CN118015857B - Road traffic planning method - Google Patents

Abstract

The invention relates to the technical field of road traffic control, in particular to a road traffic planning method, which comprises the following steps: acquiring a historical traffic flow curve of each road in each day, and further determining the traffic flow retention degree in the same period; determining the association degree of the traffic flow of any two roads according to the difference condition between the historical traffic flow curves of any two roads in each day and the traffic flow maintenance degree of any two roads in the same period; determining clustering optimization weights of any two intersections according to the traffic flow association degree, further clustering all intersections in the urban area to be planned, obtaining an optimal intersection clustering result, and further regulating and controlling intersection traffic lights; according to the invention, the clustering optimization weight is determined by quantifying the traffic flow association degree of the road, and the conventional intersection clustering result is optimized, so that the reliability of intersection traffic light regulation and control is improved, and the rationality of road traffic planning is further improved.

Description

Road traffic planning method

Technical Field

The invention relates to the technical field of road traffic control, in particular to a road traffic planning method.

Background

With the development of society, the number of motor vehicles in cities is in a trend of continuously rising, the traffic load of the cities is correspondingly increasing, and road traffic needs to be reasonably planned based on different traffic problems. For example, an unreasonable time setting of the traffic lights at an intersection or an inadaptation to a change in traffic flow for different periods of time will result in an increase in intersection cross saturation, and when there are too many vehicles waiting for the red road lights, the queuing length increases, which affects the traffic capacity of adjacent intersections and road segments, resulting in a wider traffic jam. In order to solve the traffic problem, k-means clustering is carried out on all intersections in the city, intersection relations with close relations are obtained, and intersection pre-regulation and control are achieved based on intersection clustering results.

The existing intersection clustering only considers the distance or road grade among the intersections, but in the actual scene, the situation that the traffic flow among the intersections is changed is more considered in the aspect of realizing intersection pre-regulation, so that the accuracy of the intersection clustering result in the existing traffic planning process is low, and the rationality of traffic planning is poor, namely the road traffic control capability is low.

Disclosure of Invention

In order to solve the technical problem of poor traffic planning rationality caused by low accuracy of the intersection clustering result, the invention aims to provide a road traffic planning method, and the adopted technical scheme is as follows:

One embodiment of the present invention provides a road traffic planning method, which includes the steps of:

Acquiring the historical vehicle flow of each road in the urban area to be planned in each unit time within a plurality of days, and further determining the historical vehicle flow curve of each road in each day; the road section between two adjacent intersections forms a road, and the historical traffic flow is the historical traffic flow in the same driving direction on the road;

Dividing a day into time periods with the same length, and determining the traffic flow retention degree of each road in the same time period according to each extreme point of a curve section of the historical traffic flow curve of each road in the same time period and the historical traffic flow average value of the curve section;

Determining the association degree of the traffic flow of any two roads according to the difference condition between the historical traffic flow curves of any two roads in each day and the traffic flow maintenance degree of any two roads in the same period;

Determining clustering optimization weights of any two intersections according to the association degree of the traffic flow of any two roads; clustering all intersections of the urban area to be planned by combining the clustering optimization weights of any two intersections to obtain an optimal intersection clustering result;

and regulating and controlling the intersection traffic lights in the urban area to be planned according to the optimal intersection clustering result.

Further, the determining the traffic flow retention of each road in the same period according to each extreme point of the curve segment of the historical traffic flow curve of each road in the same period and the historical traffic flow average value of the curve segment, includes:

for any road and any time period, taking a curve section of a historical traffic flow curve of the road in each day under the time period as a target curve section to obtain each corresponding target curve section of the road under the time period;

Determining each extreme point of each target curve segment, further determining the time span between each extreme point and the next extreme point adjacent to each extreme point, and obtaining the time span corresponding to each extreme point of each target curve segment;

Determining the traffic flow maintaining weight of each target curve segment corresponding to the road in the time period according to each extreme point of each target curve segment and the corresponding time span thereof;

and determining the traffic flow retention of the road in the time period according to the historical traffic flow average value and the traffic flow retention weight of each corresponding target curve segment of the road in the time period.

Further, the calculation formula of the traffic flow maintaining weight of the mth target curve segment corresponding to the road pq in the nth period is as follows:

; in the/> Maintaining weights for the traffic flow of the mth target curve segment corresponding to the road pq in the nth time period, wherein p and q are two adjacent intersections in the urban area to be planned, and the intersection p and the intersection q form the road pq,/>For the normalization function, Z is the extreme point sequence number of the mth target curve segment corresponding to the road pq in the nth period, and Z is the number of extreme points of the mth target curve segment corresponding to the road pq in the nth period,/>For the time span corresponding to the z extreme point of the corresponding mth target curve segment of the road pq under the nth time period,/>For the time span mean value corresponding to all extreme points of the mth target curve segment corresponding to the road pq in the nth period of time,/>The difference between the historical traffic flow of the z extreme point of the corresponding mth target curve segment of the road pq under the nth time period and the historical traffic flow average of the mth target curve segment is obtained.

Further, the determining the traffic flow retention of the road in the period according to the historical traffic flow average and the traffic flow retention weight of each corresponding target curve segment of the road in the period comprises:

for any one target curve segment, firstly calculating the difference between the average value of the historical traffic flow of the target curve segment and the average value of the historical traffic flow of all the target curve segments corresponding to the road in the period, and recording the difference as a target difference; calculating the product of the target difference and the traffic flow maintaining weight of the target curve segment, and recording the product as a first product;

and obtaining a first product of each corresponding target curve segment of the road in the period, and taking an average value of the first products of each corresponding target curve segments of the road in the period as the traffic flow maintenance degree of the road in the period.

