A binary geographical flow spatio-temporal interaction anomaly detection method based on spatio-temporal autocorrelation

Abstract

The application discloses a binary geographical flow spatio-temporal interaction anomaly detection method based on spatio-temporal autocorrelation, which is based on binary geographical flow OD data of Baidu migration and Baidu index, takes municipal administrative units as research units, measures global spatio-temporal autocorrelation of population flow and information flow based on a spatio-temporal autocorrelation method, selects reasonable time threshold according to the global spatio-temporal autocorrelation result and applies the time threshold to local spatio-temporal autocorrelation, and thus identifies two abnormal interaction modes of high-value abnormal interaction and low-value abnormal interaction. The application makes up for the deficiency of global autocorrelation research of binary variables, introduces a change rate of attribute values of flow with time in a spatio-temporal weight of autocorrelation, makes the spatio-temporal weight more generalized, deduces a general local binary G-statistic expression, makes the algorithm rationality and applicability of binary flow higher, and provides a powerful tool for research of binary flow in cities and between cities, and optimization of resource utilization efficiency maximization and element resource allocation of multiple elements in cities.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of geographic information science, and particularly relates to a binary geographic flow spatio-temporal interaction anomaly detection method based on spatio-temporal autocorrelation. BACKGROUND

[0002] In recent years, with the rapid development of urbanization and the advancement of regional integration, the interaction between cities of various elements is more frequent and close. In the process of exchange of various resources between cities, adjacent cities gradually develop into important urban agglomerations. The spatio-temporal interaction of diversified resources between cities forms a complex spatio-temporal interaction mode, and human flow, material flow and information flow have become the three basic element flows for measuring the degree of connection between cities and the development of urban agglomerations. However, the high level of spatio-temporal agglomeration of some regional elements leads to waste of resources, and the low level of spatio-temporal agglomeration leads to low efficiency of resource utilization, which is difficult to play the synergistic effect of the overall development of urban agglomerations. Therefore, it is of great significance to detect the spatio-temporal interaction anomaly of two geographic flows based on the method of spatio-temporal autocorrelation for realizing fair and reasonable resource allocation and sustainable urban development.

[0003] The use of autocorrelation analysis is a means to detect abnormal distribution of data. Past researches have focused on the spatial autocorrelation of geographic flow, and have used one or more autocorrelation statistics to study the combination of the autocorrelation statistics, including Moran's I, Geary's C and G statistics (Getis-Ord General and Getis-Ord Gi). However, temporal autocorrelation also exists in geographic flow, and the spatio-temporal autocorrelation research of geographic flow is an extension of spatial autocorrelation research to the time dimension. Most of the past spatio-temporal autocorrelation researches of geographic flow are directed at univariate flow, and Moran's I is mostly used to correct the spatio-temporal weight, so as to reveal the spatio-temporal dynamic geographical distribution of flow. However, Moran's I can only determine whether the spatial agglomeration of similar attributes exists in the geographic flow, and cannot distinguish whether the agglomeration is high value agglomeration or low value agglomeration. The spatio-temporal weight value in the autocorrelation statistics is mostly binary, and using time interval as the only variable to construct time weight is too single, which may overestimate the influence of spatio-temporal adjacent flow on the center flow, and ignores the change of flow value with time. The spatio-temporal autocorrelation of multi-type geographic flow (such as binary flow) has not been fully explored, and the spatial autocorrelation of multi-type flow is mostly studied by local autocorrelation, lacking global autocorrelation research.

[0004] Therefore, it is urgent to construct a method that can simultaneously perform spatio-temporal autocorrelation analysis of binary geographic flow data at global and local scales, and detect abnormal interaction of spatio-temporal flow data between regions, so as to further reasonably allocate and optimize resource elements, and realize maximum resource utilization efficiency and sustainable development. SUMMARY

[0005] In order to overcome the above-mentioned deficiencies of the prior art, the application provides a binary geographical flow spatio-temporal interaction anomaly detection method based on spatio-temporal autocorrelation, which takes a municipal area in a single urban agglomeration as a research unit, uses a unary global G statistic to construct a binary global G statistic, and selects a reasonable time threshold through a global spatio-temporal autocorrelation result; then, an improved BiFlowAMOEBA (binary flow multi-direction optimal ecological zone) algorithm is used to perform spatio-temporal clustering of HH, HL, LH and LL modes under the time threshold, and the HH and LL clusters are taken as the interaction anomaly area of the binary flow.

