CN115577306B - Self-adaptive density clustering-based method for detecting travel tide area of shared bicycle - Google Patents

Abstract

The embodiment of the present invention provides a tidal area detection method for shared bicycle travel based on adaptive density clustering, which belongs to the field of computing technology, and specifically includes: Step 1, performing data cleaning on the trajectory points generated by shared bicycle travel orders in the target area Then extract the potential borrowing and returning areas of shared bicycles; step 2, design the kernel density estimation function to construct the spatial density field of trajectory points, and adaptively construct the spatial neighborhood of trajectory points by measuring the instability of the spatial density field, and construct the spatial density accordingly The significance test function extracts the spatially dense neighborhood; step 3, according to the significance test of the spatially dense neighborhood and the trajectory point density feature, adaptively identifies the tidal neighborhood of the trajectory point and extracts the tidal area of shared bicycle travel based on the density reachable extension. Through the solution of the present invention, the starting and ending points of the shared bicycles are adaptively identified to gather areas, so that the tidal regions of the shared bicycles are adaptively distinguished and extracted, and the adaptability and accuracy of the detection process are improved.

Description

A method for detecting tidal areas of shared bicycle travel based on adaptive density clustering

Technical Field

The embodiments of the present invention relate to the field of computing technology, and in particular to a method for detecting tidal areas for shared bicycle travel based on adaptive density clustering.

Background Art

At present, the detection methods of shared bicycle tidal areas are mainly divided into two types: based on spatial unit feature measurement and based on spatial trajectory clustering detection. Among them, the method based on spatial unit feature measurement relies on hard spatial divisions such as electronic fences or spatial grids, and extracts spatial units with excessive net flow or retention density by setting parameter thresholds. However, due to the errors of the spatial positioning equipment itself and the tolerance of the parking and locking system, shared bicycles are not always strictly parked within the scope of the electronic fence. Therefore, spatial unit division is very likely to underestimate the tidal range and imbalance of shared bicycles. At the same time, the setting of spatial unit feature thresholds depends on expert experience and knowledge, which greatly limits the reliability of the results and model transferability of such methods. The method based on spatial trajectory clustering detection avoids the problem of spatial unit division and directly reflects the real bicycle borrowing and returning aggregation area through the clustering of shared bicycle start and end positioning points. However, the existing methods of this type split the distribution characteristics of the shared bicycle start and end positioning points in the clustering process, and the identification of tidal areas has not yet gotten rid of the manual setting of thresholds, which can easily mistakenly identify areas with large shared bicycle flow and dynamic balance of borrowing and returning as tidal areas.

It can be seen that there is an urgent need for a shared bicycle travel tidal area detection method based on adaptive density clustering that fully takes into account the travel space characteristics and shared bicycle trajectory characteristics.

Summary of the invention

In view of this, an embodiment of the present invention provides a method for detecting tidal areas for shared bicycle travel based on adaptive density clustering, which at least partially solves the problems of poor adaptability and accuracy in the prior art.

The embodiment of the present invention provides a method for detecting tidal areas of shared bicycle travel based on adaptive density clustering, comprising:

Step 1: Clean the data of the trajectory points generated by the shared bicycle travel orders in the target area and extract the potential borrowing and returning areas of the shared bicycles;

Step 2: Design a kernel density estimation function to construct the spatial density field of trajectory points, adaptively construct the spatial neighborhood of trajectory points by measuring the instability of the spatial density field, and construct a spatial density significance test function to extract the spatial dense neighborhood based on it;

Step 3: Adaptively identify the tidal neighborhood of the trajectory points based on the significance test of spatial dense neighborhood and trajectory point density features, and extract the tidal area of shared bicycle travel based on density reachable expansion.

According to a specific implementation of the embodiment of the present invention, step 1 specifically includes:

Step 1.1, delete the trajectory data outside the target area, within the water system and buildings;

Step 1.2, delete the trajectory data with abnormal timestamp format, sort the trajectory points of each shared bicycle order in ascending order by timestamp, and only keep one record for the abnormal situation of continuous switch lock status;

Step 1.3: For the same shared bicycle track points with repeated location coordinates, only the record with the earliest timestamp is retained;

Step 1.4, deleting a pair of trajectory points whose timestamp interval is outside the preset range;

Step 1.5, generate a spatial buffer zone for each trajectory point according to a preset radius, and overlap and merge the spatial buffer zones generated by all trajectory points to construct a potential bicycle borrowing and returning area.

According to a specific implementation of the embodiment of the present invention, step 2 specifically includes:

Step 2.1, design a kernel density estimation function, and given any trajectory point in the potential borrowing and returning area of a bicycle, obtain an approximate expression of the relative value of the kernel density at its spatial position as the spatial density field of the trajectory point;

Step 2.2, constructing the spatial neighborhood of the trajectory point based on the instability of the spatial kernel density of all trajectory points at the locations of the trajectory point spatial density field using the entropy function;

Step 2.3, under the assumption of spatial random distribution, calculate the theoretical probability distribution of k other trajectory points appearing in any spatial neighborhood, and given any trajectory point and its corresponding spatial neighborhood and calculate the probability distribution value, the spatial neighborhood with a probability distribution value greater than the significance level value is regarded as a spatial dense neighborhood.

