Road traffic operation state evaluation method and system based on data driving

Abstract

The invention discloses a road traffic operation state evaluation method and system based on data driving, and the method comprises the steps: calculating road evaluation indexes, carrying out the multi-level division, obtaining the grade of the road evaluation indexes, carrying out the dynamic incremental learning, obtaining a first index weight, carrying out the fuzzy adjustment, obtaining a second index weight, and constructing a fuzzy consistent matrix, establishing an extension matter element, performing deep learning on the road evaluation index to determine a classical domain and a node domain, calculating a classical domain distance and a node domain distance of the extension matter element of the to-be-evaluated road, calculating an evaluation association degree between the road evaluation index and the grade of the road evaluation index, determining a weight vector of the to-be-evaluated road according to the fuzzy consistency matrix, and determining a weight vector of the to-be-evaluated road according to the weight vector. And carrying out extension evaluation according to the weight vector and the evaluation association degree to obtain a road traffic operation state evaluation grade. The method not only can improve the efficiency and accuracy of road traffic operation state evaluation, but also has good interpretability, and can be directly applied to a road traffic operation state evaluation system.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of state evaluation, in particular to a highway traffic operation state evaluation method and system based on data driving. BACKGROUND

[0002] With the continuous acceleration of urbanization and the rapid growth of motor vehicle ownership, the highway traffic system is facing unprecedented pressure. Traffic congestion, frequent accidents, and low operational efficiency seriously restrict the high-quality development of regional economy and society, and also affect social safety. Therefore, it is of great significance to scientifically, accurately and timely evaluate the highway traffic operation state, so as to improve the traffic capacity of the highway network, ensure the safe and stable operation of the traffic system, and promote the construction of intelligent traffic system.

[0003] Traditional highway traffic operation state evaluation methods mostly rely on experience-based index selection, static weight distribution or simple statistical analysis models, which have many limitations. On the one hand, the index system is often not comprehensive, and the determination of index weight lacks dynamic adaptability, which cannot respond to real-time changes in traffic flow. On the other hand, the evaluation model does not consider the uncertainty and fuzziness of traffic data, has weak generalization ability, and is difficult to cope with complex and variable traffic scenarios, resulting in poor accuracy and timeliness of the evaluation results. In recent years, with the rapid development of big data, artificial intelligence and other technologies, new ideas have been provided to solve the above problems. The present application proposes a highway traffic operation state evaluation method and system based on data driving, which divides data and index levels by decision tree, determines index weight by dynamic incremental learning and fuzzy adjustment, introduces extension matter element and deep learning to construct evaluation correlation degree, and realizes final level determination by fuzzy consistent matrix and extension evaluation. This method not only can comprehensively integrate key indicators such as flow, speed and density, improve the dynamicity and rationality of weight determination, but also can enhance the adaptability and anti-interference ability of the evaluation model to complex traffic environment, effectively improve the accuracy and reliability of the evaluation results, and provide stronger technical support for highway traffic management and decision-making, which has important practical value for promoting the intelligent and fine management of highway traffic system. SUMMARY

[0004] The purpose of the present application is to provide a highway traffic operation state evaluation method and system based on data driving.

[0005] To achieve the above purpose, the present application is implemented according to the following technical solutions: The present application comprises the following steps: The highway traffic operation data is divided by decision tree and the highway evaluation index is calculated, and the highway evaluation index is divided into multiple levels to obtain the highway evaluation index level; the highway evaluation index includes flow type evaluation index, speed type evaluation index and density type evaluation index; The first index weight is obtained by dynamic incremental learning of the highway evaluation index, the second index weight is obtained by fuzzy adjustment of the first index weight, and a fuzzy consistent matrix is constructed according to the second index weight; According to the highway evaluation index grade and the highway evaluation index, an extension matter element is established, a classic field and a section field are determined by deep learning of the highway evaluation index, a classic field distance and a section field distance of the extension matter element of the to-be-evaluated highway are calculated, and an evaluation correlation degree of the highway evaluation index and the highway evaluation index grade is calculated according to the classic field distance and the section field distance; The weight vector of the to-be-evaluated highway is determined according to the fuzzy consistent matrix, and a highway traffic running state evaluation grade is obtained by extension evaluation according to the weight vector and the evaluation correlation degree.