[0006] According to an aspect of the application, a binary geographical flow spatio-temporal interaction anomaly detection method based on spatio-temporal autocorrelation is provided, comprising:

[0007] Step 1: obtaining a research area data set, including population flow and information flow group flow OD spatio-temporal data of a preset time period represented by a Baidu migration index and a Baidu index respectively in a municipal scale of a research area;

[0008] Step 2: constructing a binary global G statistic by multiplying two unary global G statistics;

[0009] Step 3: constructing a spatio-temporal weight using the spatial adjacency relationship of the area and the rate of change of the flow value over time, replacing the weight in the statistic of step 2 to obtain a binary flow global spatio-temporal autocorrelation result under different time thresholds;

[0010] Step 4: deriving a binary local G statistic formula in a weight generalization form using an L statistic and a unary local G statistic;

[0011] Step 5: combining the binary flow global spatio-temporal autocorrelation result in step 3, selecting a set time threshold and applying it to the improved BiFlowAMOEBA algorithm, which selects a seed flow under four clustering modes, uses a hierarchical clustering strategy under each mode, and merges areas under the assumption that there is no in-situ correlation, with the increase of the binary local G statistic as the target;

[0012] Step 6: obtaining spatio-temporal clustering results of HH, HL, LH and LL modes respectively, and taking the spatio-temporal clustering results of HH and LL as the interaction anomaly result of the binary flow.

[0013] As a further technical solution, obtaining the research area data set further comprises:

[0014] uniformly projecting and converting the spatial vector data to form a vector data set with consistent spatial coordinate systems;

[0015] The Baidu migration and Baidu index OD data are cleaned, the data of continuous time periods with more effective data are selected, the two kinds of geographic flow data are combined into the same city and time point, and the city ID and time ID fields are established, wherein the minimum time unit of the OD data is day.

[0016] As a further technical solution, a binary global G statistic is constructed in the form of multiplication of two unary global G statistics as follows:

[0017]

[0018] wherein w ij is a spatio-temporal weight constructed by spatial adjacency relationship and time correlation; x i and y i are two attribute values of the geographic flow standardized by z-score; x j and y j are two attribute values of the z-score standardized geographic flow in the i-th regional spatio-temporal neighborhood.

[0019] As a further technical solution, a spatio-temporal weight is constructed by regional spatial adjacency relationship and the rate of change of the flow value over time as follows:

[0020] w ij =sw ij ×tw ij

[0021] wherein w ij is a spatio-temporal weight; sw ij is a spatial weight defined by spatial adjacency relationship; tw ij is a temporal weight defined by the time distance between flows; the spatial weight and the temporal weight are respectively defined as follows:

[0022]

[0023] wherein tw i(t) and tw i(t) are the temporal weights corresponding to the binary attribute values of the flow; x j(t-Δt) and y j(t-Δt) represent two attribute values of the flow at t time; x θ and y ij are two attribute values of the adjacent flow at t-Δt time; Δt is a time interval; t ij is a time threshold; tw * is a temporal weight, which is the average value of the temporal weights corresponding to the binary attribute values of the flow.

[0024] As a further technical solution, a binary local G statistic formula of weight generalization form is derived by using L statistic and unary local G statistic as follows:

[0025]

[0026] wherein, and are the binary attribute values after Z-score standardization; w ij is the spatio-temporal weight of the ith flow and the jth flow within the spatio-temporal neighborhood; N * is the number of all flows; n * is the number of adjacent flows within the spatio-temporal neighborhood of the ith flow.

[0027] As a further technical solution, step 5 further comprises:

[0028] applying the time threshold value selected in step 3 and the binary local G statistics in step 4 to the binary flow multi-direction optimal ecological zone algorithm, using the strategy of hierarchical clustering to filter seed flows under HH, HL, LH and L four clustering modes, and if both attribute values of the binary flow after z-scores standardization are positive, it is HH mode; if both attribute values are negative, it is LL mode; if one attribute value is positive and the other is negative, it is HL or LH mode.

[0029] In the initialization stage, the value of the seed flow under each mode is 0, and the regional merging is performed with the increase of the binary local G statistics as the target, and the merging rule is as follows:

[0030]

[0031] wherein, is the number of binary flows in the merged region BFC i ∪BFC j , the Gain(BFC For two clustering clusters BFC i and BFC j , if is the maximum value, and Gain(BFC i ∪BFC j ) > 0, it is considered that the two clustering clusters can be merged into a new clustering cluster.

[0032] As a further technical solution, the method further comprises: outputting the abnormal interaction recognition result and showing the spatio-temporal distribution thereof.