，式中，

表示点

和点

间的欧氏空间距离，

表示研究区域内所有满足

的轨迹点，h表示核密度函数带宽；According to a specific implementation of an embodiment of the present invention, the approximate expression of the relative value of the kernel density is:

, where

Indicate point

and Point

The Euclidean distance between

Indicates that all the

The trajectory points, h represents the bandwidth of the kernel density function;

The expression of the entropy function is:

，式中，N表示目标区域内所有轨迹点数量；

, where N represents the number of all trajectory points in the target area;

The theoretical probability distribution of k other trajectory points appearing in any spatial neighborhood is expressed as

，式中，N表示研究区域内所有OD点数量，Area()表示面积计算函数，SN和ODA分别表示任一轨迹点空间邻域和单车潜在借还区域；

, where N represents the number of all OD points in the study area, Area () represents the area calculation function, SN and ODA represent the spatial neighborhood of any trajectory point and the potential borrowing and returning area of a single vehicle, respectively;

The expression of the probability distribution value is:

表示点

空间邻域

内其他轨迹点的数量。In the formula,

Indicate point

Spatial neighborhood

The number of other trajectory points within.

According to a specific implementation of the embodiment of the present invention, step 2.2 specifically includes:

取值范围

和步长

，迭代计算

，将

最小值对应的

设置为目标区域轨迹点的空间邻域长度；The entropy function is used to measure the spatial kernel density instability of the locations of all trajectory points in the study area.

Value range

and step length

, iterative calculation

,Will

The minimum value corresponds to

Set to the spatial neighborhood length of the trajectory points in the target area;

The spatial neighborhood of a trajectory point is constructed with any trajectory point as the center and the length of the spatial neighborhood as the radius.

According to a specific implementation of the embodiment of the present invention, step 3 specifically includes:

Step 3.1, calculate the theoretical probability of k shared bicycle trip starting points appearing in any spatially dense neighborhood;

Step 3.2, according to the calculation results of step 3.1, calculate the probability value of the spatial dense neighborhood corresponding to any trajectory point becoming the spatial tidal source neighborhood in the target area, and take the spatial dense neighborhood with a probability value greater than the significance level value as the spatial tidal source neighborhood, and take the spatial dense neighborhood with a probability value less than the significance level value as the spatial tidal sink neighborhood;

为种子点，定义

内其他起点

与

密度可达，将所有与

密度可达点聚合为空间潮汐源区域

，对于

内满足空间潮汐源邻域判别的其他点

，继续执行聚合操作更新

，直到

内所有满足空间潮汐源邻域判别的其他点全部访问完毕，重新选取目标区域内未被访问的其他空间潮汐源邻域重复上述过程，最终得到目标区域内提取的全部空间潮汐源区域表示为

；Step 3.3, take any spatial tidal source neighborhood

is the seed point, and defines

Other starting points

and

The density can be reached, all

Density-reachable points are aggregated into spatial tidal source regions

,for

Other points that meet the spatial tidal source neighborhood judgment

, continue to perform aggregate operation updates

,until

After all other points in the target area that meet the spatial tidal source neighborhood judgment are visited, other spatial tidal source neighborhoods that have not been visited in the target area are reselected to repeat the above process. Finally, all spatial tidal source areas extracted in the target area are expressed as

;

为种子点，定义

内其他终点

与

密度可达，将所有与

密度可达点聚合为空间潮汐汇区域

，对于

内满足空间潮汐汇邻域判别的其他点

，继续执行聚合操作更新

，直到

内所有满足空间潮汐汇邻域判别的其他点全部访问完毕,重新选取目标区域内未被访问的其他空间潮汐汇邻域重复上述过程，最终得到目标区域内提取的全部空间潮汐汇区域表示为

，目标区域内提取的全部空间潮汐区域表示为潮汐源区域和潮汐汇区域的并集Step 3.4, take any spatial tidal sink neighborhood

is the seed point, and defines

Other endpoints

and

The density can be reached, all

Density-reachable points are aggregated into spatial tidal sink areas

,for

Other points that meet the spatial tidal sink neighborhood judgment

, continue to perform aggregate operation updates

,until

After all other points in the target area that meet the spatial tidal sink neighborhood judgment are visited, other spatial tidal sink neighborhoods that have not been visited in the target area are reselected to repeat the above process. Finally, all spatial tidal sink areas extracted in the target area are expressed as

, all spatial tidal areas extracted within the target area are represented as the union of tidal source areas and tidal sink areas

。

.

According to a specific implementation of an embodiment of the present invention, the expression for the theoretical probability of k shared bicycle travel starting points appearing in any spatially dense neighborhood is:

表示空间密集邻域内起点数量，

表示空间密集邻域内轨迹点总数量，p表示目标区域内出行起点比例;In the formula,

represents the number of starting points in a spatially dense neighborhood,

represents the total number of trajectory points in the spatially dense neighborhood, and p represents the proportion of travel starting points in the target area;

The expression of the probability value is:

表示

内起点数量，

表示

内轨迹点总数量。In the formula,

express

The number of internal starting points,

express

The total number of inner track points.