[0006] Further, the method for obtaining the highway evaluation index grade comprises: The highway traffic running data is divided into a flow class, a speed class and a density class by using a decision tree, and a flow class evaluation index, a speed class evaluation index and a density class evaluation index are calculated according to the division result; the flow class evaluation index comprises a flow growth rate, a saturation degree and a saturation balance coefficient; the speed class evaluation index comprises a speed change rate, a free flow speed ratio and a travel time stability; and the density class evaluation index comprises a space occupancy rate and a space-time occupancy rate. The basic three-level threshold of each highway evaluation index is determined according to an industry standard, the correlation coefficient between each index is calculated respectively, two highway evaluation indexes with a correlation coefficient greater than 0.8 are taken as strong correlation indexes, and the rest are taken as weak correlation indexes, time series analysis is performed on the historical data of the strong correlation indexes to obtain a time series fluctuation coefficient, the historical data of the strong correlation indexes is divided into four indexes according to the basic three-level threshold, the data skewness and the data standard deviation of the same strong correlation index at different levels are calculated, the product of the data skewness mean and the data standard deviation mean of the same strong correlation index at different levels is taken as a data fluctuation coefficient, the basic three-level threshold of the strong correlation index is adjusted to obtain an optimized three-level threshold according to the time series fluctuation coefficient and the data fluctuation coefficient, the weak correlation indexes are divided according to the basic three-level threshold, and the strong correlation indexes are divided according to the optimized three-level threshold to obtain the highway evaluation index grade. Further, the method for obtaining the first index weight comprises: The initial weight is obtained by processing the historical highway evaluation index by using an entropy weight method, the first index weight is obtained by adjusting the initial weight according to The feature matrix is constructed by using the real-time traffic data sample , The feature matrix set is The sample number is The data dimension is vehicle type, highway evaluation index, spatial position and time stamp; The current data sample is calculated according to the historical sample reference distribution Wasserstein distance at time t, expressed as: where is the time Wasserstein distance at time t, is the historical reference data sample, is the set of all possible joint distributions, is the current data feature mapping function, is the joint distribution lower feature mapping and the expectation of squared Euclidean distance, is the spatial feature weight matrix, is the temporal weight matrix, is the function that encodes spatial coordinates into a string, is the timestamp; drift determination: when trigger weight update, where is the dynamic threshold, taking the 90th percentile of historical difference, is the noise tolerance; update the initial weight to obtain the first weight through a double-channel learning mechanism; the double-channel learning mechanism includes short-term incremental adjustment and long-term incremental adjustment, expressed as: where is the highway evaluation index the first index weight updated at time t, is the initial weight of the highway evaluation index is the short-term weight, is the short-term adjustment amount at time t, is the long-term weight, is the long-term adjustment amount of the highway evaluation index at time t, is the learning rate, is the sign function, is 1 when is -1 when is 0 when ​​​For Cross-entropy loss function at time t Gradient vector with respect to weights, is a smoothing factor, is a highway evaluation index and state Time-varying mutual information, is a highway evaluation index at the previous time t-1 First index weight.

[0007] Further, the method for constructing the fuzzy consistent matrix comprises: According to the radial basis-Gaussian mixed fuzzy kernel algorithm, the kernel function between the first index weights is calculated, and the second index weights are obtained by adjusting the first index weights through the kernel function mapping, and the expression is: Wherein is a highway evaluation index At The second index weight updated at time t, is an adjustment intensity, is the number of highway evaluation indexes, is a highway evaluation index And Radial basis-Gaussian mixed fuzzy kernel function, is a space-time adjustment factor, is a weight similarity sensitivity coefficient, is the number of Gaussian components, is the component weight, is a multivariate Gaussian distribution, is the mean of the Gaussian component, is the covariance matrix. According to the second index weight, the fuzzy consistent matrix is constructed , the expression is: Wherein is the fuzzy consistent matrix Element, , The two largest second index weights in the second index weight.

[0008] Further, the method for calculating the classical domain distance and the node domain distance of the extension matter-element of the road to be evaluated comprises: According to the highway evaluation index grade and the highway evaluation index definition extension matter-element , wherein is a set of highway evaluation index grades, is a set of highway evaluation indexes, for about 's evaluation set; Construct a boundary generation network and input multi-source traffic data tensors into the boundary generation network to obtain the classic domain interval and section interval ; The multi-source traffic data tensor includes highway evaluation indicators, time steps and number of road sections; The boundary generation network includes an encoder and a decoder. The encoder receives multi-source traffic data tensors and uses a bidirectional LSTM to extract temporal features to obtain a fixed-length context vector. The decoder processes the context vector according to the domain generation signal and outputs the classical domain and section domain of the corresponding level of the domain generation signal. The boundary generation network is obtained by training with a double-layer loss function. The double-layer loss function includes the classical domain boundary loss function and the section domain boundary loss function, which are expressed as: in is a double-layer loss function, For level The sample set of highway evaluation indicators is is the classic domain boundary loss function, is the section boundary loss function, is the section loss weight, For level The lower bound of the classical domain, For level The upper bound of the classical domain, is the interval width penalty coefficient, is the lower bound of the node domain, is the upper bound of the node domain; The classical domain distance of the extension matter-element of the highway to be evaluated is obtained by adaptively calculating the weighted distance according to the output classical domain interval, and the nodal domain distance of the extension matter-element of the highway to be evaluated is obtained by dynamically compensating the output nodal domain interval. The expression is: in 、 The index values ​​are Highway evaluation index The classical domain distance and node domain distance of is the interval width weight, is the attention sensitivity coefficient, For level The cluster center of is the drift compensation coefficient, is the time window length, for Moment highway evaluation index The value of .