[0033] According to an aspect of the present application, a binary geographical flow spatio-temporal interaction anomaly detection device based on spatio-temporal autocorrelation is provided, comprising:

[0034] ​The first main module is used for acquiring a research area data set, including group flow OD spatio-temporal data of population flow and information flow represented by a baidu migration index and a baidu index respectively in a preset time period in a city-level scale of the research area;

[0035] The second main module is used for constructing a binary global G statistic in a form of multiplication of two unary global G statistics;

[0036] The third main module is used for constructing a spatio-temporal weight by using a regional spatial adjacency relationship and a change rate of the value of the flow with time, replacing the weight in the statistic of the second main module, and obtaining a binary flow global spatio-temporal autocorrelation result under different time thresholds;

[0037] The fourth main module is used for deducing a binary local G statistic formula of a weight generalization form by using an L statistic and a unary local G statistic;

[0038] The fifth main module is used for combining the binary flow global spatio-temporal autocorrelation result in the third main module, selecting a set time threshold, and applying the improved BiFlowAMOEBA algorithm, which selects a seed flow in four clustering modes, applies a hierarchical clustering strategy in each mode, and performs regional merging under the assumption that there is no in-situ correlation, to increase the binary local G statistic.

[0039] The sixth main module is used for obtaining spatio-temporal clustering results of HH, HL, LH and LL four modes respectively, and taking the spatio-temporal clustering results of HH and LL as binary flow interactive anomaly results.

[0040] According to an aspect of the present application, a kind of binary geographic flow spatio-temporal interactive anomaly detection system based on spatio-temporal autocorrelation is provided, including memory and processor;The memory stores the program instruction executed by the processor, and the processor calls the program instruction to execute the steps of the kind of binary geographic flow spatio-temporal interactive anomaly detection method based on spatio-temporal autocorrelation.

[0041] According to an aspect of the present application, a kind of binary geographic flow spatio-temporal interactive anomaly detection system based on spatio-temporal autocorrelation is provided, including memory and processor;The memory stores the program instruction executed by the processor, and the processor calls the program instruction to execute the steps of the kind of binary geographic flow spatio-temporal interactive anomaly detection method based on spatio-temporal autocorrelation.

[0042] Compared with prior art, the beneficial effects of the present application are that:

[0043] 1. The application is based on binary geographic flow OD data in the aspect of binary geographic flow spatio-temporal interaction anomaly detection, taking municipal administrative units as research units, and measuring the global spatio-temporal autocorrelation of population flow and information flow based on the method of spatio-temporal autocorrelation, and selecting reasonable time threshold according to the global spatio-temporal autocorrelation result for local spatio-temporal autocorrelation, so as to identify the two abnormal interaction modes of HH and LL.

[0044] 2. The application introduces the time dimension into the autocorrelation research of binary flow, and expands the spatio-temporal autocorrelation research method of unary flow; the global binary G statistic is developed and constructed through unary global G statistic, which makes up for the deficiency of global autocorrelation research of binary variables; the concept of change rate of flow attribute value with time is introduced into the autocorrelation spatio-temporal weight, which makes the spatio-temporal weight more general, and deduces the general expression of local binary G statistic, so that the rationality and applicability of the algorithm of binary flow are higher, which provides a powerful tool for the research of binary flow in cities and between cities, and the optimization of urban multi-element resource utilization efficiency and element resource allocation. BRIEF DESCRIPTION OF DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0046] Figure 1 The flow chart of the binary geographic flow spatio-temporal interaction anomaly detection method based on spatio-temporal autocorrelation provided by the embodiment of the present application.

[0047] Figure 2 The flow chart of the BiFlowAMOEBA algorithm based on global autocorrelation provided by the embodiment of the present application.

[0048] Figure 3 And Figure 4 Both are spatio-temporal distribution maps of abnormal interaction flow with high population flow and information flow provided by the embodiment of the present application.

[0049] Figure 5 The structural schematic diagram of the binary geographic flow spatio-temporal interaction anomaly detection equipment based on spatio-temporal autocorrelation provided by the embodiment of the present application. DETAILED DESCRIPTION

[0050] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work are within the protection scope of the present application. In addition, the technical features in each embodiment or in a single embodiment provided by the present application can be combined with each other at will to form new technical solutions, and such combination is not restricted by the order of steps and / or structure composition mode, but should be based on the implementation by those of ordinary skill in the art. When the combination of technical solutions appears contradictory or unimplementable, it should be considered that the combination of technical solutions does not exist, and is not within the protection scope required by the present application.