Further, the determining the association degree of the traffic flow of any two roads according to the difference condition between the historical traffic flow curves of any two roads in each day and the traffic flow maintenance degree of any two roads in the same period comprises the following steps:

Performing DTW (draw-off line) matching on historical traffic flow curves of any two roads in the same day to obtain matched data points, and further determining traffic flow retention of a target curve segment of data acquisition time of two data points in each pair of data points under the period of time;

Determining the actual distance between the matched pairs of data points corresponding to any two roads according to the historical traffic flow of the matched pairs of data points and the traffic flow retention of the target curve segment of two data points in the pairs of data points under the period of time;

Determining an initial road and a final road in any two roads, acquiring each split road corresponding to the initial road, and further determining the actual distance between each pair of data points after the final road and each split road corresponding to the initial road are matched;

determining the association degree of the traffic flow of any two roads on the same day according to the actual distance of each pair of data points after the matching corresponding to any two roads, the actual distance of each pair of data points after the matching corresponding to each shunt road corresponding to the starting road and the historical traffic flow difference of each pair of data points;

And taking the average value of the vehicle flow correlation degrees of any two roads at each day as the vehicle flow correlation degree of any two roads.

Further, a first road in a route formed by any two roads along the same driving direction is a starting road, and a first intersection of the starting road is a starting intersection; the last road in the route formed by any two roads along the same driving direction is a termination road, and the first intersection of the termination road is a termination intersection; each road adjacent to a route formed by any two roads along the same driving direction is used as each split road corresponding to the starting road.

Further, the calculation formula of the actual distance of the h-th data point after matching corresponding to the road pq and the road wv is as follows:

; in the method, in the process of the invention, For the actual distance between the h-th pair of data points after the matching of the road pq and the road wv, w and v are two adjacent intersections in the urban area to be planned, and the intersection w and the intersection v form the road wv,/>For the history traffic flow difference of the h-th data point after the matching corresponding to the road pq and the road wv,/>As a normalization function,/>For the average value of the vehicle flow retention of the target curve segment under the belonged period of the data acquisition time of two data points in the h data points after the matching of the road pq and the road wv,/>For the difference of the vehicle flow retention degree of the target curve segment under the belonged period of the data acquisition time of two data points in the h data points after the matching of the road pq and the road wv, the method comprises the following steps of/>And the historical vehicle flow average value of the h-th data point after the matching of the road pq and the road wv is obtained.

Further, the calculation formula of the association degree of the traffic flow of the road pq and the road wv on the same day is as follows:

; in the method, in the process of the invention, For the association degree of the traffic flow of the road pq and the road wv on the same day,/>For the actual distance between the H-th pair of data points after matching corresponding to the road pq and the road wv, H is the serial number of each pair of data points after matching corresponding to the two roads, and H is the pair of data points after matching corresponding to the two roads,/>For the normalization function, I is the serial number of each split road corresponding to the initial road, I is the number of all split roads corresponding to the initial road,/>To terminate the actual distance of the h-th pair of data points after matching corresponding to the i-th split road corresponding to the initial road and the difference between the actual distance of the h-th pair of data points after matching corresponding to the road pq and the road wv,/>And the historical vehicle flow difference of the h-th data point after matching corresponding to the ith split road corresponding to the ending road and the starting road is obtained.

Further, determining the cluster optimization weight of any two intersections according to the association degree of the traffic flows of any two roads comprises:

Acquiring all routes formed by any two intersections along the driving direction;

for any one route of any two intersections, calculating the accumulated sum of the association degrees of the traffic flow of all the roads except the ending road passing through in the route and the ending road, and recording the accumulated sum as the initial association degree of the traffic flow of the route; calculating the reciprocal of the number of roads passing through the route, and taking the product of the initial vehicle flow association degree of the route and the reciprocal of the number of roads passing through the route as the first vehicle flow association degree of the route;

taking the average value of the first vehicle flow correlation degree of each route of any two intersections as the second vehicle flow correlation degree of all routes formed by the two intersections along the driving direction; calculating the average value of the second vehicle flow correlation degrees of all routes formed by the two intersections along different driving directions, carrying out normalization processing on the average value of the second vehicle flow correlation degrees corresponding to the different driving directions, and taking the value after normalization processing as the clustering optimization weight of the two intersections.

Further, the clustering optimization weight of any two intersections is combined, all intersections in the urban area to be planned are clustered, and an optimal intersection clustering result is obtained, including:

Determining the initial distance measurement characteristics of any two intersections and the inverse proportion value of the clustering optimization weights of any two intersections, and taking the product of the initial distance measurement characteristics of any two intersections and the inverse proportion value of the clustering optimization weights of the corresponding two intersections as the optimized distance measurement characteristics of the corresponding two intersections;

and carrying out k-means clustering on all the intersections in the urban area to be planned based on the optimized distance measurement characteristics of any two intersections to obtain a clustering result, and taking the clustering result as an optimal intersection clustering result.

The invention has the following beneficial effects:

The conventional relationship between the intersections only considers the distance or road grade between the intersections, and the influence of the variation trend difference of the traffic flow between the intersections on the intersection clustering result is not considered, so that the accuracy of the intersection clustering result is low. The invention provides a road traffic planning method, which further improves the rationality of traffic planning through more reliable traffic light regulation and control; specifically, based on a historical traffic flow curve of a road formed by adjacent intersections for a plurality of days, quantifying the traffic flow retention of the road in the same period, and taking the traffic flow retention as a regulation reference, the method can accurately reflect the traffic flow change stability condition of the road in a local period, and is mainly used for subsequently determining the traffic flow association degree; in order to further improve the accuracy of the vehicle flow correlation, not only the stable condition of the vehicle flow change in the local period is analyzed, but also the difference condition of the vehicle flow change in the whole day is analyzed, so that the vehicle flow correlation of any two roads is determined, and the calculation factors in multiple aspects are considered when the vehicle flow correlation is calculated, so that the calculation accuracy of the vehicle flow correlation can be effectively improved; the clustering optimization weight of the two intersections based on the vehicle flow association degree can effectively improve the accuracy of clustering results of all intersections; and determining the optimal intersection clustering result is beneficial to improving the reliability of intersection traffic light regulation and control in the urban area to be planned, and is beneficial to avoiding special situations of wide traffic jams.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of a road traffic planning method according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a driving route pqwv according to an embodiment of the present invention;

fig. 3 is a schematic view of each split road of the driving route pqwv according to an embodiment of the present invention.

Detailed Description

In order to further describe the technical means and effects adopted by the present invention to achieve the preset purpose, the following detailed description is given below of the specific implementation, structure, features and effects of the technical solution according to the present invention with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

The application scene aimed by the invention is as follows:

The conventional intersection clustering only considers the road grade or the traffic flow difference under a single moment, and the like, ignores the similarity degree of traffic flow trend between intersections with the real traffic light reference value, causes low accuracy of intersection clustering results, further causes poor traffic light regulation and control value and poor traffic planning rationality.