[0009] Furthermore, the method for performing extension evaluation to obtain the highway traffic operation status evaluation grade includes: The evaluation correlation between highway evaluation index and highway evaluation index grade is calculated based on the classic domain distance and node domain distance. The expression is: in The indicator value is Highway evaluation index Corresponding level The evaluation correlation of is the variance sensitivity, For indicators The time series variance of is the smoothing factor; According to the evaluation correlation of the highway evaluation index levels, a correlation matrix is ​​constructed with the indicators as rows and the levels as columns. The weight vector of the highway to be evaluated is determined according to the fuzzy consistency matrix. The weighted correlation of the same highway evaluation index at different levels is calculated by the weight vector and the correlation matrix. The level corresponding to the maximum weighted correlation of the highway evaluation index is taken as the evaluation level. The evaluation level of the highway traffic operation status is obtained by taking the average of the evaluation levels of the highway evaluation indicators of the same type. The highway traffic operation status evaluation level includes highway flow level, highway speed level and highway density level.

[0010] The second aspect is the data-driven highway traffic operation status evaluation system, which includes: Grading module: used to divide highway traffic operation data using a decision tree and calculate highway evaluation indicators, and to perform multi-level division on the highway evaluation indicators to obtain highway evaluation indicator grades; A weight module is configured to perform dynamic incremental learning on the highway evaluation index to obtain a first index weight, perform fuzzy adjustment on the first index weight to obtain a second index weight, and construct a fuzzy consistency matrix based on the second index weight; A correlation module is used to establish an extension matter-element according to the highway evaluation index level and the highway evaluation index, perform deep learning on the highway evaluation index to determine the classical domain and the nodal domain, calculate the classical domain distance and the nodal domain distance of the extension matter-element of the highway to be evaluated, and calculate the evaluation correlation between the highway evaluation index and the highway evaluation index level according to the classical domain distance and the nodal domain distance; Evaluation module: used to determine the weight vector of the road to be evaluated according to the fuzzy consistency matrix, and perform extension evaluation according to the weight vector and the evaluation correlation to obtain the evaluation grade of the road traffic operation status; Intelligent supervision module: for storing, viewing and managing the fuzzy consistent matrix, the evaluation correlation degree and the highway traffic operation state evaluation grade, and conducting traffic control according to the highway traffic operation state evaluation grade.

[0011] The beneficial effects of the present application are: The present application is a data-driven highway traffic operation state evaluation method and system, which has the following technical effects compared with the prior art: The present application can improve the data preprocessing capability and enhance the model adaptability in the precision evaluation of highway traffic operation state evaluation through the steps of grade division, dynamic incremental learning, fuzzy adjustment, deep learning and extension evaluation, thereby improving the efficiency and accuracy of highway traffic operation state evaluation, optimizing the highway traffic operation state evaluation technology, greatly saving resources, improving work efficiency, realizing the evaluation of highway traffic operation state, providing stronger technical support for highway traffic management and decision-making, having important practical value for promoting the intelligent and fine management of highway traffic system, and being suitable for different highway traffic operation state evaluation systems and the highway traffic operation state evaluation needs of different users, and having certain universality. BRIEF DESCRIPTION OF DRAWINGS

[0012] Figure 1 The present application is a data-driven highway traffic operation state evaluation method and system, which has the following technical effects compared with the prior art: DETAILED DESCRIPTION

[0013] The present application will be further described below through specific embodiments, and the illustrative embodiments of the present application and the description are used to explain the present application, but not as a limitation of the present application.