[0051] The present application provides a binary geographical flow spatio-temporal interaction anomaly detection method based on spatio-temporal autocorrelation. The method takes the municipal area within a single urban agglomeration as the research unit, uses a unary global G statistic to construct a binary global G statistic, and selects a reasonable time threshold through the global spatio-temporal autocorrelation result. Then, the improved BiFlowAMOEBA (Binary Flow Multi-Direction Optimal Ecological Zone) algorithm is used to perform spatio-temporal clustering of HH, HL, LH and LL modes under the time threshold, and the HH and LL clusters are taken as the interaction anomaly areas of the binary flow. Finally, on this basis, optimization measures for reasonable resource allocation can be proposed. Specifically, the following steps are included:

[0052] Step 1: Collect and process the data set of the study area, including the population flow and information flow group flow OD spatio-temporal data within 1 week represented by the Baidu migration index and the Baidu index respectively within the municipal scale of the study area. The OD data has a minimum time unit of day.

[0053] As preferred, the data processing in step 1 mainly includes: (1) uniform projection conversion processing of spatial vector data to form a vector data set with consistent spatial coordinate system; (2) statistics of the binary flow data (population flow and information flow) between the municipal areas in the study area within 7 days.

[0054] Step 2: Construct a binary global G statistic by multiplying two unary global G statistics.

[0055] As preferred, the construction of the binary global G statistic in step 2 can be obtained from the following formula:

[0056]

[0057] where w ij is a spatio-temporal weight constructed by spatial adjacency relationship and temporal correlation; x iwith y i are two attribute values of the z-score standardized geographic flow; x j and y j are two attribute values of the z-score standardized geographic flow within the i-th regional spatio-temporal neighborhood.

[0058] Step 3: Construct the spatio-temporal weights by using the spatial adjacency relationship and the rate of change of the flow values over time, replace the weights in the original statistics, and obtain the global spatio-temporal autocorrelation results of the binary flow under different time thresholds.

[0059] As a preference, the spatio-temporal weights in Step 3 are constructed by multiplying the spatial weights and the temporal weights, as shown in the following formula:

[0060] w ij = sw ij × tw ij

[0061] where w ij is the spatio-temporal weight; sw ij is the spatial weight defined by the spatial adjacency relationship; and tw ij is the temporal weight defined by the temporal distance between the flows.

[0062] (1) For the origin-destination (OD) point region of the flow, the spatial weight is based on the Queen adjacency rule and is defined as follows:

[0063]

[0064] (2) For the construction of the temporal weight, it is believed here that the flow at time t is simultaneously affected by the adjacent flow at time t-Δt within a certain time threshold and the current time t, because each flow has a time stamp and a binary attribute value, and the attribute values of the flows at different time points are different, so the temporal distance between the flows can be calculated by using the time difference and the rate of change of the flow over time, as shown in the following formula:

[0065]

[0066]

[0067] wherein and are the temporal weights corresponding to the binary attribute values of the flow; x i(t) and y i(t) represent the two attribute values of the flow at time t; x j(t-Δt) and y j(t-Δt) are the two attribute values of the adjacent flow at time t-Δt; Δt is the time interval; t θ is the time threshold; and tw ijis the time weight, which is the average value of the time weight corresponding to the binary attribute value of the flow.

[0068] Step 4: Use the L statistic and the univariate local G statistic to derive the weighted generalized form of the bivariate local G statistic formula.

[0069] Preferably, the binary local G statistic formula in step 4 is derived as follows:

[0070] (1) The bivariate local L statistic and the univariate local G statistic are used to construct the bivariate local G statistic. The bivariate local L statistic is used to measure the bivariate spatial dependence, and its formula is as follows:

[0071]

[0072] When the weight matrix is ​​row-normalized, its formula can be simplified as follows:

[0073]

[0074] in, and Represents the local weighted z-scores values ​​of variables x and y respectively; S X and S Y is the standard deviation of variables x and y.

[0075] (2) After the weight matrix is ​​row-normalized, the formula for the univariate local G statistic is as follows:

[0076]

[0077] (3) Receive Inspired by G*, the binary local G statistic should have both the spatial correlation of the two variables and the number of local units to represent the local spatial correlation, and then be extended to the time dimension to show the spatiotemporal correlation. When the spatiotemporal weight matrix is ​​row-normalized, its formula is as follows:

[0078]

[0079] in, and are the binary attribute values ​​after Z-score standardization; w ij is the spatiotemporal weight of the i-th flow and the j-th flow in the spatiotemporal neighborhood; N * is the number of all flows; n * is the number of adjacent flows in the spatiotemporal neighborhood of the i-th flow.

[0080] Step 5: Combined with the global spatiotemporal autocorrelation results in step 3, select a reasonable time threshold and apply it to the improved BiFlowAMOEBA algorithm, such asFigure 2 As shown, the algorithm selects seed flows in four clustering modes, applies the strategy of hierarchical clustering in each mode, and merges regions with the goal of increasing the binary local G statistic in the absence of the assumption of in-situ correlation.

