffic status into three groups. At the same time, according to the observation set, construct the chaos index Hzs in each monitoring period. If the chaos index Hzs exceeds the danger threshold Q, issue a hidden state recognition instruction to the outside; S2 specifically includes: S21. Using big data technology to obtain historical data, the historical data includes the number of vehicles running ahead in each monitoring period during the historical period , Lane change frequency , Traffic density and the average speed of vehicles ; S211. Based on historical data, construct traffic flow factors for each monitoring period in the historical period , which is obtained by the following formula: ； In the formula, is the maximum value of vehicle density on the traffic road; S212, the traffic flow factor in each monitoring period of the historical period Perform statistics, summarize the traffic flow range and sort by value. The specific sorting method is: sort the values ​​in ascending order to generate a monitoring sequence, and divide the monitoring sequence into three equal parts, and obtain the first sub-monitoring sequence, the second sub-monitoring sequence and the third sub-monitoring sequence in turn. At the same time, mark the first sub-monitoring sequence, the second sub-monitoring sequence and the third sub-monitoring sequence as "low flow", "medium flow" and "high flow" respectively to obtain three groups of traffic states; S2 also includes: S22, based on the traffic flow factors in each monitoring period in the historical period The calculation method is used to obtain the traffic flow factor in each monitoring period. At the same time, according to the obtained observation set, the number of vehicles running ahead is and lane change frequency All are dimensionless and the corresponding data values ​​are mapped to the interval To obtain the chaos index Hzs in each monitoring period, the specific method is as follows: ； In the formula, , and Represented as the number of vehicles running ahead , Lane change frequency and traffic flow factor The weight value of , C is represented as the correction constant; S3. After receiving the hidden state recognition instruction, according to the statistical analysis of the observation set and combined with the three groups of traffic states, the transformation probability matrix Bgz and the actual probability matrix Sgz are obtained respectively, and the initial analysis monitoring point is set to obtain the initial probability distribution matrix Cgz by statistics. Based on the initial probability distribution matrix Cgz and the actual probability matrix Sgz, the forward variable is constructed. , and analyze the probability of the optimal path under different traffic conditions during the corresponding monitoring period ; S3 specifically includes: S31. Preset the initial analysis monitoring points in a day, and calculate the probabilities of different traffic conditions at the initial analysis monitoring points according to the observation set to obtain the initial probability distribution matrix Cgz. The specific contents are as follows: ； in, , and They represent the probability of the initial monitoring point being at low flow, medium flow, and high flow respectively; S32, after receiving the hidden state recognition instruction issued in S23, according to the statistical analysis of the observation set and combined with the three groups of traffic states, respectively obtain the transformation probability matrix Bgz and the actual probability matrix Sgz, the specific contents are as follows: ； ； in, Expressed as the transition probability from "low flow" to "low flow", Expressed as the transition probability from "low flow" to "medium flow"; It is expressed as the actual probability of observing the traffic flow factor in the first monitoring period under the "low flow" state, It is expressed as the actual probability of observing the traffic flow factor in the first monitoring period under the "medium flow" state; S3 also includes: S33, based on the initial probability distribution matrix Cgz and the actual probability matrix Sgz, analyze the forward variables of each traffic state at t=1 , the specific contents are as follows: ； In the formula, Expressed as the probability of being in state i at the initial analysis monitoring point, It is expressed as the probability of observing the traffic flow factor at the initial analysis monitoring point under state i; S3 also includes: S34, based on the traffic flow factors in each monitoring period , analyze the probability of a better path under each traffic condition during the corresponding monitoring period The specific acquisition method is as follows: ； In the formula, It is expressed as the probability of the better path being in state j during the t-1th monitoring period; is the transition probability, which indicates the probability of transitioning from state j to state i; is the actual probability, which means the actual probability of observing the traffic flow factor in the tth monitoring period under state i; S4: Based on the obtained optimal path probability in S3 , through backtracking, the hidden state of each monitoring period is obtained, and a better state sequence is constructed.

2. According to the big data-based digital safety monitoring and early warning method of claim 1, it is characterized by: S2 also includes: S23. Preset a danger threshold Q, and compare and analyze the chaos index Hzs with the danger threshold Q to determine whether the traffic road in the corresponding monitoring period is in an abnormal state. The specific determination content is as follows: S231, if the chaos index Hzs exceeds the danger threshold Q, the traffic road in the corresponding monitoring period will be judged to be in an abnormal state, and a hidden state recognition instruction will be issued externally; S232. If the chaos index Hzs does not exceed the danger threshold Q, it will be determined that the traffic road in the corresponding monitoring period is not in an abnormal state, and no hidden state recognition instruction will be issued to the outside for the time being.

3. According to the big data-based digital safety monitoring and early warning method of claim 1, it is characterized by: S4 specifically includes: S41, according to the probability of the optimal path under each traffic state in the corresponding monitoring period obtained in S34 , compare the probability of the optimal path under each traffic condition in the same monitoring period , from which the probability of a better path is selected The maximum value of the better path probability The traffic state corresponding to the maximum value of is taken as the hidden state of the corresponding monitoring period; S42. According to the method of obtaining the hidden state of the corresponding monitoring period in S41, the hidden states of different monitoring periods are obtained respectively to construct a better state sequence, and based on the obtained better state sequence, a line graph or a bar graph is used to show the change trend of the traffic state over time.

4. A digital safety monitoring and early warning system based on big data, used to implement a digital safety monitoring and early warning method based on big data as described in any one of claims 1 to 3 above, characterized in that: It includes data module, instruction issuing module, hidden state analysis module and backtracking module; The data module is used to pre-select a section of traffic road, and use a number of monitoring instruments to collect traffic data on the traffic road, and convert the traffic data into a digital format to obtain an observation set, and after dimensionless processing, upload the observation set to the cloud platform; The instruction issuing module is used to use big data technology to obtain historical data, summarize the traffic flow range based on the historical data and divide the traffic status into three groups, and at the same time, construct the chaos index Hzs in each monitoring period according to the observation set. If the chaos index Hzs exceeds the danger threshold Q, a hidden state recognition instruction is issued externally; The hidden state analysis module is used to obtain the transformation probability matrix Bgz and the actual probability matrix Sgz respectively according to the statistical analysis of historical data and the three groups of traffic states after receiving the hidden state recognition instruction, and set the initial monitoring point to obtain the initial probability distribution matrix Cgz by statistics. Based on the initial probability distribution matrix Cgz and the actual probability matrix Sgz, the forward variable According to the chaos index Hzs in each monitoring period, the probability of the optimal path under different traffic conditions in the corresponding monitoring period is analyzed. ; The backtracking module is based on the obtained better path probability in S3 , through backtracking, the hidden state of each monitoring period is obtained, and a better state sequence is constructed.

Images