[0014] The present application is a data-driven highway traffic operation state evaluation method and system, which has the following technical effects compared with the prior art: As shown in the figure, in the present embodiment, the following steps are included: Figure 1 The decision tree is used to divide the highway traffic operation data and calculate the highway evaluation index, and the highway evaluation index is divided into multiple levels to obtain the highway evaluation index grade; the highway evaluation index includes the flow evaluation index, the speed evaluation index and the density evaluation index; The dynamic incremental learning is performed on the highway evaluation index to obtain the first index weight, the fuzzy adjustment is performed on the first index weight to obtain the second index weight, and the fuzzy consistent matrix is constructed according to the second index weight; The dynamic incremental learning is performed on the highway evaluation index to obtain the first index weight, the fuzzy adjustment is performed on the first index weight to obtain the second index weight, and the fuzzy consistent matrix is constructed according to the second index weight; According to the highway evaluation index grade and the highway evaluation index, an extension matter element is established, a depth learning of the highway evaluation index is performed to determine a classic field and a node field, a classic field distance and a node field distance of the extension matter element of the to-be-evaluated highway are calculated, and an evaluation correlation degree of the highway evaluation index and the highway evaluation index grade is calculated according to the classic field distance and the node field distance. A weight vector of the to-be-evaluated highway is determined according to the fuzzy consistent matrix, and a highway traffic running state evaluation grade is obtained through extension evaluation according to the weight vector and the evaluation correlation degree.

[0015] In the embodiment, the method for obtaining the highway evaluation index grade comprises: The highway traffic running data is divided into a flow class, a speed class and a density class by using a decision tree, and a flow class evaluation index, a speed class evaluation index and a density class evaluation index are calculated according to the division result; the flow class evaluation index comprises a flow growth rate, a saturation degree and a saturation degree balance coefficient; the speed class evaluation index comprises a speed change rate, a free flow speed ratio and a travel time stability; and the density class evaluation index comprises a space occupancy rate and a space-time occupancy rate. The basic three-level threshold values of the highway evaluation indexes are determined according to the industry standard, the correlation coefficients between the indexes are calculated respectively, two highway evaluation indexes with a correlation coefficient greater than 0.8 are taken as strong correlation indexes, and the rest of the indexes are taken as weak correlation indexes, time series analysis is performed on the historical data of the strong correlation indexes to obtain a time series fluctuation coefficient, the historical data of the strong correlation indexes are divided into four-level indexes according to the basic three-level threshold values, the data skewness and the data standard deviation of the same strong correlation index at different levels are calculated, the product of the data skewness mean and the data standard deviation mean of the same strong correlation index at different levels is taken as a data fluctuation coefficient, the basic three-level threshold values of the strong correlation indexes are adjusted to obtain optimized three-level threshold values according to the time series fluctuation coefficient and the data fluctuation coefficient, and the weak correlation indexes are divided according to the basic three-level threshold values, and the strong correlation indexes are divided according to the optimized three-level threshold values to obtain the highway evaluation index grade. In actual evaluation, when the flow class evaluation index is calculated, the flow growth rate comprises a same-period and a same-period growth rate of the flow in a road section / channel / road network, the saturation degree comprises a road section saturation degree = a cross-section traffic capacity / traffic capacity, a channel saturation degree = (∑road section saturation degree*road section length) / channel length, and a road network saturation degree = (∑road section saturation degree*road section length) / road network length, and the saturation degree balance coefficient comprises a channel saturation degree balance coefficient = a standard deviation of the saturation degrees of each road section of the channel / an average saturation degree of each road section of the channel, and a road network saturation degree balance coefficient = a standard deviation of the saturation degrees of each road section of the road network / an average saturation degree of each road section of the road network. When calculating speed evaluation indicators, the speed change rate (by vehicle type) includes the year-on-year and month-on-month change rates of average speeds for sections, corridors, and the road network. The free-flow speed ratio (by vehicle type) includes the section free-flow speed ratio = average vehicle section speed / section speed limit, corridor free-flow speed ratio = (∑ section free-flow speed ratio * section mileage) / total corridor mileage, and network free-flow speed ratio = (∑ section free-flow speed ratio * section mileage) / total network mileage. Travel time stability includes travel time stability in the temporal dimension (standard deviation of the multi-day average travel time for a section or corridor) and travel time stability in the spatial dimension (standard deviation of the average travel time for each section on a corridor). When calculating density evaluation indicators, spatial occupancy includes section spatial occupancy = ∑ vehicle length / section mileage, road network spatial occupancy = (∑ section spatial occupancy * section length) / total road network mileage, and spatiotemporal occupancy includes section spatiotemporal occupancy = (∑ vehicle length * vehicle on-road time during the study period) / (section mileage * study time), channel spatiotemporal occupancy = (∑ section spatiotemporal occupancy * section length) / total channel mileage, and network spatiotemporal occupancy = (∑ section spatiotemporal occupancy * section length) / total road network mileage. The autoregressive moving average model is used to perform time series analysis to obtain the time series fluctuation coefficient. The basic three-level thresholds of the strongly correlated indicators were scaled to obtain the optimized three-level thresholds, and the highway evaluation indicators were divided into four levels according to the optimized three-level thresholds.