According to the method, the association characteristics among the collected historical traffic flows are analyzed, the traffic flow change relation among the intersections is obtained, and the intersection clustering result is optimized based on the traffic flow change relation. Specifically, a road traffic planning method is provided, as shown in fig. 1, including the following steps:

S1, acquiring the historical traffic flow of each road in the urban area to be planned in each unit time in a plurality of days, and further determining the historical traffic flow curve of each road in each day.

In the embodiment, firstly, acquiring historical traffic flow of each road of a city area to be planned of the city area to be planned in each unit time in a plurality of days through GPS (Global Positioning System ) equipment; in order to facilitate analysis of the trend of the historical traffic flow data every day, curve fitting is performed on a plurality of historical traffic flows every day by using least square fitting, and a historical traffic flow curve of each road in every day can be obtained. The road section between two adjacent intersections forms a road, the number of the collected historical traffic flow is 15 days before the day, namely, the number of days of a plurality of days is 15, and the unit time of a single historical traffic flow is set to be 5 minutes; the horizontal axis of the historical traffic flow curve is data acquisition time, the vertical axis is the historical traffic flow of each data acquisition time, and the time span between adjacent data acquisition times is 1 and 5 minutes.

It should be noted that, the closer the time of collection of the historical traffic flow is to the day, the greater the reference value of the collected historical traffic flow data, so the historical traffic flow data is analyzed to be the historical traffic flow data of the first 15 days of the day in this embodiment; the number of days for collecting the historical traffic flow and the unit time of the single historical traffic flow can be set by an implementer according to specific practical conditions, and are not particularly limited herein; the least square fitting can remove the fluctuation to a certain extent, and the implementation process is the prior art and is not in the scope of the invention, and detailed description is omitted here.

S2, dividing a day into time periods with the same length, and determining the traffic flow retention degree of each road in the same time period according to each extreme point of the curve section of the historical traffic flow curve of each road in the same time period and the historical traffic flow average value of the curve section.

When the traffic light time of the intersection is regulated, the traffic flow condition of a certain road can be known in advance by the traffic flow conditions of the first few roads in the passing direction of the intersection based on the communication relationship between the roads in the urban area to be planned. Therefore, the traffic flow association factor among the intersections is added to perform intersection clustering, so that the reliability of the subsequent intersection traffic light time regulation is improved.

Analyzing the change rule of the historical traffic flow curve of each road, and regarding the road with the regularity of the historical traffic flow, the reference value is high when the traffic flow of the traffic light at the regulated crossing is early-warned, wherein the regularity refers to the fact that the traffic flow has early and late peak periods, for example, before the traffic flow of a certain road is suddenly increased, the related road which can clearly cause the traffic flow of the road to be suddenly increased can refer to the road formed by the adjacent crossing in the same cluster; for roads with historical traffic flow in the early and late peak periods, even if the distance between the road and the road to be analyzed is relatively short, the reference value of the road to be analyzed is relatively weak when the road is used for adjusting the traffic flow early warning of the road to be analyzed.

Based on the analysis of the change rule of the historical traffic flow curve of the road, it is known that the traffic flow retention of the road under a fixed period needs to be quantified, so that the traffic flow association degree of two roads is determined based on the traffic flow retention, the traffic flow retention can represent the historical traffic flow change stability degree of the road every day under the same period, and the greater the historical traffic flow change stability degree, the stronger the traffic flow rule degree of the corresponding road is indicated.

In this embodiment, for a certain road, in order to analyze the traffic flow condition of a historical traffic flow curve of each day in the same period, a day is first divided into each period with a length of 1 hour, and 24 periods are obtained; for convenience of description, taking the calculation process of the traffic flow maintenance degree of each road under the same period as an example of determining the traffic flow maintenance degree of the road pq under the nth period, the specific implementation process may include:

And in the first step, taking a curve section of the historical traffic flow curve of the road pq in each day under the nth period as a target curve section, and obtaining each corresponding target curve section of the road pq under the nth period.

In the present embodiment, the road pq is formed by a road section between the adjacent intersection p and the intersection q, and the road pq may refer to a vehicle traveling from the intersection p to the intersection q; uniformly marking a curve section under an nth period in a 15-day historical traffic flow curve corresponding to the road pq as a target curve section, which is equivalent to uniformly dividing the 15-day historical traffic flow curve into curve sections with the length of 1 hour, and selecting the curve sections with the same serial numbers as the target curve sections; the number of target curve segments in the same period is consistent with the number of historical traffic flow curves.

And a second step of determining each extreme point of each target curve segment, further determining the time span between each extreme point and the next extreme point adjacent to each extreme point, and obtaining the time span corresponding to each extreme point of each target curve segment.

In this embodiment, in order to analyze the change situation of the historical traffic flow data of each target curve segment, each extremum point in each target curve segment is calculated by using a derivative method, and the process of determining the extremum point by using the derivative method is the prior art, and will not be described here again; for the last extreme point of the target curve segment, the next extreme point adjacent to the extreme point is the first extreme point of the next target curve segment of the target curve segment on the same historical traffic flow curve, and if the target curve segment is the last curve segment on the historical traffic flow curve, the last extreme point of the target curve segment is not analyzed; the time span is the difference between the data acquisition time between two adjacent extreme points, and the larger the time span is, the larger the difference of the horizontal coordinates between the two extreme points is.

And thirdly, determining the traffic flow maintaining weight of each target curve segment corresponding to the road pq in the nth period according to each extreme point of each target curve segment and the corresponding time span thereof.

In this embodiment, the traffic flow maintaining weight may be used to measure the traffic flow stability of a plurality of historical traffic flow curves corresponding to the road in the same period, where the more uniform the extreme point distribution is, the weaker the fluctuation of the curve segment is, and the higher the stability of the traffic flow in the period is.

