A subway transfer data analysis system and method based on big data

Abstract

The application discloses a subway transfer data analysis system and method based on big data, and belongs to the technical field of data analysis. A station transfer path is drawn, and station transfer paths passing through the same escalator are integrated; the importance of the escalator is analyzed, time segments are divided, and people flow data in the time segments passing through the escalator is monitored; test operation data of the escalator is adjusted, and a people flow dynamic data set is updated and marked; in different time cycle periods, the same test operation data is used for people flow dispersion testing, the average distance between test data clusters is obtained, the centroid position is dynamically changed, the maximum intracorpus distance is analyzed and obtained, and the centroid position when the intracorpus distance is maximum is analyzed and obtained; the test operation speed of the escalator during different time segments is intelligently analyzed and optimized; furthermore, the effectiveness of the subway station transfer path for people flow dispersion can be evaluated, and the operation speed of the escalator in different time segments can be intelligently optimized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis, in particular to a subway transfer data analysis system and method based on big data. BACKGROUND

[0002] With the acceleration of urbanization, the urban population is increasing, and the traffic demand is also growing. Subway, as an efficient and convenient means of transportation, can meet the travel demand of a large number of people, especially during peak hours, the operation capacity of subway is more obvious.

[0003] In the prior art, the change of passenger flow is predicted intelligently by collecting and analyzing passenger travel data, and the train operation plan and transfer scheme are automatically adjusted to meet the travel demand of passengers. For some large subway stations, there are multiple subway lines crossing and transferring, and the station is still crowded. The control method for crowd evacuation is still insufficient. SUMMARY

[0004] The purpose of the present application is to provide a subway transfer data analysis system and method based on big data to solve the problems raised in the background.

[0005] In order to solve the above technical problems, the present application provides the following technical scheme:

[0006] A subway transfer data analysis system based on big data, the system comprises: a transfer path processing module, a data acquisition module, a data cluster analysis module and an intelligent evaluation module;

[0007] The transfer path processing module draws the transfer path in the station according to the topographic map in the station, generates a transfer path set, and integrates the transfer paths in the station that pass through the same escalator based on the escalator to generate a path aggregation point set.

[0008] The data acquisition module is used to construct an escalator evacuation importance analysis model to calculate the evacuation importance of the escalator, divide time segments, monitor the passenger flow data passing through the escalator in the time segment, and take the evacuation importance of the escalator as the weight coefficient of the passenger flow data to generate a passenger flow dynamic data set.

[0009] The data cluster analysis module is used to adjust the test running data of the escalator, update and mark the passenger flow dynamic data set to generate a test data cluster, and use the same test running data to test the passenger flow evacuation in different time cycle periods to obtain the inter-cluster average distance between the test data clusters and generate an evaluation data cluster.

[0010] The intelligent evaluation module dynamically changes the centroid position based on the evaluation data cluster, analyzes the maximum intra-cluster distance, and the centroid position when the intra-cluster distance is maximum; in different time cycles, with the adjustment of the escalator test running data, intelligently analyzes and optimizes the test running speed of the escalator during different time segments.

[0011] Further, the transfer path processing module further comprises a path generation unit and a path aggregation unit.

[0012] The path generation unit is configured to compile a station metro transfer station waiting floor area code, and record any one station metro transfer station waiting floor area code as I i , wherein i represents the sequence number of the station metro transfer station waiting floor area; draw a station transfer path, the station transfer path is all paths between one station metro transfer station waiting floor area and another station metro transfer station waiting floor area, and each station transfer path is a path passing through an escalator; obtain all station transfer paths drawn between station metro transfer station waiting floor area I i and station metro transfer station waiting floor area I j , and generate a transfer path set, denoted as TP(I i →I j ) = {tp x | x ∈ [1, a]}, wherein j represents the sequence number of the station metro transfer station waiting floor area, and i ≠ j, tp x represents the xth station transfer path, and a represents the sequence number of the station transfer path.

[0013] The path aggregation unit is configured to uniformly code all escalators existing in the station transfer path, and record any one escalator as E e , wherein e represents the code number of the escalator; according to the station transfer path, integrate the station transfer paths passing through the same escalator E e , and generate a path aggregation point set, denoted as CP(E e ) = {tp x | x ∈ [1, b]}, wherein b represents the sequence number of the station transfer path.

[0014] Further, the data collection module further comprises a channeling analysis unit and a dynamic perception unit.