The shared bicycle travel tidal area detection scheme based on adaptive density clustering in the embodiment of the present invention includes: step 1, extracting the potential borrowing and returning areas of shared bicycles after data cleaning of the trajectory points generated by shared bicycle travel orders in the target area; step 2, designing a kernel density estimation function to construct the spatial density field of the trajectory points, adaptively constructing the spatial neighborhood of the trajectory points by measuring the instability of the spatial density field, and constructing a spatial density significance test function based on this to extract the spatial dense neighborhood; step 3, adaptively identifying the tidal neighborhood of the trajectory points according to the significance test of the spatial dense neighborhood and the trajectory point density features, and extracting the shared bicycle travel tidal area based on the density reachable expansion.

The beneficial effects of the embodiments of the present invention are as follows: through the scheme of the present invention, the travel space characteristics and the shared bicycle trajectory characteristics are fully taken into account, and the shared bicycle starting and ending point aggregation areas are adaptively identified, so that the shared bicycle travel tidal areas can be adaptively identified and extracted, thereby improving the adaptability and accuracy of the detection process.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

FIG1 is a schematic flow chart of a method for detecting tidal areas for shared bicycle travel based on adaptive density clustering according to an embodiment of the present invention;

FIG2 is a schematic diagram of a specific implementation process of a method for detecting tidal areas for shared bicycle travel based on adaptive density clustering provided by an embodiment of the present invention;

FIG3 is a schematic diagram of a certain place as a target area provided by an embodiment of the present invention;

FIG4 is a schematic diagram of an overall detection result of a tidal area provided by an embodiment of the present invention;

FIG5 is a schematic diagram of tidal area detection results around a subway station in a certain area provided by an embodiment of the present invention.

DETAILED DESCRIPTION

The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

The following describes the embodiments of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and/or function described herein is merely illustrative. Based on the present invention, it should be understood by those skilled in the art that an aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement the device and/or practice the method. In addition, other structures and/or functionalities other than one or more of the aspects described herein can be used to implement this device and/or practice this method.

It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention. The drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

Additionally, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, one skilled in the art will appreciate that the aspects described may be practiced without these specific details.

An embodiment of the present invention provides a method for detecting tidal areas for shared bicycle travel based on adaptive density clustering, which can be applied to the public transportation optimization process in urban traffic planning scenarios.

Referring to FIG1 , a flow chart of a method for detecting tidal areas for shared bicycle travel based on adaptive density clustering is provided in an embodiment of the present invention. As shown in FIG1 and FIG2 , the method mainly includes the following steps:

Step 1: Clean the data of the trajectory points generated by the shared bicycle travel orders in the target area and extract the potential borrowing and returning areas of the shared bicycles;

Furthermore, the step 1 specifically includes:

Step 1.1, delete the trajectory data outside the target area, within the water system and buildings;

Step 1.2, delete the trajectory data with abnormal timestamp format, sort the trajectory points of each shared bicycle order in ascending order by timestamp, and only keep one record for the abnormal situation of continuous switch lock status;

Step 1.3: For the same shared bicycle track points with repeated location coordinates, only the record with the earliest timestamp is retained;

Step 1.4, deleting a pair of trajectory points whose timestamp interval is outside the preset range;

Step 1.5, generate a spatial buffer zone for each trajectory point according to a preset radius, and overlap and merge the spatial buffer zones generated by all trajectory points to construct a potential bicycle borrowing and returning area.

In the specific implementation, the OD point (i.e., track point) trajectory generated by the shared bicycle travel order is cleaned and the potential bicycle borrowing and returning area is extracted. Specifically, it includes:

1.1 Data Cleansing

First, delete the trajectory data outside the study area, in the water system, and in the building. Then, delete the trajectory data with abnormal timestamp format. Furthermore, sort the OD points of each shared bicycle order in ascending order by timestamp. For the abnormal situation of continuous switch lock status, only one record is retained. At the same time, for the same shared bicycle trajectory points with repeated location coordinates, only the record with the earliest timestamp is retained. Finally, delete a pair of OD points with an OD point timestamp interval of less than 1 minute or greater than 1 hour.

1.2 Extraction of potential bicycle borrowing and returning areas

A spatial buffer with a radius of 50 meters is generated for each OD point. The spatial buffers generated by all OD points are superimposed and merged to construct the bicycle potential borrowing and returning area ODA , which further refines the constraints on all spatial ranges where bicycles may be borrowed and returned.