As an example, the calculation formula of the traffic flow maintaining weight of the mth target curve segment corresponding to the road pq at the nth period may be:

; in the/> Maintaining weights for the traffic flow of the mth target curve segment corresponding to the road pq in the nth time period, wherein p and q are two adjacent intersections in the urban area to be planned, and the intersection p and the intersection q form the road pq,/>For the normalization function, Z is the extreme point sequence number of the mth target curve segment corresponding to the road pq in the nth period, and Z is the number of extreme points of the mth target curve segment corresponding to the road pq in the nth period,/>For the time span corresponding to the z extreme point of the corresponding mth target curve segment of the road pq under the nth time period,/>For the time span mean value corresponding to all extreme points of the mth target curve segment corresponding to the road pq in the nth period of time,/>The difference between the historical traffic flow of the z extreme point of the corresponding mth target curve segment of the road pq under the nth time period and the historical traffic flow average of the mth target curve segment is obtained.

In the calculation formula of the traffic flow maintaining weight,Can be used for characterizing the time span difference of the z extreme point of the corresponding mth target curve segment of the road pq under the nth time period,/>The difference between the z extreme point of the mth target curve segment corresponding to the road pq in the nth time period and the integral mean value of the mth target curve segment can be represented; the smaller the overall time span difference is, the more uniform the extreme point distribution of the mth target curve segment is, the weaker the fluctuation of the mth target curve segment is, and the more stable the traffic flow change of the mth target curve segment is; /(I)The larger the fluctuation amplitude of the z extreme point is, the higher the amplitude of the fluctuation is, the higher the weight should be given to the z extreme point, so/>For acting as/>Is a weight of (2).

It should be noted that, the difference between two data in this embodiment refers to an actual difference between two data, for example, an absolute value of a difference between two data.

And fourthly, determining the traffic flow retention degree of the road pq in the nth period according to the historical traffic flow mean value and the traffic flow retention weight of each corresponding target curve segment of the road pq in the nth period.

It should be noted that, the traffic flow retention mainly provides data for the correlation degree of the traffic flow calculated later, the traffic flow retention refers to the traffic flow retention condition of a certain road in a long time sequence historical traffic flow data in the same period every day, when the traffic flow difference of the road in a certain period is smaller, the better the traffic flow retention condition of the road is indicated, and the higher the reference value and the confidence degree are when the traffic light regulation of the relevant road is performed later based on the traffic flow of the road in the period.

In this embodiment, for any one target curve segment corresponding to the road pq in the nth period, firstly, calculating the difference between the average value of the historical traffic flow of the target curve segment and the average value of the historical traffic flow of all the target curve segments corresponding to the road pq in the nth period, and recording the difference as a target difference, so as to obtain the target difference of each corresponding target curve segment in the nth period; and further calculating the product of the target difference of each corresponding target curve segment and the traffic flow holding weight of the corresponding target curve segment in the nth period, marking the product as a first product, and taking the average value of the first products of each corresponding target curve segment in the nth period as the traffic flow holding degree of the road pq in the nth period.

As an example, the calculation formula of the traffic flow retention of the road pq at the nth period may be:

; in the/> For the vehicle flow retention of the road pq in the nth period, M is the number of corresponding target curve segments in the nth period, and M is the number of corresponding target curve segments in the nth period,/>For the historical traffic flow average of the mth target curve segment of the road pq under the nth period of time,/>For the average value of the historic traffic flow average values of all target curve segments of the road pq under the nth period of time,/>In order to find the absolute value function,For the target difference of the mth target curve segment corresponding to the road pq in the nth period of time,/>Maintaining weights for the traffic flow of the mth target curve segment corresponding to the road pq in the nth period of time,/>Is the first product of the corresponding mth target curve segment of the road pq under the nth period.

In the calculation formula of the vehicle flow retention,The difference between the historical traffic flow average value of the mth target curve segment and the average value of the historical traffic flow average values of all target curve segments in the nth period of time in multiple days can be represented, and the smaller the traffic flow average value difference is, the stronger the traffic flow change stability of the road pq in the nth period of time is, the better the traffic flow retention is, and the higher the reference value of traffic flow data in the subsequent nth period of time is when traffic light regulation is carried out; /(I)It can be characterized that the possibility of the road pq having a peak period under the nth period and the road pq/>, of the nth period of a plurality of daysThe larger the road pq is, the more likely the road pq has a peak period in the nth period, and the higher the weight of the mth target curve segment in the nth period is when the mth target curve segment participates in the calculation of the traffic flow retention in the nth period; the vehicle flow keeps the weight in the range of 0 to 1.

And S3, determining the association degree of the traffic flow of any two roads according to the difference condition between the historical traffic flow curves of any two roads in each day and the traffic flow maintenance degree of any two roads in the same period.

It should be noted that, the traffic flow retention mainly characterizes the traffic flow retention of the road itself in a certain period, but for a plurality of roads, the traffic flow retention of the road in a certain period does not fully characterize the correlation between the roads, and the road actually having the correlation usually has the correlation in the road of the whole day, for example, the road having the early peak and the road generally having the late peak. Therefore, it is necessary to quantify the traffic flow correlation between roads, that is, determine the traffic flow correlation of any two roads, based on the difference between the historical traffic flow curves of the whole of each day, in combination with the traffic flow retention of the roads in the same period.

In this embodiment, the calculation process of the association degree of the traffic flow of any two roads is consistent, and for convenience of description, the analysis of the association degree of the traffic flow of the road pq and the road wv may include:

firstly, performing DTW matching on a historical traffic flow curve of a road pq and a historical traffic flow curve of a road wv in the same day to obtain matched pairs of data points, and further determining the traffic flow retention degree of a target curve segment of the data acquisition time of two data points in the pairs of data points in a period of time.

The road pq and the road wv can form a driving route along the same driving direction, the road pq can represent that the intersection p drives to the intersection q, and the road wv can represent that the intersection w drives to the intersection v; if the intersection p is the initial intersection, the road pq is the initial road, the intersection v is the ending intersection, the road wv is the ending road, and the driving route can be from the intersection p to the intersection v, i.e. from the intersection p, through the intersection q and the intersection w, to the intersection v, and stop, and the driving route pqwv is shown in fig. 2.

In the first substep, DTW matching is performed on the historical traffic flow curves of the road pq and the road wv in the same day.