[0015] The channeling analysis unit is configured to construct an escalator channeling importance analysis model based on the transfer path set and the path aggregation point set, and calculate the channeling importance of the escalator, and the specific calculation formula is as follows:

[0016]

[0017] , wherein EI(Ee ) represents the importance of the evacuation of the escalator E e , NUM[TP(I i →I j )] represents the total number of in-station transfer paths included in the transfer path set TP(I i →I j ), NUM[CP(E e )∩TP(I i →I j )] represents the total number of in-station transfer paths included in the intersection set between the path aggregation point set CP(E e ) and the transfer path set TP(I i →I j ), and N represents the maximum value of the sequence number of the in-station transfer path;

[0018] The dynamic perception unit is configured to divide a day into k continuous time segments in a time cycle period of one day, and any one time segment is recorded as T u , wherein u represents the sequence number of the time segment and u≤k; a fixed flow observation point is arranged on each escalator, and the flow data passing through each escalator in the time segment T u is observed in real time, and the flow data refers to the number of pedestrians passing through the escalator in each time segment; the flow dynamic data set is generated according to the in-station transfer path, and is recorded as FD(T u |t)={EI(E e )×fd(E e )|e∈[1,c]},wherein FD(T u |t) represents the corresponding flow dynamic data set generated in the time segment T u in the tth time cycle period, fd(E e ) represents the flow data passing through the escalator E u in the time segment T e in the tth time cycle period, and c represents the total number of escalators existing in the in-station transfer path.

[0019] Further, the data cluster analysis module further comprises a data updating unit and a data correlation processing unit;

[0020] The data updating unit is configured to adjust the test running speed of the escalator and make the test running speeds of the escalators the same, record the dynamic changes of the flow dynamic data set, and update and mark the flow dynamic data set; the running speed obtained by adjusting the escalator for the yth time is recorded as V y , and the updated and marked flow dynamic data set is recorded as test data cluster FD(T u |t,V y);

[0021] The data correlation processing unit obtains a test data cluster FD(T u |t, V y ) in a time cycle t and at a test running speed V y , performs a crowd evacuation test, and obtains an inter-cluster average distance between the test data cluster FD(T u |t, V y ) and the test data cluster FD(T u+1 |t, V y ), denoted as CS[(T u →T u+1 )|t, V y ], collects the inter-cluster average distance, and generates an evaluation data cluster, denoted as AD(t, V y )={CS[(T u →T u+1 )|t, V y ]|u∈[1, k-1]}.

[0022] Further, the intelligent evaluation module further comprises a data screening unit and an evaluation optimization unit.

[0023] The data screening unit initializes a centroid of the evaluation data cluster AD(t, V y ) based on the evaluation data cluster, denoted as u:u→u+1, and the cluster element corresponding to the centroid position is CS[(T u →T u+1 )|t, V y ], calculates an intra-cluster distance of the evaluation data cluster AD(t, V y ) based on the centroid u:u→u+1, adjusts the centroid position, and iteratively calculates the intra-cluster distance until the intra-cluster distance is maximum, and the iteration calculation stops, the maximum intra-cluster distance is denoted as \max[AD(t, V y )], and the centroid at the time when the intra-cluster distance is maximum is denoted as u max .

[0024] The evaluation optimization unit is configured to set t=t+n and y=y+n, n is an integer greater than or equal to 1, take u max as the centroid of the evaluation data cluster AD(t+n, V y+n ), calculate the intra-cluster distance of the evaluation data cluster AD(t+n, V y+n ), denoted as \[AD(t+n, V y+n )], and compare the size of \[AD(t+n, V y+n )] and \max[AD(t, V y )], if there is any \[AD(t+n, V y+n)] < max [AD(t, V y )], then output the centroid u max and the test running speed V of the escalator y , otherwise output the centroid u max and the test running speed V of the escalator y+n corresponding to the maximum value of \[AD(t+n, V y+n ].

[0025] A subway transfer data analysis method based on big data, the method comprising the following steps:

[0026] Step S100: according to the station terrain map, draw the station transfer path, generate the transfer path set, and based on the escalator, integrate the station transfer paths passing through the same escalator, and generate the path aggregation point set;

[0027] Step S200: construct an escalator evacuation importance analysis model, calculate the evacuation importance of the escalator; divide the time segment, monitor the passenger flow data passing through the escalator in the time segment, and take the evacuation importance of the escalator as the weight coefficient of the passenger flow data, and generate the passenger flow dynamic data set;

[0028] Step S300: adjust the test running data of the escalator, and update and mark the passenger flow dynamic data set to generate the test data cluster; in different time cycle periods, use the same test running data to test the passenger flow evacuation, obtain the average distance between the test data clusters, and generate the evaluation data cluster;

[0029] Step S400: based on the evaluation data cluster, dynamically change the centroid position, analyze to obtain the maximum intrac cluster distance, and the centroid position when the intrac cluster distance is maximum; in different time cycle periods, with the adjustment of the test running data of the escalator, intelligently analyze and optimize the test running speed of the escalator during different time segments.