[0081] As preferred, the strategy of hierarchical clustering is applied in step 5 in each clustering mode, and the initialization stage sets the seed flow in each mode to 0, and merges regions with the goal of increasing the binary local G statistic, and the merging rule is shown as follows:

[0082]

[0083] wherein, is the number of binary flows in the merged region BFC i ∪BFC j , the number of binary flows between each two adjacent clustering clusters is calculated. For two clustering clusters BFC i and BFC j , if is the maximum value, and Gain(BFC i ∪BFC j ) > 0, it is considered that the two clustering clusters can be merged into a new clustering cluster.

[0084] Step 6: Obtain the spatio-temporal clustering results of HH, HL, LH and LL modes respectively, and take the spatio-temporal clustering results of HH and LL as the interaction anomaly results of binary flows. Output the abnormal interaction recognition results and show the spatio-temporal distribution thereof.

[0085] As a preferred embodiment, please refer to Figure 1 , the binary geographic flow spatio-temporal interaction anomaly detection method based on spatio-temporal autocorrelation mentioned in the application comprises the following steps:

[0086] Step 1: Take the OD data of Baidu migration and Baidu index (PC terminal plus mobile terminal) in the Guangdong-Hong Kong-Macao Greater Bay Area as the representation of population flow and information flow, and the data set contains city names (due to data availability, Hong Kong and Macao two special administrative regions are excluded), index value representing flow value, and time. The data processing mainly includes: (1) uniform projection conversion processing is performed on the spatial vector data to form a vector data set with consistent spatial coordinate system; (2) the Baidu migration and Baidu index OD data are cleaned, and the data of a continuous week (October 1, 2021 to October 7, 2021) with more valid data (not 0) are selected, the two kinds of geographic flow data are merged into the same city and time point, and the city ID and time ID fields are established.

[0087] Step 2: Calculate the spatio-temporal weight of each flow with its neighbors within the spatio-temporal neighborhood at different time thresholds without considering the original correlation. The temporal weight is applied to the constructed global autocorrelation expression, which is defined as follows:

[0088]

[0089] where d represents a certain spatio-temporal distance range, w ij is the spatio-temporal weight constructed using spatial adjacency and temporal correlation; x i and y i are two attribute values of the geographical flow standardized by z-score; x j and y j are two attribute values of the z-score standardized geographical flow within the spatio-temporal neighborhood of the ith region.

[0090] In this embodiment, the global spatio-temporal autocorrelation results are shown in Table 1.

[0091] Table 1 Global spatio-temporal autocorrelation results at different time thresholds

[0092]

[0093] In Table 1, *, **, and *** indicate that BG(d) is significant at the 0.05, 0.01, and 0.005 levels, respectively.

[0094] Step 3: According to the spatio-temporal autocorrelation results obtained in Step 2 and the time threshold, if the purpose of the analysis is to identify short-term anomalies or fluctuations, a smaller time threshold (such as days) may be more appropriate; if the purpose is to capture long-term trends, a larger time threshold (such as weeks) should be selected. Here, the time threshold of t θ = 2 is selected to capture short-term abnormal interactions.

[0095] Step 4: Based on the derived local autocorrelation G statistic G and G * , the formula is as follows:

[0096]

[0097] where x and y are the binary attribute values after z-score standardization; w ij is the spatio-temporal weight of the ith flow with the jth flow within the spatio-temporal neighborhood; N * is the number of all flows; n * is the number of adjacent flows within the spatio-temporal neighborhood of the ith flow.

[0098] Step 5: The time threshold selected in Step 3 and the BG in Step 4 * The BiFlowAMOEBA algorithm is applied to the hierarchical clustering strategy to screen seed flows under four clustering modes (HH, HL, LH, and LL). If both attribute values of the binary flow after z-score standardization are positive, it is HH mode; if both attribute values are negative, it is LL mode; if one attribute value is positive and the other is negative, it is HL or LH mode.

[0099] In the initialization phase, the seed flow under each mode is initialized to 0, and the regional merging is performed with the goal of increasing the binary local G statistic. The merging rules are as follows:

[0100]

[0101] wherein, is the number of binary flows in the merged region BFC i ∪BFC j . The G statistic of each two adjacent clustering clusters is calculated. For two clustering clusters BFC i and BFC j , if is the maximum value, and Gain(BFC i ∪BFC j ) > 0, then the two clustering clusters can be merged into a new clustering cluster.

[0102] Step 6: Output the abnormal interaction recognition result and show its spatiotemporal distribution.