In this embodiment, for a certain day of several days, DTW (DYNAMIC TIME WARPING ) matching is performed on the historical traffic flow curve of the road pq and the historical traffic flow curve of the road wv of the day, so that each pair of data points can be obtained, each data point on the historical traffic flow curve of the road pq can determine the data point matched with each data point on the historical traffic flow curve of the road wv, and the data point and the corresponding matched data point can form a pair of data points; the implementation process of DTW matching is the prior art, and is not in the scope of the present invention, and will not be described in detail here.

And a second substep, determining the traffic flow retention of the target curve segment of the matched data acquisition time of two data points in each pair of data points under the period.

The data acquisition time of two data points in the historical traffic flow curve of the data points is determined, and then the time period of the data point acquisition time of the two data points is determined, so that the traffic flow retention of the target curve segment of the data acquisition time of the two data points under the time period is obtained, namely the traffic flow retention of the two data points is obtained.

For example, when the data acquisition time of one data point in a pair of data points is 1:19, the data acquisition time of the data point belongs to the 2 nd time period in one day, the 2 nd time period is 1:00 to 2:00, and the traffic flow maintenance of the target curve segment of the data acquisition time of the data point under the 2 nd time period can be obtained, namely the traffic flow maintenance of the data point is obtained.

And secondly, determining the actual distance between each pair of data points after matching corresponding to the road pq and the road wv according to the historical traffic flow of each pair of data points after matching and the traffic flow retention degree of the target curve segment of two data points in each pair of data points under the period of time.

In this embodiment, by analyzing the historical traffic flow of two data points in a pair of data points and the traffic flow retention of a target curve segment under the period of time, the actual distance of the pair of data points is quantified, and the larger the actual distance is, the smaller the characteristic similarity of traffic flow data of the pair of data points is, and the larger the difference is; the actual distance may be used to calculate a vehicle flow correlation subsequently.

As an example, taking determining the actual distance between any pair of data points as an example, the calculation formula of the actual distance between the h-th pair of data points after matching corresponding to the road pq and the road wv may be:

; in the method, in the process of the invention, For the actual distance between the h-th pair of data points after the matching of the road pq and the road wv, w and v are two adjacent intersections in the urban area to be planned, and the intersection w and the intersection v form a stop road wv,/>For the history traffic flow difference of the h-th data point after the matching corresponding to the road pq and the road wv,/>As a normalization function,/>For the average value of the vehicle flow retention of the target curve segment under the belonged period of the data acquisition time of two data points in the h data points after the matching of the road pq and the road wv,/>For the difference of the vehicle flow retention degree of the target curve segment under the belonged period of the data acquisition time of two data points in the h data points after the matching of the road pq and the road wv, the method comprises the following steps of/>And the historical vehicle flow average value of the h-th data point after the matching of the road pq and the road wv is obtained.

In the calculation formula of the actual distance,The vehicle flow difference degree of the h-th data point after matching can be represented, and the smaller the vehicle flow difference degree is, the higher the association degree is; /(I)The overall traffic flow retention degree of the h-th data point after matching can be characterized, and the greater the overall traffic flow retention degree of a pair of data points, the higher the weight of the data points when analyzing the traffic flow difference condition of the road pq and the road wv on the same day is, so the/>As/>Weights of (2); similarly, a greater overall traffic flow retention for a pair of data points means that the corresponding traffic flow retention for two of the pair of data points is simultaneously greater; in a similar situation where the traffic flow retention between two data points is analyzed, rather than one of the data points having a higher traffic flow retention and the other data point having a lower traffic flow retention, the/>, will beAlso as/>Weights of/>The larger the difference of the vehicle flow retention of the h data point is, the higher the weight constraint of the historical vehicle flow difference of the h data point is, namely the smaller the weight is; The historical traffic flow of the h-th data point can be represented, the actual distance is used for subsequently determining the association degree of the traffic flows, and then the traffic light time between roads is adjusted, if the historical traffic flows are all extremely small, the traffic load at the corresponding data acquisition moment is indicated not to need to be adjusted and controlled in the traffic light time.

And thirdly, determining an initial road and a final road in the road pq and the road wv, acquiring each shunt road corresponding to the initial road, and further determining the actual distance between each pair of data points after matching corresponding to each shunt road corresponding to the initial road and each shunt road corresponding to the final road.

In this embodiment, a first road in a route formed by the road pq and the road wv along the same driving direction is a starting road, and a first intersection of the starting road is a starting intersection; the last road in the route formed along the same driving direction is a termination road, and the first intersection of the termination road is a termination intersection; and taking each road which is communicated with the route and is adjacent to the route as each diversion road corresponding to the initial road. As shown in fig. 3, the schematic diagram of each of the split roads of the travel route pqwv is shown in fig. 3, in which assuming that the overall driving direction is rightward, the road pq is the start road, the road wv is the end road, and each of the adjacent roads that communicate with the travel route pqwv includes the road qv3, the road qv4, the road wv1, and the road wv2, and each of the split roads corresponding to the start road pq is qv3, qv4, wv1, and wv 2.

Referring to the calculation process of the actual distances between the matched pairs of data points corresponding to the road pq and the road wv corresponding to the first step to the second step in step S3, the actual distances between the matched pairs of data points corresponding to the split roads corresponding to the ending road and the starting road can be obtained, and the detailed description is not repeated here.

And fourthly, determining the vehicle flow association degree of the road pq and the road wv on the same day according to the actual distance of each pair of data points after matching corresponding to the road pq and the road wv, the actual distance of each pair of data points after matching corresponding to each shunt road corresponding to the starting road and the ending road and the historical vehicle flow difference of each pair of data points.

In fig. 3, when the traffic flow of the starting road pq is large, the corresponding traffic load is not completely applied to the ending road wv, and there is also traffic flow split for the starting road pq on the road qv3, the road qv4, the road wv1, and the road wv2 on the traveling route corresponding to the starting road pq and the ending road wv, and the larger the traffic flow split for the split road is, the smaller the traffic load influence of the ending road wv on the starting road pq is.