[0016] In this embodiment, the method for obtaining the first indicator weight includes: The entropy weight method is used to process the historical highway evaluation index to obtain the initial weight. Constructing feature matrix from traffic data samples at each moment , is the feature matrix set, is the sample size, is a data dimension; the data dimension includes vehicle type, highway evaluation index, spatial location and timestamp; Calculate the current data sample Reference distribution with historical samples The Wasserstein distance is expressed as: in For time The Wasserstein distance at time, is a sample of historical reference data, is the set of all possible joint distributions, For current data The feature mapping function, is the joint distribution Lower feature map and The expectation of the square of the Euclidean distance, is the spatial feature weight matrix, is the time weight matrix, To convert the spatial coordinates Encoded as string, is the timestamp; Drift determination: When The weight update is triggered when is a dynamic threshold, taking the 90% quantile of the historical difference, is the noise tolerance; The first weight is obtained by updating the initial weight through a dual-channel learning mechanism; the dual-channel learning mechanism includes short-term incremental adjustment and long-term incremental adjustment, and the expression is: in Highway evaluation index exist The first indicator weight is updated at all times, Highway evaluation index The initial weight of is the short-term weight, for Time short-term adjustment amount, is the long-term weight, Highway evaluation index exist Long-term adjustment of time, is the learning rate, is a symbolic function, When taking 1, When taking -1, When 0, for Moment-wise cross entropy loss function The gradient vector with respect to the weights, is the smoothing factor, Highway evaluation index and status The time-varying mutual information of is the highway evaluation index at the previous moment The weight of the first indicator; In actual evaluation, the cross entropy loss function The expression is: in is the number of rating levels, for The real state label at the moment, Is the model prediction function, used to output The probability of the level, is the indicator weight vector; Evaluation indicators and status Time-varying mutual information The expression is: in is a set of status labels, i.e., indicator levels 1-4, is a set of highway evaluation indicators, for The joint probability distribution of time, is the marginal distribution of the indicator, is the marginal distribution of states; Take short-term weight 0.5, long-term weight The learning rate is 0.3. is 0.01, smoothing factor The first indicator weight is calculated as 0.9.

[0017] In this embodiment, the method for constructing a fuzzy consensus matrix includes: The kernel function between the first indicator weights is calculated according to the radial basis-Gaussian mixed fuzzy kernel algorithm. The second indicator weight is obtained by adjusting the first indicator weight through kernel function mapping. The expression is: in Highway evaluation index exist The second indicator weight is updated at all times. To adjust the intensity, is the number of highway evaluation indicators, Highway evaluation index and The radial basis-Gaussian mixture fuzzy kernel function, is the spatiotemporal regulatory factor, is the weighted similarity sensitivity coefficient, is the number of Gaussian components, is the component weight, is a multivariate Gaussian distribution, is the mean of the Gaussian components, is the covariance matrix; According to the second index weight, a fuzzy consistent matrix is constructed , and the expression is: wherein is the element of the fuzzy consistent matrix , , is the maximum of the two second index weights in the second index weight; In actual evaluation, the number of highway evaluation indexes is 28, the adjustment intensity is 0.1, the space-time adjustment factor is 0.5, the weight similarity sensitivity coefficient is 0.1, and the Gaussian component number is 5. The kernel function between the first index weights is calculated, and the first index weights are adjusted to obtain the second index weights through kernel function mapping.

[0018] In the embodiment, the method for calculating the classical domain distance and the node domain distance of the extension matter element of the road to be evaluated comprises: According to the highway evaluation index level and the highway evaluation index definition, the extension matter element is defined wherein is the highway evaluation index level set, is the highway evaluation index set, is the evaluation set of ; A boundary generation network is constructed, and the multi-source traffic data tensor is input into the boundary generation network to obtain the classical domain interval and the node domain interval ; the multi-source traffic data tensor comprises a highway evaluation index, a time step, and a number of road sections; The boundary generation network comprises an encoder and a decoder, the encoder receives the multi-source traffic data tensor, extracts time sequence features by using a bidirectional LSTM to obtain a fixed-length context vector, and the decoder processes the context vector according to a domain generation signal to output the classical domain and the node domain corresponding to the level of the domain generation signal; the boundary generation network is trained by using a double-layer loss function; the double-layer loss function comprises a classical domain boundary loss function and a node domain boundary loss function, and the expression is: wherein is the double-layer loss function, is a highway evaluation index sample set with a level of , is the classical domain boundary loss function,​​​ is a loss function of the domain boundary, is a loss weight of the domain, is a level is a lower bound of the classical domain, is a level is an upper bound of the classical domain, is an interval width penalty coefficient, is a lower bound of the domain, is an upper bound of the domain; The classical domain interval of the output is used for adaptive weighted distance calculation to obtain the classical domain distance of the road extension matter element to be evaluated, and the domain interval of the output is dynamically compensated to obtain the domain distance of the road extension matter element to be evaluated, and the expression is: wherein , are the classical domain distance and the domain distance of the road evaluation index with the index value , is an interval width weight, is an attention sensitivity coefficient, is a clustering center of the level , is a drift compensation coefficient, is a time window length, is the value of the road evaluation index at the moment; In actual evaluation, the level set of the road evaluation index respectively represents unobstructed, mild congestion, severe congestion and paralysis, and the evaluation set takes “ ” as an example to represent the value interval of the road evaluation index of the 3rd level; In the boundary generation network, 28-dimensional road evaluation indexes of 120 road sections of a certain highway within 1440 minutes are input, the domain loss weight is taken as 0.8, the interval width penalty coefficient is taken as 0.3, and the classical domain and the domain of each road evaluation index are output; The interval width weight is taken as 1.2, the attention sensitivity coefficient is taken as 0.8, the drift compensation coefficient is taken as 0.15, and the time window length is taken as 24h, and the classical domain distance and the domain distance of the road extension matter element to be evaluated are calculated.