[0103] Based on the BiFlowAMOEBA algorithm, Figure 3 the abnormal interaction mode with high population flow and information flow is shown. In this mode, only one clustering cluster is finally detected, and the area with high flow of the two types can be found in Guangzhou, Foshan, Dongguan, Shenzhen, and Huizhou. On October 2, the Shenzhen-Dongguan and Shenzhen-Huizhou flows are higher than those on October 1, as well as the Shenzhen-Huizhou, Shenzhen-Guangzhou, Dongguan-Guangzhou, Dongguan-Huizhou, and Guangzhou-Foshan flows. This indicates that during the first two days of the National Day holiday, there is a strong population and information flow, and the Guangzhou-Foshan urbanization synergy effect is obvious. The population in Shenzhen mainly flows out, and they pay more attention to information in neighboring cities. Compared with other dates, the first two days of the National Day holiday are the period of abnormally high population and information flow, and the outflow effect in Shenzhen is more obvious.

[0104] Figure 4The abnormal interaction mode with high population flow and information flow is shown, and three cluster clusters are finally detected in the mode, and the lower abnormal interaction has a large time span within a week. The first cluster cluster includes Foshan-Dongguan, Zhaoqing-Dongguan, Zhaoqing-Huizhou, Jiangmen-Shenzhen, Jiangmen-Huizhou, Zhaoqing-Shenzhen and Jiangmen-Dongguan on October 2, Zhaoqing-Huizhou, Jiangmen-Huizhou and Zhuhai-Huizhou on October 1. The origin and destination of most of these binary flows are not adjacent counties and cities, and the distance is far. The second cluster cluster includes Foshan-Zhongshan, Zhaoqing-Jiangmen, Zhaoqing-Zhongshan, Zhaoqing-Zhuhai, Jiangmen-Zhuhai and Zhuhai-Jiangmen on October 7, Zhaoqing-Zhuhai, Zhaoqing-Jiangmen and Zhaoqing-Zhongshan on October 6. This shows that compared with other cities, the population flow out of Zhaoqing City and the degree of attention to other cities are lower in the last two days of the National Day holiday. The third cluster cluster includes Huizhou-Zhongshan, Huizhou-Jiangmen, Huizhou-Zhuhai and Huizhou-Foshan on October 5, and Huizhou-Jiangmen, Huizhou-Zhuhai and Huizhou-Zhaoqing on October 4. This shows that compared with other cities, the population flow out of Huizhou City and the degree of attention to other cities are lower during the National Day holiday. The spatio-temporal adjacency of the two abnormal interaction modes can help the urban planning department to improve the configuration and optimization of transportation infrastructure and improve the utilization efficiency of resources.

[0105] The implementation basis of each embodiment of the present application is realized by the programmed processing of a device with processor function. Therefore, in engineering practice, the technical solutions and functions of each embodiment of the present application are packaged into various modules. Based on this actual situation, on the basis of each of the above embodiments, an embodiment of the present application provides a binary geographical flow spatio-temporal interaction anomaly detection device based on spatio-temporal autocorrelation, which is used to execute a binary geographical flow spatio-temporal interaction anomaly detection method based on spatio-temporal autocorrelation in the above method embodiment.

[0106] Reference is made to Figure 5The device comprises: a first main module for acquiring a research area data set, including group flow OD spatio-temporal data of population flow and information flow represented by a baidu migration index and a baidu index respectively in a preset time period in a city-level scale within a research area; a second main module for constructing a binary global G statistic quantity in the form of multiplication of two unary global G statistic quantities; a third main module for constructing a spatio-temporal weight by using a regional spatial adjacency relationship and a change rate of the value of the flow with time, replacing the weight in the second main module statistic quantity, and obtaining a binary flow global spatio-temporal autocorrelation result under different time thresholds; a fourth main module for deducing a binary local G statistic quantity formula of the weight generalization form by using an L statistic quantity and a unary local G statistic quantity; a fifth main module for combining the binary flow global spatio-temporal autocorrelation result in the third main module, selecting a set time threshold, and applying it to the improved BiFlowAMOEBA algorithm, which selects a seed flow in four clustering modes, applies a hierarchical clustering strategy in each mode, and performs regional merging under the assumption that there is no in-situ correlation, with the increase of the binary local G statistic quantity as the target; and a sixth main module for obtaining spatio-temporal clustering results of HH, HL, LH and LL four modes respectively, and taking the spatio-temporal clustering results of HH and LL as interaction anomaly results of the binary flow.

[0107] The device provided by the embodiment of the application is based on spatio-temporal autocorrelation of binary geographic flow, and the spatio-temporal autocorrelation of multiple types of geographic flow is not fully explored. Figure 5 The device provided by the embodiment of the application is based on spatio-temporal autocorrelation of binary geographic flow, and the spatio-temporal autocorrelation of multiple types of geographic flow is not fully explored.