Further analysis, even if the traffic flow of the starting road pq is large, there is not necessarily a large traffic load for the ending road wv on the driving route, so when analyzing the traffic flow association condition between the starting road pq and the ending road wv, since it is necessary to pass through a plurality of intersections before reaching the ending intersection v, each intersection on the driving route has a certain degree of split condition, and therefore, it is necessary to analyze the influence degree of each split road corresponding to the starting road on the traffic flow association degree of the starting road pq and the ending road wv on the same day.

As an example, the calculation formula of the association degree of the traffic flow of the start road pq and the end road wv at the same day may be:

; in the method, in the process of the invention, For the relevance of the traffic flow of the starting road pq and the ending road wv on the same day,/>For the actual distance between the H-th pair of data points after the matching corresponding to the starting road pq and the ending road wv, H is the serial number of each pair of data points after the matching corresponding to the two roads, and H is the pair of data points after the matching corresponding to the two roads,/>For the normalization function, I is the sequence number of each split road corresponding to the initial road pq, I is the number of all split roads corresponding to the initial road pq,/>To obtain the difference between the actual distance of the h-th data point after matching corresponding to the i-th split road corresponding to the start road wv and the actual distance of the h-th data point after matching corresponding to the start road pq and the stop road wv,/>And the historical vehicle flow difference of the h-th data point after matching corresponding to the ith split road corresponding to the start road pq is obtained for the stop road wv.

In the calculation formula of the association degree of the vehicle flow,The larger the actual distance between the h data point after the matching of the starting road pq and the ending road wv is, the worse the characteristic similarity of the flow data of the h data point is,The larger the data feature similarity of the traffic flow data of the starting road pq and the stopping road wv is worse in the same day, the smaller the traffic flow association degree of the starting road pq and the stopping road wv in the same day is; /(I)，For the actual distance of the h-th data point after matching corresponding to the i-th split road corresponding to the start road pq and the stop road wv,/>To find absolute value function,/>The larger the difference between the actual distance between the h-th data point of the starting road pq and the actual distance between the h-th data point of the corresponding i-th split road is, the higher the correlation between the starting road pq and the ending road wv is for the h-th data point than the correlation between the ending road wv and the i-th split road corresponding to the starting road pq; when the traffic flow of the starting road pq suddenly increases, the main traffic load of the starting road pq is distributed to the ending road wv, but not each split road corresponding to the starting road pq, and the higher the weight of the subsequent clustering distance between the starting road pq and the ending road wv is analyzed; /(I)The larger the difference analysis of the h-th data point after the matching corresponding to the i-th split road corresponding to the starting road pq and the ending road wv is, namely/>, the more importantThe higher the confidence of/>Is thatWeights of/>For/>Is a weight of (2).

And fifthly, taking an average value of the vehicle flow correlation degree of the road pq and the road wv at each day as the vehicle flow correlation degree of the road pq and the road wv.

In this embodiment, each of the road pq and the road wv has its corresponding traffic flow association degree, and in order to measure the actual traffic flow association condition between the road pq and the road wv, the average value of the traffic flow association degrees of the road pq and the road wv at each day may be used as the traffic flow association degree of the road pq and the road wv. It is worth to say that, the traffic light time is regulated and controlled based on the traffic flow association degree between roads, mainly to relieve traffic load.

S4, determining clustering optimization weights of any two intersections according to the association degree of the traffic flow of any two roads; and clustering all the intersections in the urban area to be planned by combining the clustering optimization weights of any two intersections to obtain an optimal intersection clustering result.

After obtaining the association degree of the traffic flow of any two roads in all roads, in order to improve the accuracy of the clustering result of the intersections, the traffic flow change condition between the intersections is added to the implementation process of the clustering process of the intersections, specifically based on the association degree of the traffic flow of the roads, the clustering optimization weights of the two intersections during the k-means clustering process are quantized, and then the optimal clustering result of the intersections is obtained, and the specific implementation steps may include:

and determining the clustering optimization weight of any two intersections according to the association degree of the traffic flow of any two roads.

And a first substep, obtaining all routes formed by any two intersections along the driving direction.

In this embodiment, there are multiple different driving routes between any two intersections, and in order to comprehensively quantify the traffic flow change situation between any two intersections, all the driving routes between every two intersections in the urban area to be planned, that is, the driving routes, are obtained based on the GPS system.

When calculating the association degree of the traffic flow, at least two roads are needed to calculate, so that the cluster optimization weight of two adjacent intersections can be assigned to be 1 for the route formed by the path between the two adjacent intersections.

And a second sub-step of determining a first traffic relevance of each route of any two intersections.

In this embodiment, the calculation process of the first traffic association degree of each route of any two intersections is consistent, and taking any one route of any two intersections as an example, the determining the first traffic association degree of the route may include the following specific implementation steps:

Firstly, calculating the accumulated sum of the association degree of the traffic flow of each road except the ending road passing through in the route and the ending road, and recording the accumulated sum as the initial association degree of the traffic flow of the route; then, the reciprocal of the number of roads passed in the route is calculated, and the product of the initial traffic flow association of the route and the reciprocal of the number of roads passed in the route is taken as the first traffic flow association of the route.

When the association degree between the traffic flow of each road except the ending road and the ending road in the route is larger, the influence of the shunting road on the route is smaller, the traffic load of the starting road is mainly distributed on all the roads passing by the route, and the association degree of the initial traffic flow of the route is larger; when the number of roads passing through the route is large, the distance of the route is far and is influenced by the diversion of the road junction on the route, and the reliability of the initial traffic flow association degree determined at the moment is small, so that further analysis on the initial traffic flow association degree of the route is needed, and the product of the initial traffic flow association degree of the route and the reciprocal of the number of roads passing through the route is taken as the traffic flow association degree of the route and is recorded as the first traffic flow association degree.

And a third sub-step of determining a cluster optimization weight of any two intersections based on the first traffic relevance of each route of the two intersections.

In this embodiment, an average value of the first traffic correlation degrees of each route of any two intersections is used as the second traffic correlation degrees of all routes formed by the two intersections along the driving direction; calculating the average value of the second vehicle flow correlation degree of all routes formed by the two intersections along different driving directions, carrying out normalization processing on the average value of the second vehicle flow correlation degree, and taking the value after normalization processing as the clustering optimization weight of the two intersections.