[0019] In this embodiment, the method for obtaining the road traffic running state evaluation level by performing extension evaluation comprises: The evaluation correlation between highway evaluation index and highway evaluation index grade is calculated based on the classic domain distance and node domain distance. The expression is: in The indicator value is Highway evaluation index Corresponding level The evaluation correlation of is the variance sensitivity, For indicators The time series variance of is the smoothing factor; According to the evaluation correlation of the highway evaluation index grades, a correlation matrix is ​​constructed with the index as the row and the grade as the column; the weight vector of the highway to be evaluated is determined according to the fuzzy consistency matrix; the weighted correlation of the same highway evaluation index at different grades is calculated by the weight vector and the correlation matrix; the grade corresponding to the maximum weighted correlation of the highway evaluation index is taken as the evaluation grade; the evaluation grade of the highway traffic operation status is obtained by taking the average of the evaluation grades of the highway evaluation indicators of the same type; the highway traffic operation status evaluation grade includes the highway flow grade, the highway speed grade and the highway density grade; In actual evaluation, the variance sensitivity is taken is 0.15, smoothing factor is 0.1, and the evaluation correlation between highway evaluation index and highway evaluation index grade is calculated to construct a 28×4 correlation matrix; Taking the weighted correlation calculation of road section saturation as an example, the four-level weight vector of road section saturation is determined to be [0.35, 0.35, 0.4, 0.6] through the fuzzy consistency matrix. The correlation between road section saturation and the evaluation of each level is determined to be [0.85, 0.72, 0.65, 0.45] according to the correlation matrix. The weighted correlations of different levels of road section saturation are calculated to be 0.2975, 0.252, 0.26, and 0.27. The level (level 1) corresponding to the maximum weighted correlation of 0.2975 for road section saturation is taken as the evaluation level. The average evaluation grades of the flow, speed and density highway evaluation indicators were calculated respectively. The average grade of 11 flow indicators was 1.81, the average grade of 11 speed indicators was 1.18, and the average grade of 6 speed indicators was 2. That is, the highway traffic flow grade was 2 (stable flow, slight impact), the highway traffic speed grade was 1 (fast speed, smooth traffic), and the highway traffic density grade was 2 (moderate density, slight impact on traffic).

[0020] The second aspect is the data-driven highway traffic operation status evaluation system, which includes: Grading module: used to divide highway traffic operation data using a decision tree and calculate highway evaluation indicators, and to perform multi-level division on the highway evaluation indicators to obtain highway evaluation indicator grades; A weight module is configured to perform dynamic incremental learning on the highway evaluation index to obtain a first index weight, perform fuzzy adjustment on the first index weight to obtain a second index weight, and construct a fuzzy consistency matrix based on the second index weight; A correlation module is used to establish an extension matter-element according to the highway evaluation index level and the highway evaluation index, perform deep learning on the highway evaluation index to determine the classical domain and the nodal domain, calculate the classical domain distance and the nodal domain distance of the extension matter-element of the highway to be evaluated, and calculate the evaluation correlation between the highway evaluation index and the highway evaluation index level according to the classical domain distance and the nodal domain distance; Evaluation module: used to determine the weight vector of the road to be evaluated according to the fuzzy consistency matrix, and perform extension evaluation according to the weight vector and the evaluation correlation to obtain the evaluation grade of the road traffic operation status; Intelligent supervision module: used to store, view and manage the fuzzy consistency matrix, the evaluation correlation and the highway traffic operation status evaluation level, and perform traffic control according to the highway traffic operation status evaluation level.