[0030] Further, the specific implementation process of step S100 comprises:

[0031] Step S101: compile the station transfer station waiting floor area code, and code any station transfer station waiting floor area as I i , wherein i represents the station transfer station waiting floor area serial number; draw the station transfer path, the station transfer path is all the paths between one station transfer station waiting floor area and another station transfer station waiting floor area, and each station transfer path is a path passing through an escalator; obtain the station transfer station waiting floor area I i to the station transfer station waiting floor area I jall the intra-station transfer paths drawn between the stations in set I and generate a transfer path set, denoted as TP(I i →I j )={tp x |x∈[1,a]},wherein j represents the sequence number of the waiting area of the intra-station transfer station, i≠j, tp x represents the xth intra-station transfer path, and a represents the sequence number of the intra-station transfer paths;

[0032] Step S102: uniformly encode all escalators existing in the intra-station transfer paths, and denote any one escalator as E e , wherein e represents the encoding number of the escalator; according to the intra-station transfer paths, integrate the intra-station transfer paths passing through the same escalator E e , and generate a path aggregation point set, denoted as CP(E e )={tp x |x∈[1,b]},wherein b represents the sequence number of the intra-station transfer paths.

[0033] Further, the specific implementation process of the step S200 comprises:

[0034] Step S201: based on the transfer path set and the path aggregation point set, construct an escalator evacuation importance analysis model, and calculate the evacuation importance of the escalator, and the specific calculation formula is as follows:

[0035]

[0036] wherein EI(E e ) represents the evacuation importance of the escalator E e , NUM[TP(I i →I j )] represents the total number of intra-station transfer paths contained in the transfer path set TP(I i →I j ), NUM[CP(E e )∩TP(I i →I j )] represents the total number of intra-station transfer paths contained in the intersection set of the path aggregation point set CP(E e ) and the transfer path set TP(I i →I j ), and N represents the maximum value of the sequence number of the intra-station transfer paths;

[0037] Step S202: divide the time in a day into k continuous time segments with the day as the time cycle unit, and denote any one time segment as T u , wherein u represents the sequence number of the time segment and u≤k;

[0038] A fixed pedestrian flow observation point is set up on each escalator, and the time segment T is observed in real time. u The system collects pedestrian flow data for each escalator within the station, whereby the pedestrian flow data refers to the number of pedestrians passing through the escalator in each time segment. Based on the transfer routes within the station, a dynamic pedestrian flow dataset, denoted as FD(T), is generated. u |t)={EI(E e )×fd(E e )|e∈[1,c]}, where FD(T u |t) represents the time segment T during the t-th time cycle. u The corresponding dynamic dataset of pedestrian flow generated within the dataset, fd(E) e ) represents the time segment T during the t-th time cycle. u Inside, via escalator E e The passenger flow data, where c represents the total number of escalators in the transfer path within the station;

[0039] According to the above method, the operating speed of escalators plays a crucial role in in-station transfers. Adjusting the operating speed of escalators can effectively disperse passenger flow. By planning transfer routes within the station, passenger flow is maximized and converged onto escalators. By changing the operating speed of escalators, the user experience is improved while adaptively controlling passenger flow. Furthermore, different numbers of transfer routes converged by escalators result in different evacuation effects. By analyzing the importance of escalator guidance, the evacuation capacity of escalators can be quantified and differentiated. The greater the importance of guidance, the stronger the escalator's evacuation capacity and the greater its crucial role. Therefore, the importance of escalator guidance is used as a weighting coefficient for passenger flow data, amplifying and reducing dynamic passenger flow data, and further distinguishing data characteristics.

[0040] The data cluster analysis module is used to adjust the test operation data of the escalator, update and label the dynamic dataset of pedestrian flow, and generate test data clusters. In different time cycles, pedestrian flow management tests are conducted using the same test operation data to obtain the average distance between test data clusters and generate evaluation data clusters.

[0041] Furthermore, the specific implementation process of step S300 includes:

[0042] Step S301: Adjust the test running speed of the escalator to ensure that the test running speeds of the escalators are the same, record the dynamic changes of the pedestrian flow dynamic dataset, and update and label the pedestrian flow dynamic dataset; record the running speed obtained when the y-th escalator is adjusted as V. y The updated and labeled dynamic dataset of pedestrian flow will then be denoted as the test data cluster FD(T). u|t, V y );

[0043] Step S302: According to the test data cluster FD(T u |t, V y ), the crowd flow test is carried out at the test running speed V y in the time cycle t, and the inter-cluster average distance between the test data cluster FD(T u |t, V y ) and the test data cluster FD(T u+1 |t, V y ) is obtained based on the Euclidean distance formula, denoted as CS[(T u →T u+1 )|t, V y ], the inter-cluster average distance is collected, and the evaluation data cluster is generated, denoted as AD(t, V y )={CS[(T u →T u+1 )|t, V y ]|u∈[1, k-1]}.