It should be noted that, each route of any two intersections has a corresponding first traffic correlation, and an average value of all the first traffic correlations determined at this time is taken as a traffic correlation of the whole route formed by the two intersections along the driving direction, and is recorded as a second traffic correlation; two driving directions generally exist on a highway, each driving direction has a corresponding second vehicle flow correlation degree, an average value of the second vehicle flow correlation degrees corresponding to the two driving directions is calculated, and the average value of the second vehicle flow correlation degrees corresponding to the two driving directions is normalized by utilizing sigmod functions, so that the clustering optimization weights of the two intersections can be obtained. The normalization implementation process of sigmod functions is prior art and is not within the scope of the present invention, and will not be described in detail here.

And secondly, clustering all the intersections in the urban area to be planned by combining the clustering optimization weights of any two intersections to obtain an optimal intersection clustering result.

It should be noted that, the clustering optimization weight is determined by analyzing the correlation of the historical traffic flow changes between two roads, the correlation of the historical traffic flow changes between the two roads refers to the traffic flow correlation, and the determination of the clustering optimization weight is helpful to overcome the defect that the accuracy of the clustering result is low due to the fact that the traffic flow change trend difference between the intersections is not considered in the existing intersection clustering, and further the reliability of the intersection traffic light time regulation is improved.

The first substep, determining the initial distance measurement characteristics of any two intersections and the inverse proportion value of the clustering optimization weights of any two intersections, and taking the product of the initial distance measurement characteristics of any two intersections and the inverse proportion value of the clustering optimization weights of the corresponding two intersections as the optimized distance measurement characteristics of the corresponding two intersections.

In this embodiment, the initial distance measurement characteristics of any two intersections may be determined through operations such as road level, conventional traffic flow difference, etc., that is, the initial distance measurement characteristics may be determined through conventional intersection clustering implementation means, and the specific determination process of the initial distance measurement characteristics is in the prior art, which is not in the scope of the present invention and will not be described in detail herein.

The larger the clustering optimization weight is, the stronger the relevance between two intersections is, and the more the initial distance measurement characteristics of the two intersections in the clustering process are required to be optimized, the optimization can be realized by correspondingly weakening the initial distance measurement characteristics of the two intersections, so that the inverse operation of the clustering optimization weight is required to be performed to determine an inverse proportion value. The value range of the clustering optimization weight is between 0 and 1, so that the inverse proportion value of the clustering optimization weight can be determined by subtracting the clustering optimization weight from 1.

And a second sub-step, based on the optimized distance measurement characteristics of any two intersections, carrying out k-means clustering treatment on all the intersections in the urban area to be planned to obtain a clustering result, and taking the clustering result as an optimal intersection clustering result.

In this embodiment, the implementation process of the k-means clustering process is the prior art, which is not in the scope of the present invention and will not be described in detail here. Compared with the traditional intersection clustering process, the optimal intersection clustering result determined by the embodiment has good clustering effect, and the clustering weights of two intersections with the correlation of the traffic flow change can be enhanced to a certain extent, so that the intersection clustering result with higher accuracy is obtained.

And S5, regulating and controlling the intersection traffic lights in the urban area to be planned according to the optimal intersection clustering result.

In this embodiment, the traffic light time of the intersections in the urban area to be planned is regulated and controlled based on the optimal intersection clustering result corresponding to the urban area to be planned, that is, based on the intersections in each cluster. The detailed process of realizing the traffic light time regulation and control of the intersection through the intersection clustering result is the prior art, and is not in the protection scope of the invention, and is not described in detail here.

The invention provides a road traffic planning method, which is characterized in that the clustering optimization weight of any two intersections is determined when all the intersections are clustered by analyzing the association degree of traffic flow between every two roads, so that an optimal intersection clustering result with higher accuracy is obtained, more reliable intersection traffic light time regulation is realized, and the traffic planning rationality is further improved.

The above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the invention and are intended to be included within the scope of the invention.

Claims (9)

1. A road traffic planning method, characterized by comprising the steps of:

Acquiring the historical vehicle flow of each road in the urban area to be planned in each unit time within a plurality of days, and further determining the historical vehicle flow curve of each road in each day; the road section between two adjacent intersections forms a road, and the historical traffic flow is the historical traffic flow in the same driving direction on the road;

Dividing a day into time periods with the same length, and determining the traffic flow retention degree of each road in the same time period according to each extreme point of a curve section of the historical traffic flow curve of each road in the same time period and the historical traffic flow average value of the curve section;

Determining the association degree of the traffic flow of any two roads according to the difference condition between the historical traffic flow curves of any two roads in each day and the traffic flow maintenance degree of any two roads in the same period;

Determining clustering optimization weights of any two intersections according to the association degree of the traffic flow of any two roads; clustering all intersections of the urban area to be planned by combining the clustering optimization weights of any two intersections to obtain an optimal intersection clustering result;

Regulating and controlling intersection traffic lights in the urban area to be planned according to the optimal intersection clustering result;

the determining the traffic flow retention of each road in the same period according to each extreme point of the curve section of the historical traffic flow curve of each road in the same period and the historical traffic flow average value of the curve section, comprising:

for any road and any time period, taking a curve section of a historical traffic flow curve of the road in each day under the time period as a target curve section to obtain each corresponding target curve section of the road under the time period;

Determining each extreme point of each target curve segment, further determining the time span between each extreme point and the next extreme point adjacent to each extreme point, and obtaining the time span corresponding to each extreme point of each target curve segment;

Determining the traffic flow maintaining weight of each target curve segment corresponding to the road in the time period according to each extreme point of each target curve segment and the corresponding time span thereof;

and determining the traffic flow retention of the road in the time period according to the historical traffic flow average value and the traffic flow retention weight of each corresponding target curve segment of the road in the time period.

2. The road traffic planning method according to claim 1, wherein the calculation formula of the traffic flow holding weight of the mth target curve segment corresponding to the road pq at the nth period is:

; in the/> Maintaining weights for the traffic flow of the mth target curve segment corresponding to the road pq in the nth time period, wherein p and q are two adjacent intersections in the urban area to be planned, and the intersection p and the intersection q form the road pq,/>For the normalization function, Z is the extreme point sequence number of the mth target curve segment corresponding to the road pq in the nth period, and Z is the number of extreme points of the mth target curve segment corresponding to the road pq in the nth period,/>For the time span corresponding to the z extreme point of the corresponding mth target curve segment of the road pq under the nth time period,/>For the time span mean value corresponding to all extreme points of the mth target curve segment corresponding to the road pq in the nth period of time,/>The difference between the historical traffic flow of the z extreme point of the corresponding mth target curve segment of the road pq under the nth time period and the historical traffic flow average of the mth target curve segment is obtained.