Step 2: Design a kernel density estimation function to construct the spatial density field of trajectory points, adaptively construct the spatial neighborhood of trajectory points by measuring the instability of the spatial density field, and construct a spatial density significance test function to extract the spatial dense neighborhood based on it;

Based on the above embodiment, step 2 specifically includes:

Step 2.1, design a kernel density estimation function, and given any trajectory point in the potential borrowing and returning area of a bicycle, obtain an approximate expression of the relative value of the kernel density at its spatial position as the spatial density field of the trajectory point;

Step 2.2, constructing the spatial neighborhood of the trajectory point based on the instability of the spatial kernel density of all trajectory points at the locations of the trajectory point spatial density field using the entropy function;

Step 2.3, under the assumption of spatial random distribution, calculate the theoretical probability distribution of k other trajectory points appearing in any spatial neighborhood, and given any trajectory point and its corresponding spatial neighborhood and calculate the probability distribution value, the spatial neighborhood with a probability distribution value greater than the significance level value is regarded as a spatial dense neighborhood.

，式中，

表示点

和点

间的欧氏空间距离，

表示研究区域内所有满足

的轨迹点，h表示核密度函数带宽；Optionally, the relative value of the kernel density is approximated as:

, where

Indicate point

and Point

The Euclidean distance between

Indicates that all the

The trajectory points, h represents the bandwidth of the kernel density function;

The expression of the entropy function is:

，式中，N表示目标区域内所有轨迹点数量；

, where N represents the number of all trajectory points in the target area;

The theoretical probability distribution of k other trajectory points appearing in any spatial neighborhood is expressed as

，式中，N表示研究区域内所有OD点数量，Area()表示面积计算函数，SN和ODA分别表示任一轨迹点空间邻域和单车潜在借还区域；

, where N represents the number of all OD points in the study area, Area () represents the area calculation function, SN and ODA represent the spatial neighborhood of any trajectory point and the potential borrowing and returning area of a single vehicle, respectively;

式中，

表示点

空间邻域

内其他轨迹点的数量。The expression of the probability distribution value is:

In the formula,

Indicate point

Spatial neighborhood

The number of other trajectory points within.

Furthermore, the step 2.2 specifically includes:

取值范围

和步长

，迭代计算

，将

最小值对应的

设置为目标区域轨迹点的空间邻域长度；The entropy function is used to measure the spatial kernel density instability of the locations of all trajectory points in the study area.

Value range

and step length

, iterative calculation

,Will

The minimum value corresponds to

Set to the spatial neighborhood length of the trajectory points in the target area;

The spatial neighborhood of a trajectory point is constructed with any trajectory point as the center and the length of the spatial neighborhood as the radius.

In specific implementation, the present invention adopts an adaptive density test method to identify the densely clustered neighborhood of OD points as the candidate sub-area for shared bicycle travel tidal detection. First, a kernel density estimation function is designed to construct the spatial density field of OD points, and then the spatial neighborhood of OD points is adaptively constructed by measuring the instability of spatial density, and then a spatial density significance test function is constructed to extract the dense neighborhood of OD points. Specifically, it includes:

2.1, OD point spatial neighborhood construction

Given any trajectory point p _i , the relative value of the kernel density at its spatial position is approximately expressed as:

表示点

和点

间的欧氏空间距离，

表示研究区域内所有满足

的轨迹点，h表示核密度函数带宽。在此基础上，利用熵函数度量研究区域内所有OD点所在位置的空间核密度不稳定性：In the formula,

Indicate point

and Point

The Euclidean distance between

Indicates that all the

The trajectory points are represented by h, and h represents the bandwidth of the kernel density function. On this basis, the entropy function is used to measure the spatial kernel density instability of all OD points in the study area:

宽松的取值范围

和步长

，迭代计算

，将

最小值对应的

设置为研究区域轨迹点空间邻域eps。基于此，以任一轨迹点

为圆心、eps的长度为半径构建的圆形空间区域表示

的空间邻域，记为

。Where N represents the number of all OD points in the study area.

Loose value range

and step length

, iterative calculation

,Will

The minimum value corresponds to

Set to the spatial neighborhood eps of the trajectory points in the study area. Based on this, any trajectory point

The circular space area is represented by the center and the length of eps as the radius.

The spatial neighborhood of

.

2.2, OD point dense neighborhood identification

Under the assumption of random spatial distribution, the theoretical probability of k other OD points appearing in any spatial neighborhood SN is the Poisson distribution density function:

，

成为研究区域内空间密集邻域的概率计算为泊松分布的概率分布函数：Where N represents the number of all OD points in the study area; Area () represents the area calculation function; SN and ODA represent the spatial neighborhood of any trajectory point and the potential borrowing and returning area of a single vehicle, respectively. Given any OD point pi and its spatial neighborhood

,

The probability of being a spatially dense neighbor within the study area is calculated as the probability distribution function of the Poisson distribution:

表示点

空间邻域

内其他OD点的数量。若

大于给定具有统计意义的显著性水平α（如0.95），则

标记为空间密集邻域

。In the formula,

Indicate point

Spatial neighborhood

The number of other OD points within.

is greater than a given statistically significant level α (such as 0.95), then

Marked as a spatially dense neighborhood

.

Step 3: Adaptively identify the tidal neighborhood of the trajectory points based on the significance test of spatial dense neighborhood and trajectory point density features, and extract the tidal area of shared bicycle travel based on density reachable expansion.