[0108] It should be noted that the device provided by the embodiment of the application is used to implement the method in the method embodiment and the method in other method embodiments provided by the application, and the difference is only that the corresponding functional modules are set, the principle is basically the same as that of the above-mentioned device embodiments provided by the application, as long as the person skilled in the art improves the device in the above-mentioned device embodiments on the basis of the above-mentioned device embodiments, refers to the specific technical solutions in other method embodiments, obtains the corresponding technical means by combining technical features, and the technical solutions composed of these technical means, as long as the technical solutions have practicality, the device in the above-mentioned device embodiments is improved to obtain the corresponding device class embodiment, which is used to implement the method in other method class embodiments.

[0109] Based on the same inventive concept as the above embodiments, the embodiments of the present application also provide a binary geographical flow spatio-temporal interaction anomaly detection system based on spatio-temporal autocorrelation, comprising a memory and a processor; the memory stores program instructions executed by the processor, and the processor calls the program instructions to execute the steps of the binary geographical flow spatio-temporal interaction anomaly detection method based on spatio-temporal autocorrelation.

[0110] Based on the same inventive concept as the above embodiments, the embodiments of the present application also provide a non-transitory computer readable storage medium storing computer instructions, which make the computer execute the steps of the binary geographical flow spatio-temporal interaction anomaly detection method based on spatio-temporal autocorrelation.

[0111] In summary of the above embodiments, the present application measures the global spatio-temporal autocorrelation of population flow and information flow based on the binary geographical flow migration and the baidu index OD data, taking the municipal administrative unit as the research unit, based on the method of spatio-temporal autocorrelation. According to the global spatio-temporal autocorrelation result, a reasonable time threshold is selected and applied to the local spatio-temporal autocorrelation, so as to identify two abnormal interaction modes of high-value abnormal interaction and low-value abnormal interaction. The present application develops and constructs the global binary G statistic based on the unary global G statistic, which makes up for the deficiency of the global autocorrelation research of binary variables. The concept of the change rate of the flow attribute value with time is introduced into the autocorrelation spatio-temporal weight, which makes the spatio-temporal weight more general, and deduces the general expression of the local binary G statistic, so that the rationality and applicability of the binary flow algorithm are higher, which provides a powerful tool for the research of binary flow in cities and between cities, and the maximization of urban multi-element resource utilization efficiency and the optimization of element resource allocation.

[0112] It should be understood that parts not elaborated in the specification can be regarded as prior art.

[0113] Finally, it should be pointed out that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present application.

Claims

1. A method for detecting spatiotemporal interaction anomalies in binary geographic flows based on spatiotemporal autocorrelation, characterized by: include: Step 1: Obtain the study area dataset, including the group flow OD spatiotemporal data of population flow and information flow within a preset period at the city level in the study area; Step 2: Construct a binary global G statistic by multiplying two unary global G statistics; Step 3: Use the regional spatial adjacency relationship and the rate of change of the flow attribute value over time to construct the spatiotemporal weight, replace the weight in the statistic in step 2, and obtain the global spatiotemporal autocorrelation results of binary geographic flow under different time thresholds; among them, the spatiotemporal weight constructed by using the regional spatial adjacency relationship and the rate of change of the flow attribute value over time is: In ij =sw ij ×tw ij Among them, w ij is the spatiotemporal weight; sw ij is the spatial weight defined by the spatial adjacency; tw ij is the time weight defined by the time distance between flows; the spatial weight and time weight are defined as follows: in, and are the time weights corresponding to the binary attribute values ​​of the flow; x i(t) with y i(t) Represents two attribute values ​​of the flow at time t; x j(t-Δt) and y j(t-Δt) are the two attribute values ​​of adjacent flows at time t-Δt; Δt is the time interval; t θ is the time threshold; tw ij is the time weight, which is the average value of the time weight corresponding to the binary attribute value of the flow; Step 4: Use the L statistic and the univariate local G statistic to derive the weighted generalized form of the bivariate local G statistic formula; Step 5: Combined with the global spatiotemporal autocorrelation results of the binary geographic flow in step 3, a set time threshold is selected and applied to the improved BiFlowAMOEBA algorithm. This algorithm selects seed flows under four clustering modes and uses a hierarchical clustering strategy in each mode. Under the assumption that there is no in-situ correlation, regions are merged with the goal of increasing the binary local G statistic. Step 6: Obtain the spatiotemporal clustering results of the four modes HH, HL, LH, and LL respectively, and use the spatiotemporal clustering results of HH and LL as the interaction anomaly results of the binary geographic flow.