3. The method for road traffic planning according to claim 1, wherein the determining the traffic flow retention of the road in the period according to the historical traffic flow average and the traffic flow retention weight of each corresponding target curve segment of the road in the period comprises:

for any one target curve segment, firstly calculating the difference between the average value of the historical traffic flow of the target curve segment and the average value of the historical traffic flow of all the target curve segments corresponding to the road in the period, and recording the difference as a target difference; calculating the product of the target difference and the traffic flow maintaining weight of the target curve segment, and recording the product as a first product;

and obtaining a first product of each corresponding target curve segment of the road in the period, and taking an average value of the first products of each corresponding target curve segments of the road in the period as the traffic flow maintenance degree of the road in the period.

4. The method for planning road traffic according to claim 1, wherein determining the association of traffic flows of any two roads according to the difference between the historic traffic flow curves of any two roads in each day and the traffic flow maintenance of any two roads in the same period comprises:

Performing DTW (draw-off line) matching on historical traffic flow curves of any two roads in the same day to obtain matched data points, and further determining traffic flow retention of a target curve segment of data acquisition time of two data points in each pair of data points under the period of time;

Determining the actual distance between the matched pairs of data points corresponding to any two roads according to the historical traffic flow of the matched pairs of data points and the traffic flow retention of the target curve segment of two data points in the pairs of data points under the period of time;

Determining an initial road and a final road in any two roads, acquiring each split road corresponding to the initial road, and further determining the actual distance between each pair of data points after the final road and each split road corresponding to the initial road are matched;

determining the association degree of the traffic flow of any two roads on the same day according to the actual distance of each pair of data points after the matching corresponding to any two roads, the actual distance of each pair of data points after the matching corresponding to each shunt road corresponding to the starting road and the historical traffic flow difference of each pair of data points;

And taking the average value of the vehicle flow correlation degrees of any two roads at each day as the vehicle flow correlation degree of any two roads.

5. The road traffic planning method according to claim 4, wherein a first road in a route formed by any two roads along the same driving direction is a starting road, and a first intersection of the starting road is a starting intersection; the last road in the route formed by any two roads along the same driving direction is a termination road, and the first intersection of the termination road is a termination intersection; each road adjacent to a route formed by any two roads along the same driving direction is used as each split road corresponding to the starting road.

6. The road traffic planning method according to claim 5, wherein the calculation formula of the actual distance of the h-th data point after matching corresponding to the road pq and the road wv is:

; in the method, in the process of the invention, For the actual distance between the h-th pair of data points after the matching of the road pq and the road wv, w and v are two adjacent intersections in the urban area to be planned, and the intersection w and the intersection v form the road wv,/>For the history traffic flow difference of the h-th data point after the matching corresponding to the road pq and the road wv,/>As a normalization function,/>For the average value of the vehicle flow retention of the target curve segment under the belonged period of the data acquisition time of two data points in the h data points after the matching of the road pq and the road wv,/>For the difference of the vehicle flow retention degree of the target curve segment under the belonged period of the data acquisition time of two data points in the h data points after the matching of the road pq and the road wv, the method comprises the following steps of/>And the historical vehicle flow average value of the h-th data point after the matching of the road pq and the road wv is obtained.

7. The road traffic planning method according to claim 6, wherein the calculation formula of the association degree of the traffic flow of the road pq and the road wv on the same day is:

; in the/> For the association degree of the traffic flow of the road pq and the road wv on the same day,/>For the actual distance between the H-th pair of data points after matching corresponding to the road pq and the road wv, H is the serial number of each pair of data points after matching corresponding to the two roads, and H is the pair of data points after matching corresponding to the two roads,/>For the normalization function, I is the serial number of each split road corresponding to the initial road, I is the number of all split roads corresponding to the initial road,/>To terminate the actual distance of the h-th pair of data points after matching corresponding to the i-th split road corresponding to the initial road and the difference between the actual distance of the h-th pair of data points after matching corresponding to the road pq and the road wv,/>And the historical vehicle flow difference of the h-th data point after matching corresponding to the ith split road corresponding to the ending road and the starting road is obtained.

8. The method for planning road traffic according to claim 5, wherein determining the cluster optimization weights of any two intersections according to the association degree of the traffic flows of any two roads comprises:

Acquiring all routes formed by any two intersections along the driving direction;

for any one route of any two intersections, calculating the accumulated sum of the association degrees of the traffic flow of all the roads except the ending road passing through in the route and the ending road, and recording the accumulated sum as the initial association degree of the traffic flow of the route; calculating the reciprocal of the number of roads passing through the route, and taking the product of the initial vehicle flow association degree of the route and the reciprocal of the number of roads passing through the route as the first vehicle flow association degree of the route;

taking the average value of the first vehicle flow correlation degree of each route of any two intersections as the second vehicle flow correlation degree of all routes formed by the two intersections along the driving direction; calculating the average value of the second vehicle flow correlation degrees of all routes formed by the two intersections along different driving directions, carrying out normalization processing on the average value of the second vehicle flow correlation degrees corresponding to the different driving directions, and taking the value after normalization processing as the clustering optimization weight of the two intersections.

9. The method for planning road traffic according to claim 1, wherein the clustering of all intersections in the urban area to be planned by combining the clustering optimization weights of any two intersections to obtain an optimal intersection clustering result comprises:

Determining the initial distance measurement characteristics of any two intersections and the inverse proportion value of the clustering optimization weights of any two intersections, and taking the product of the initial distance measurement characteristics of any two intersections and the inverse proportion value of the clustering optimization weights of the corresponding two intersections as the optimized distance measurement characteristics of the corresponding two intersections;

and carrying out k-means clustering on all the intersections in the urban area to be planned based on the optimized distance measurement characteristics of any two intersections to obtain a clustering result, and taking the clustering result as an optimal intersection clustering result.