Based on the above embodiment, step 3 specifically includes:

Step 3.1, calculate the theoretical probability of k shared bicycle trip starting points appearing in any spatially dense neighborhood;

Step 3.2, according to the calculation results of step 3.1, calculate the probability value of the spatial dense neighborhood corresponding to any trajectory point becoming the spatial tidal source neighborhood in the target area, and take the spatial dense neighborhood with a probability value greater than the significance level value as the spatial tidal source neighborhood, and take the spatial dense neighborhood with a probability value less than the significance level value as the spatial tidal sink neighborhood;

为种子点，定义

内其他起点

与

密度可达，将所有与

密度可达点聚合为空间潮汐源区域

，对于

内满足空间潮汐源邻域判别的其他点

，继续执行聚合操作更新

，直到

内所有满足空间潮汐源邻域判别的其他点全部访问完毕,重新选取目标区域内未被访问的其他空间潮汐源邻域重复上述过程，最终得到目标区域内提取的全部空间潮汐源区域表示为

；Step 3.3, take any spatial tidal source neighborhood

is the seed point, and defines

Other starting points

and

The density can be reached, all

Density-reachable points are aggregated into spatial tidal source regions

,for

Other points that meet the spatial tidal source neighborhood judgment

, continue to perform aggregate operation updates

,until

After all other points in the target area that meet the spatial tidal source neighborhood judgment are visited, other spatial tidal source neighborhoods that have not been visited in the target area are reselected to repeat the above process. Finally, all spatial tidal source areas extracted in the target area are expressed as

;

为种子点，定义

内其他终点

与

密度可达，将所有与

密度可达点聚合为空间潮汐汇区域

，对于

内满足空间潮汐汇邻域判别的其他点

，继续执行聚合操作更新

，直到

内所有满足空间潮汐汇邻域判别的其他点全部访问完毕，重新选取目标区域内未被访问的其他空间潮汐汇邻域重复上述过程，最终得到目标区域内提取的全部空间潮汐汇区域表示为

，目标区域内提取的全部空间潮汐区域表示为潮汐源区域和潮汐汇区域的并集Step 3.4, take any spatial tidal sink neighborhood

is the seed point, and defines

Other endpoints

and

The density can be reached, all

Density-reachable points are aggregated into spatial tidal sink areas

,for

Other points that meet the spatial tidal sink neighborhood judgment

, continue to perform aggregate operation updates

,until

After all other points in the target area that meet the spatial tidal sink neighborhood judgment are visited, other spatial tidal sink neighborhoods that have not been visited in the target area are reselected to repeat the above process. Finally, all spatial tidal sink areas extracted in the target area are expressed as

, all spatial tidal areas extracted within the target area are represented as the union of tidal source areas and tidal sink areas

。

.

Furthermore, the theoretical probability of k shared bicycle trip starting points appearing in any spatially dense neighborhood is expressed as

表示空间密集邻域内起点数量,

表示空间密集邻域内轨迹点总数量，p表示目标区域内出行起点比例;In the formula,

represents the number of starting points in a spatially dense neighborhood,

represents the total number of trajectory points in the spatially dense neighborhood, and p represents the proportion of travel starting points in the target area;

The expression of the probability value is:

表示

内起点数量，

表示

内轨迹点总数量。In the formula,

express

The number of internal starting points,

express

The total number of inner track points.

In specific implementation, based on the above-mentioned identification of densely clustered neighborhoods of OD points, the present invention adopts a method based on OD point density feature clustering to detect the shared bicycle travel tidal area. First, the OD point tidal neighborhood is adaptively identified based on the OD point density feature significance test, and then the shared bicycle travel tidal area is extracted based on density reachable expansion. Specifically including:

3.1, OD point tidal neighborhood identification

Given any spatially dense neighborhood SDN , the theoretical probability of k bicycle trip starting points appearing in the SDN is the binomial distribution density function:

表示

内起点数量，

表示空间密集邻域内轨迹点总数量，p表示目标区域内出行起点比例。给定任一OD点

及其空间密集邻域

，

成为研究区域内空间潮汐源邻域的可能性计算为：In the formula,

express

The number of internal starting points,

represents the total number of trajectory points in the spatially dense neighborhood, and p represents the proportion of travel starting points in the target area.

and its spatially dense neighborhood

,

The probability of becoming a neighbor of a spatial tidal source in the study area is calculated as:

表示

内起点数量，

表示

内轨迹点总数量。若

大于给定具有统计意义的显著性水平α（如0.95），则

标记为空间潮汐源邻域

；若

小于给定具有统计意义的显著性水平1-α，则

标记为空间潮汐汇邻域

。In the formula,

express

The number of internal starting points,

express

The total number of inner track points.

is greater than a given statistically significant level α (such as 0.95), then

Marked as spatial tidal source neighborhood

;like

is less than the given statistically significant level 1- α , then

Marked as a spatial tidal sink neighborhood

.