2. According to claim 1, a method for detecting spatiotemporal anomalies in binary geographic streams based on spatiotemporal autocorrelation is characterized by: Get the study area dataset, including: Perform unified projection transformation on spatial vector data to form a vector data set with consistent spatial coordinate system; The OD data is cleaned, and data from continuous periods with more valid data are selected. The two geographic flow data are merged into the same city and time point, and the city ID and time ID fields are established. The minimum time unit of the OD data is day.

3. The method for detecting spatiotemporal anomalies in binary geographic streams based on spatiotemporal autocorrelation according to claim 1 is characterized in that: The binary global G statistic is constructed by multiplying two unary global G statistics: Among them, w ij It is the spatiotemporal weight constructed by using spatial adjacency and temporal correlation; x i with y i are two attribute values ​​of geographic flow standardized using z-score; x j and y j are the two attribute values ​​of the z-score-normalized geographic flow within the spatiotemporal neighborhood of the i-th region.

4. The method for detecting spatiotemporal anomalies in binary geographic streams based on spatiotemporal autocorrelation according to claim 3 is characterized in that: The formula of the binary local G statistic in the weight generalized form is derived using the L statistic and the univariate local G statistic: in, and are the binary attribute values ​​after Z-score standardization; w ij is the spatiotemporal weight of the i-th flow and the j-th flow in the spatiotemporal neighborhood; N * is the number of all flows; n * is the number of adjacent flows in the spatiotemporal neighborhood of the i-th flow.

5. The method for detecting spatiotemporal anomalies in binary geographic streams based on spatiotemporal autocorrelation according to claim 4 is characterized in that: Step 5 also includes: The time threshold selected in step 3 and the binary local G statistic in step 4 are combined Applied to the multi-directional optimal ecological zone algorithm for binary geographical flows, the hierarchical clustering strategy was used to screen seed flows under four clustering modes: HH, HL, LH, and L. If both attribute values ​​of the binary geographical flow after z-score standardization were positive, it was the HH mode; if both attribute values ​​were negative, it was the LL mode; if one attribute value was positive and the other was negative, it was the HL or LH mode. The initialization phase will be the seed flow of each mode The value is 0, and regions are merged with the goal of increasing the binary local G statistic. The merging rules are as follows: in, It is in the merge area BFC i ∪BFC j The number of binary geographic flows in the cluster is calculated by calculating the For two clusters BFC i and BFC j ,if is the maximum value, and Gain(BFC i ∪BFC j )>0, it is considered that the two clusters can be merged into a new cluster.

6. The method for detecting spatiotemporal anomalies in binary geographic streams based on spatiotemporal autocorrelation according to claim 1, characterized in that: The method further includes: outputting abnormal interaction recognition results and displaying their spatiotemporal distribution.

7. A device for detecting spatiotemporal interaction anomalies of binary geographic streams based on spatiotemporal autocorrelation, used to implement the method according to any one of claims 1 to 6, characterized in that: include: The first main module is used to obtain the dataset of the study area, including the group flow OD spatiotemporal data of population flow and information flow within a preset time period at the city level in the study area; The second main module is used to construct a binary global G statistic by multiplying two unary global G statistics; The third main module is used to construct spatiotemporal weights using the regional spatial adjacency relationship and the rate of change of the flow value over time, replacing the weights in the statistics of the second main module to obtain the global spatiotemporal autocorrelation results of binary geographic flows under different time thresholds; The fourth main module is used to derive a weight-generalized binary local G statistic formula using the L statistic and the univariate local G statistic; The fifth main module is used to combine the global spatiotemporal autocorrelation results of the binary geographic flow in the third main module, select a set time threshold and apply it to the improved BiFlowAMOEBA algorithm. This algorithm selects seed flows under four clustering modes and uses a hierarchical clustering strategy in each mode. Under the assumption that there is no in-situ correlation, the algorithm merges regions with the goal of increasing the binary local G statistic. The sixth main module is used to obtain the spatiotemporal clustering results of the four modes of HH, HL, LH and LL respectively, and the spatiotemporal clustering results of HH and LL are used as the interaction anomaly results of binary geographic flows.

8. A spatiotemporal interaction anomaly detection system for binary geographic flows based on spatiotemporal autocorrelation, characterized by: It includes a memory and a processor; the memory stores program instructions executed by the processor, and the processor calls the program instructions to execute the steps of a binary geographic flow spatiotemporal interaction anomaly detection method based on spatiotemporal autocorrelation as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, which enable the computer to execute the steps of the method for detecting spatiotemporal interaction anomalies of binary geographic flows based on spatiotemporal autocorrelation as described in any one of claims 1 to 6.

Images