Claims

1. A data-driven highway traffic operation status evaluation method, characterized by: The following steps are involved: S1. Using a decision tree to divide highway traffic operation data and calculate highway evaluation indicators, the highway evaluation indicators are divided into multiple levels to obtain highway evaluation indicator grades; the highway evaluation indicators include flow evaluation indicators, speed evaluation indicators, and density evaluation indicators; S2. Performing dynamic incremental learning on the highway evaluation index to obtain a first index weight, performing fuzzy adjustment on the first index weight to obtain a second index weight, and constructing a fuzzy consistency matrix based on the second index weight; S3. Establishing an extension matter-element based on the highway evaluation index level and the highway evaluation index, performing deep learning on the highway evaluation index to determine a classical domain and a nodal domain, calculating a classical domain distance and a nodal domain distance of the extension matter-element of the highway to be evaluated, and calculating an evaluation correlation between the highway evaluation index and the highway evaluation index level based on the classical domain distance and the nodal domain distance; S4. Determine a weight vector of the road to be evaluated according to the fuzzy consistent matrix, and perform extension evaluation according to the weight vector and the evaluation correlation to obtain an evaluation grade of the road traffic operation status.

2. The data-driven highway traffic operation status evaluation method according to claim 1 is characterized in that: The method for obtaining the highway evaluation index grade includes: A decision tree is used to classify highway traffic operation data into flow, speed, and density categories, and flow evaluation indicators, speed evaluation indicators, and density evaluation indicators are calculated based on the classification results; the flow evaluation indicators include flow growth rate, saturation, and saturation equilibrium coefficient; the speed evaluation indicators include speed change rate, free flow speed ratio, and travel time stability; and the density evaluation indicators include spatial occupancy and spatiotemporal occupancy. According to industry standards, the basic three-level thresholds of each highway evaluation indicator are determined, and the correlation coefficients between each indicator are calculated respectively. Two highway evaluation indicators with correlation coefficients greater than 0.8 are taken as strongly correlated indicators, and the remaining indicators are weakly correlated indicators. The historical data of the strongly correlated indicators are analyzed in time series to obtain the time series fluctuation coefficient. The historical data of the strongly correlated indicators are divided into four-level indicators according to the basic three-level thresholds. The data skewness and data standard deviation of the same strongly correlated indicator at different levels are calculated. The product of the mean data skewness and the mean data standard deviation of the same strongly correlated indicator at different levels is taken as the data fluctuation coefficient. The basic three-level thresholds of the strongly correlated indicators are adjusted according to the time series fluctuation coefficient and the data fluctuation coefficient to obtain the optimized three-level threshold. The weakly correlated indicators are divided according to the basic three-level thresholds, and the strongly correlated indicators are divided according to the optimized three-level thresholds to obtain the highway evaluation indicator grades.

3. The data-driven highway traffic operation status evaluation method according to claim 1 is characterized in that: The method for obtaining the first indicator weight includes: The entropy weight method is used to process the historical highway evaluation index to obtain the initial weight. Constructing feature matrix from traffic data samples at each moment , is the feature matrix set, is the sample size, is a data dimension; the data dimension includes vehicle type, highway evaluation index, spatial location and timestamp; Calculate the current data sample Reference distribution with historical samples The Wasserstein distance is expressed as: in For time The Wasserstein distance at time, is a sample of historical reference data, is the set of all possible joint distributions, For current data The feature mapping function, is the joint distribution Lower feature map and The expectation of the square of the Euclidean distance, is the spatial feature weight matrix, is the time weight matrix, To convert the spatial coordinates Encoded as string, is the timestamp; Drift determination: When The weight update is triggered when is a dynamic threshold, taking the 90% quantile of the historical difference, is the noise tolerance; The first weight is obtained by updating the initial weight through a dual-channel learning mechanism; the dual-channel learning mechanism includes short-term incremental adjustment and long-term incremental adjustment, and the expression is: in Highway evaluation index exist The first indicator weight is updated at all times, Highway evaluation index The initial weight of is the short-term weight, for Time short-term adjustment amount, is the long-term weight, Highway evaluation index exist Long-term adjustment of time, is the learning rate, is a symbolic function, When taking 1, When taking -1, When 0, for Moment-wise cross entropy loss function The gradient vector with respect to the weights, is the smoothing factor, Highway evaluation index and status The time-varying mutual information of is the highway evaluation index at the previous moment The first indicator weight.

4. The data-driven highway traffic operation status evaluation method according to claim 1 is characterized in that: The method for constructing a fuzzy consistency matrix comprises: The kernel function between the first indicator weights is calculated according to the radial basis-Gaussian mixed fuzzy kernel algorithm. The second indicator weight is obtained by adjusting the first indicator weight through kernel function mapping. The expression is: in Highway evaluation index exist The second indicator weight is updated at all times. To adjust the intensity, is the number of highway evaluation indicators, Highway evaluation index and The radial basis-Gaussian mixture fuzzy kernel function, is the spatiotemporal regulatory factor, is the weighted similarity sensitivity coefficient, is the number of Gaussian components, is the component weight, is a multivariate Gaussian distribution, is the mean of the Gaussian components, is the covariance matrix; Constructing fuzzy consistency matrix based on the second index weight , the expression is: in is the fuzzy consistent matrix element, 、 are the two largest second indicator weights among the second indicator weights.