3.2, OD point tidal area extraction

为种子点，定义

内其他起点

与

密度可达，将所有与

密度可达点聚合为空间潮汐源区域

，对于

内满足空间潮汐源邻域判别的其他点

，继续执行聚合操作更新

，直到

内所有满足空间潮汐源邻域判别的其他点全部访问完毕,重新选取目标区域内未被访问的其他空间潮汐源邻域重复上述过程，最终得到目标区域内提取的全部空间潮汐源区域表示为

。Take any spatial tidal source neighborhood

is the seed point, and defines

Other starting points

and

The density can be reached, all

Density-reachable points are aggregated into spatial tidal source regions

,for

Other points that meet the spatial tidal source neighborhood judgment

, continue to perform aggregate operation updates

,until

After all other points in the target area that meet the spatial tidal source neighborhood judgment are visited, other spatial tidal source neighborhoods that have not been visited in the target area are reselected to repeat the above process. Finally, all spatial tidal source areas extracted in the target area are expressed as

.

为种子点，定义

内其他终点

与

密度可达，将所有与

密度可达点聚合为空间潮汐汇区域

，基于以上空间扩展策略，最终，研究区域内提取的全部空间潮汐汇区域表示为

。研究区域内提取的全部空间潮汐区域表示为潮汐源区域和潮汐汇区域的并集Similarly, any spatial tidal sink neighborhood

is the seed point, and defines

Other endpoints

and

The density can be reached, all

Density-reachable points are aggregated into spatial tidal sink areas

, based on the above spatial expansion strategy, finally, all the spatial tidal sink areas extracted in the study area are expressed as

The total spatial tidal area extracted in the study area is represented as the union of the tidal source area and the tidal sink area.

。

.

The shared bicycle travel tidal area detection method based on adaptive density clustering provided in this embodiment fully considers the travel space characteristics and shared bicycle trajectory characteristics, and adaptively identifies the shared bicycle starting and ending point aggregation areas, so as to adaptively distinguish and extract the shared bicycle travel tidal area, thereby improving the adaptability and accuracy of the detection process.

The present solution will be described below in conjunction with a specific embodiment, and the specific implementation process of the present invention will be described using the OD point data of shared bicycles in a certain city on December 21, 2020:

(1) In this embodiment, District A and District B of a city are selected to constitute the target area, and the OD point data of shared bicycles is used. The data time is 6:00:00-10:00:00 on December 21, 2020. The trajectory data and the study area are shown in Figure 3.

(2) Delete trajectory data outside the study area and within water systems and buildings; delete trajectory data with abnormal timestamp format; arrange the OD points of each shared bicycle order in ascending order by timestamp, and only retain one record for abnormal situations where the switch lock status is continuous; for the same shared bicycle trajectory point with repeated location coordinates, only retain the record with the earliest timestamp; delete a pair of OD points with an OD point timestamp interval of less than 1 minute or greater than 1 hour.

(3) Generate a spatial buffer with a radius of 50 meters for each OD point. Overlay and merge the spatial buffers generated by all OD points to construct the bicycle potential borrowing and returning area (ODA) , further refining the constraints on all spatial ranges where bicycles may be borrowed and returned.

，其空间位置处的核密度相对值近似表达为：(4) Construction of OD point spatial neighborhood. Given any trajectory point

, the relative value of the kernel density at its spatial position is approximately expressed as:

表示点

和点

间的欧氏空间距离，

表示研究区域内所有满足

的轨迹点，h表示核密度函数带宽。在此基础上，利用熵函数度量研究区域内所有OD点所在位置的空间核密度不稳定性：In the formula,

Indicate point

and Point

The Euclidean distance between

Indicates that all the

The trajectory points are represented by h, and h represents the bandwidth of the kernel density function. On this basis, the entropy function is used to measure the spatial kernel density instability of all OD points in the study area:

宽松的取值范围

和步长

，迭代计算

，将

最小值对应的

设置为研究区域轨迹点空间邻域eps。基于此，以任一轨迹点

为圆心、eps为半径构建的圆形空间区域表示

的空间邻域，记为

。其中，设置

和

分别为1 m、200m和1 m。Where N represents the number of all OD points in the study area.

Loose value range

and step length

, iterative calculation

,Will

The minimum value corresponds to

Set to the spatial neighborhood eps of the trajectory points in the study area. Based on this, any trajectory point

The circular space area is represented by the center and eps is the radius.

The spatial neighborhood of

Among them, setting

and

They are 1 m , 200 m and 1 m respectively.

(5) Identification of dense neighborhoods of OD points. Under the assumption of spatial random distribution, the theoretical probability of k other OD points appearing in any spatial neighborhood SN is the Poisson distribution density function:

，

成为研究区域内空间密集邻域的概率计算为泊松分布的概率分布函数：Where N represents the number of all OD points in the study area; Area () represents the area calculation function; SN and ODA represent the spatial neighborhood of any trajectory point and the potential borrowing and returning area of a single vehicle, respectively. Given any OD point p _i and its spatial neighborhood

,

The probability of being a spatially dense neighbor within the study area is calculated as the probability distribution function of the Poisson distribution:

表示点p _i 空间邻域

内其他OD点的数量。若

大于给定具有统计意义的显著性水平α，则

标记为空间密集邻域

。其中，设置α=0.95。In the formula,

Represents the spatial neighborhood of point p _i

The number of other OD points within.

is greater than a given statistically significant level α , then

Marked as a spatially dense neighborhood

. Among them, α is set to 0.95.

(6) Identification of OD point tidal neighborhood. Given any spatially dense neighborhood SDN , the theoretical probability of k bicycle trip starting points appearing in the SDN is the binomial distribution density function:

表示

内起点数量，

表示空间密集邻域内轨迹点总数量，p表示目标区域内出行起点比例。给定任一OD点

及其空间密集邻域

，

成为研究区域内空间潮汐源邻域的可能性计算为：In the formula,

express

The number of internal starting points,

represents the total number of trajectory points in the spatially dense neighborhood, and p represents the proportion of travel starting points in the target area.

and its spatially dense neighborhood

,

The probability of becoming a neighbor of a spatial tidal source in the study area is calculated as:

表示

内起点数量，

表示

内轨迹点总数量。若

大于给定具有统计意义的显著性水平α，则

标记为空间潮汐源邻域

；若

小于给定具有统计意义的显著性水平1-α，则

标记为空间潮汐汇邻域

。其中，设置α=0.95。In the formula,

express

The number of internal starting points,

express

The total number of inner track points.

is greater than a given statistically significant level α , then

Marked as spatial tidal source neighborhood

;like

is less than the given statistically significant level 1- α , then

Marked as a spatial tidal sink neighborhood

. Among them, α is set to 0.95.

为种子点，定义

内其他起点

与

密度可达，将所有与

密度可达点聚合为空间潮汐源区域

，对于

内满足空间潮汐源邻域判别的其他点

，继续执行聚合操作更新

，直到

内所有满足空间潮汐源邻域判别的其他点全部访问完毕,重新选取目标区域内未被访问的其他空间潮汐源邻域重复上述过程，最终得到目标区域内提取的全部空间潮汐源区域表示为

。(7) Extraction of tidal area at OD point. Take any spatial tidal source neighborhood as

is the seed point, and defines

Other starting points

and

The density can be reached, all

Density-reachable points are aggregated into spatial tidal source regions

,for

Other points that meet the spatial tidal source neighborhood judgment

, continue to perform aggregate operation updates

,until

After all other points in the target area that meet the spatial tidal source neighborhood judgment are visited, other spatial tidal source neighborhoods that have not been visited in the target area are reselected to repeat the above process. Finally, all spatial tidal source areas extracted in the target area are expressed as

.

为种子点，定义

内其他终点

与

密度可达，将所有与

密度可达点聚合为空间潮汐汇区域

，基于以上空间扩展策略，如图4和图5所示，最终，研究区域内提取的全部空间潮汐汇区域表示为

。研究区域内提取的全部空间潮汐区域表示为潮汐源区域和潮汐汇区域的并集Similarly, any spatial tidal sink neighborhood

is the seed point, and defines

Other endpoints

and

The density can be reached, all

Density-reachable points are aggregated into spatial tidal sink areas

,Based on the above spatial expansion strategy, as shown in Figures 4 and 5, finally, all the spatial tidal sink areas extracted in the study area are expressed as

The total spatial tidal area extracted in the study area is represented as the union of the tidal source area and the tidal sink area.

。

.

It can be seen from this that the method of the present invention has the following advantages over the existing solutions:

(1) Existing methods based on spatial unit feature measurement require spatial unit division. At the same time, the tidal degree judgment relies on manual experience settings, which makes it difficult to accurately and reliably detect the tidal area of shared bicycle travel. The present invention effectively avoids the spatial division operation and adaptively extracts the fine spatial range of the shared bicycle travel tidal phenomenon based on a statistically significant significance test.

(2) The existing methods based on spatial trajectory clustering detection ignore the distribution characteristics of the starting and ending points of shared bicycles, and still use characteristics such as shared bicycle density or retention to set thresholds to distinguish tidal phenomena. This makes it difficult to truly reflect the degree of tidal imbalance in real urban space and is prone to misjudgment of the dynamic equilibrium area of borrowing and returning. The present invention performs tidal significance testing based on the real spatial distribution and actual relative quantity characteristics of shared bicycle OD points. It has rigorous mathematical and geographical interpretability and can reliably and stably extract shared bicycle travel tidal areas.

The units involved in the embodiments of the present invention may be implemented in software or hardware.

It should be understood that various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof.