5. The data-driven highway traffic operation status evaluation method according to claim 1 is characterized in that: The method for calculating the classical domain distance and the nodal domain distance of the extension matter-element of the highway to be evaluated includes: Define extension matter-element based on highway evaluation index level and highway evaluation index ,in is the set of highway evaluation index levels, is a set of highway evaluation indicators, for about 's evaluation set; Construct a boundary generation network and input multi-source traffic data tensors into the boundary generation network to obtain the classic domain interval and section interval ; The multi-source traffic data tensor includes highway evaluation indicators, time steps and number of road sections; The boundary generation network includes an encoder and a decoder. The encoder receives multi-source traffic data tensors and uses a bidirectional LSTM to extract temporal features to obtain a fixed-length context vector. The decoder processes the context vector according to the domain generation signal and outputs the classical domain and section domain of the corresponding level of the domain generation signal. The boundary generation network is obtained by training with a double-layer loss function. The double-layer loss function includes the classical domain boundary loss function and the section domain boundary loss function, which are expressed as: in is a double-layer loss function, For level The sample set of highway evaluation indicators is is the classic domain boundary loss function, is the section boundary loss function, is the section loss weight, For level The lower bound of the classical domain, For level The upper bound of the classical domain, is the interval width penalty coefficient, is the lower bound of the node domain, is the upper bound of the node domain; The classical domain distance of the extension matter-element of the highway to be evaluated is obtained by adaptively calculating the weighted distance according to the output classical domain interval, and the nodal domain distance of the extension matter-element of the highway to be evaluated is obtained by dynamically compensating the output nodal domain interval. The expression is: in 、 The index values ​​are Highway evaluation index The classical domain distance and node domain distance of is the interval width weight, is the attention sensitivity coefficient, For level The cluster center of is the drift compensation coefficient, is the time window length, for Moment highway evaluation index The value of .

6. The data-driven highway traffic operation status evaluation method according to claim 1 is characterized in that: The method for performing extension evaluation to obtain a highway traffic operation status evaluation grade comprises: The evaluation correlation between highway evaluation index and highway evaluation index grade is calculated based on the classic domain distance and node domain distance. The expression is: in The indicator value is Highway evaluation index Corresponding level The evaluation correlation of is the variance sensitivity, For indicators The time series variance of is the smoothing factor; According to the evaluation correlation of the highway evaluation index levels, a correlation matrix is ​​constructed with the indicators as rows and the levels as columns. The weight vector of the highway to be evaluated is determined according to the fuzzy consistency matrix. The weighted correlation of the same highway evaluation index at different levels is calculated by the weight vector and the correlation matrix. The level corresponding to the maximum weighted correlation of the highway evaluation index is taken as the evaluation level. The evaluation level of the highway traffic operation status is obtained by taking the average of the evaluation levels of the highway evaluation indicators of the same type. The highway traffic operation status evaluation level includes highway flow level, highway speed level and highway density level.

7. The data-driven highway traffic operation status evaluation system is characterized by: Used to perform the method according to any one of claims 1 to 6, comprising: Grading module: used to divide highway traffic operation data using a decision tree and calculate highway evaluation indicators, and to perform multi-level division on the highway evaluation indicators to obtain highway evaluation indicator grades; A weight module is configured to perform dynamic incremental learning on the highway evaluation index to obtain a first index weight, perform fuzzy adjustment on the first index weight to obtain a second index weight, and construct a fuzzy consistency matrix based on the second index weight; A correlation module is used to establish an extension matter-element according to the highway evaluation index level and the highway evaluation index, perform deep learning on the highway evaluation index to determine the classical domain and the nodal domain, calculate the classical domain distance and the nodal domain distance of the extension matter-element of the highway to be evaluated, and calculate the evaluation correlation between the highway evaluation index and the highway evaluation index level according to the classical domain distance and the nodal domain distance; Evaluation module: used to determine the weight vector of the road to be evaluated according to the fuzzy consistency matrix, and perform extension evaluation according to the weight vector and the evaluation correlation to obtain the evaluation grade of the road traffic operation status; Intelligent supervision module: used to store, view and manage the fuzzy consistency matrix, the evaluation correlation and the highway traffic operation status evaluation level, and perform traffic control according to the highway traffic operation status evaluation level.

Images