The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (3)

1. The utility model provides a sharing bicycle trip tide area detection method based on self-adaptation density clustering which is characterized by comprising the following steps:

step 1, data cleaning is carried out on track points generated by a travel order of a shared bicycle in a target area, and then a potential borrowing and returning area of the shared bicycle is extracted;

step 2, designing a kernel density estimation function to construct a track point space density field, adaptively constructing a track point space neighborhood by measuring the instability of the space density field, and constructing a space density significance test function to extract a space density neighborhood according to the space density estimation function;

the step 2 specifically includes:

step 2.1, designing a kernel density estimation function, and giving a kernel density relative value approximate expression of any track point in a potential borrowing and returning area of a bicycle to obtain a spatial position of the track point as a track point spatial density field, wherein the kernel density relative value approximate expression is as follows

Wherein->

Representation dot->

And (4) point->

Euclidean space distance between->

Indicating all fulfilments within the investigation region +.>

Track points of>

Representing the bandwidth of the kernel density function;

step 2.2, measuring the spatial kernel density instability of the positions of the track points in the track point spatial density places by utilizing an entropy function, and constructing a track point spatial neighborhood;

measuring the spatial kernel density instability of all track points in the research area by using entropy function, and giving the bandwidth of kernel density function

Value range->

And step size->

Iterative calculation of entropy function->

Will->

Corresponding to the minimum value

Setting the space neighborhood length of the track point of the target area;

constructing a track point space neighborhood by taking any track point as a circle center and the space neighborhood length as a radius;

step 2.3, under the assumption of space random distribution, calculating theoretical probability distribution of k other track points in any space neighborhood, giving any track point and a corresponding space neighborhood thereof, calculating probability distribution value of the space neighborhood being a dense neighborhood, and taking the space neighborhood with the probability distribution value larger than a significance level value as the space dense neighborhood;

step 3, adaptively identifying the track point tide neighborhood according to the space dense neighborhood and the track point density characteristic significance test and extracting the shared bicycle trip tide region based on density reachable expansion;

the step 3 specifically includes:

step 3.1, calculating the occurrence in any one of the spatially dense neighborskTheoretical probability of individual sharing of a bicycle travel starting point; wherein the occurrence in any one of the spatially dense neighborskThe expression of the theoretical probability of each shared bicycle travel starting point is

In the method, in the process of the invention,

representing the number of start points in a spatially dense neighborhood, +.>

Represents the total number of trace points in a spatially dense neighborhood, +.>

Representing the travel starting point proportion in the target area;

step 3.2, calculating a spatial dense neighborhood corresponding to any track point according to the calculation result of the step 3.1 to become a probability value of a spatial tide source neighborhood in the target area, taking the spatial dense neighborhood with the probability value larger than the significance level value as the spatial tide source neighborhood, and taking the spatial dense neighborhood with the probability value smaller than the significance level value as the spatial tide sink neighborhood; wherein, the expression of the probability value that the space dense neighborhood corresponding to any track point becomes the space tide source neighborhood in the target area is that

In the method, in the process of the invention,

representation->

Number of internal origin->

Representation->

Total number of inner trace points>

Representation dot->

Is a spatially dense neighborhood of (1); />

Step 3.3, using any space tidal source neighborhood

For seed points, define->

Other starting points->

And->

The density is reachable, all are combined with +.>

The density reachable points are aggregated into a spatial tidal source area +.>

For->

Other points satisfying the space tidal source neighborhood discrimination in the interior continue to execute aggregation operation update +.>

Up to->

After all other points meeting the space tidal source neighborhood discrimination in the target area are accessed, the other space tidal source neighborhood which is not accessed in the target area is selected again, and the process is repeated, so that the whole space tidal source area extracted in the target area is finally obtained and expressed as +.>

；

Step 3.4, converging the neighborhood by any space tide

For seed points, define->

Other end point->

And->

The density is reachable, all are combined with +.>

The density reachable points are aggregated into a space tidal current area +.>

For->

Other points satisfying the space tide sink neighborhood discrimination in the interior continue to execute aggregation operation update +.>

Up to->

After all other points meeting the space tide sink neighborhood discrimination are accessed, the other space tide sink neighborhood which is not accessed in the target area is selected again, the process is repeated, and finally, all the space tide sink areas extracted in the target area are expressed as +.>

The total spatial tidal area extracted within the target area is represented as the union of the tidal Source area and the tidal sink area

。

2. The method according to claim 1, wherein the step 1 specifically comprises:

step 1.1, deleting track data outside the range of a target area, in a water system and in a building;

step 1.2, deleting track data with abnormal timestamp formats, arranging track points of each shared bicycle order in ascending order according to the timestamps, and keeping only one record for abnormal conditions of continuous unlocking and locking states;

step 1.3, for the same shared bicycle track point with repeated position coordinates, only one record with the earliest time stamp is reserved;

step 1.4, deleting a pair of track points with track point time stamp intervals outside a preset range;

and 1.5, generating a space buffer area according to a preset radius by each track point, and superposing and combining the space buffer areas generated by all track points to construct a bicycle potential borrowing and returning area.

3. The method of claim 2, wherein the entropy function has an expression of

In which, in the process,Nrepresenting the number of all track points in the target area;

occurrence in any one of the spatial neighborskThe theoretical probability distribution of each other trace point is expressed as

In which, in the process,Nrepresenting the number of all OD points in the investigation region, the OD points representing the trace points,Area() Representing the area calculation function,SNandODArespectively representing a space neighborhood of any track point and a bicycle potential borrowing and returning area;

the expression for calculating the probability distribution value of the spatial neighborhood being a dense neighborhood is as follows

In the method, in the process of the invention,

representation dot->

Spatial neighborhood->

The number of other trace points in the track. />

Images