[0044] Further, the specific implementation process of the step S400 includes:

[0045] Step S401: Based on the evaluation data cluster, the centroid of the evaluation data cluster AD(t, V y ) is initialized, denoted as u: u→u+1, and the cluster element corresponding to the centroid position is CS[(T u →T u+1 )|t, V y ], based on the centroid u: u→u+1, the inter-cluster distance of the evaluation data cluster AD(t, V y ) is calculated; the centroid position is adjusted, and the inter-cluster distance is iteratively calculated until the inter-cluster distance is maximum, and the iteration calculation stops, the maximum inter-cluster distance is denoted as \max[AD(t, V y )], and the centroid at the time when the inter-cluster distance is maximum is denoted as u max ;

[0046] Step S402: Let t=t+n, y=y+n, n is an integer greater than or equal to 1, u max is taken as the centroid of the evaluation data cluster AD(t+n, V y+n ), and the inter-cluster distance of the evaluation data cluster AD(t+n, V y+n ) is calculated, denoted as \[AD(t+n, V y+n )]; the size of \[AD(t+n, V y+n )] and \max[AD(t, V y )] is compared, if there is any \[AD(t+n, V y+n)] < max [AD(t, V y )], output the centroid u max and the test running speed V y of the escalator max , otherwise output the centroid u y+n and the test running speed V y+n corresponding to the maximum value of \[AD(t+n, V y+n )].

[0047] According to the above method, there are three variables in the test data cluster, namely the time segment, the time period and the test running speed, the test is carried out at a running speed in a time period, the time period and the running speed are fixed, only the time segment is changed, and the evacuation capacity of each escalator in different time segments is observed to form a test data cluster; theoretically, the larger the cluster average distance, the greater the difference between the data points in different clusters, and the better the overall evacuation effect of the escalator during the flow transfer between adjacent two time segments, so that the difference between the flow data characteristics in the two time segments is more obvious after one time segment to the next time segment, the density of the flow can be reduced, and the evaluation data cluster is generated by collecting the cluster average distance corresponding to adjacent two time segments to record the data characteristics in a complete time period; the theoretical characteristics of the cluster average distance are grasped, and the analysis of the cluster distance is continued to find the time segment that has a key impact when the running speed test is carried out; due to the difference in the cluster average distance, the data in the evaluation data cluster exists continuous change fluctuation, the larger the cluster average distance, the better the evacuation effect, the smaller the cluster average distance, the worse the evacuation effect, theoretically, the larger the cluster distance, the more dispersed the data points in the cluster, when the running data is tested, the maximum cluster distance can filter out the time segment with the best evacuation effect, that is, the centroid position, and the centroid position is used as a quantitative, the cluster distance of the evaluation data cluster AD(t+n, V y+n ) is calculated in the next time period, and a maximum cluster distance of the evaluation data cluster AD(t+n, V y+n ) is found, if the maximum cluster distance of the evaluation data cluster AD(t+n, V y ) is greater than or equal to \[AD(t, V y+n )], the test running speed V y at the centroid position reaches the better evacuation effect.

[0048] Compared with the prior art, the beneficial effects achieved by the present application are: in the subway transfer data analysis system and method based on big data provided by the present application, the station transfer path is drawn, and the station transfer paths passing through the same escalator are integrated; the importance of the escalator is analyzed, the time segment is divided, and the people flow data passing through the escalator in the time segment is monitored; the test running data of the escalator is adjusted, and the people flow dynamic data set is updated and marked; in different time cycle periods, the same test running data is used for people flow dispersion test, the average distance between test data clusters is obtained; the centroid position is dynamically changed, the maximum intrac cluster distance is analyzed and obtained, and the centroid position when the intrac cluster distance is maximum, the test running speed of the escalator during different time segments is intelligently analyzed and optimized; further, the effectiveness of the subway station transfer path for people flow dispersion can be evaluated, and the running speed of the escalator during different time periods can be intelligently optimized, thereby improving the people flow dispersion capability. BRIEF DESCRIPTION OF DRAWINGS

[0049] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, which together with the embodiments of the present application, serve to explain the present application, and do not constitute a limitation to the present application. In the drawings:

[0050] Fig. 1 is a structural schematic diagram of a subway transfer data analysis system based on big data of the present application;

[0051] Fig. 2 is a step schematic diagram of a subway transfer data analysis method based on big data of the present application. DETAILED DESCRIPTION

[0052] 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 a part 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 skilled in the art without creative labor are within the scope of protection of the present application.

[0053] Please refer to Figs. 1-2 , the present application provides technical solutions:

[0054] Please refer to Fig. 1 , in the first embodiment: a subway transfer data analysis system based on big data is provided, which comprises: a transfer path processing module, a data acquisition module, a data cluster analysis module and an intelligent evaluation module;

[0055] The transfer path processing module draws a transfer path in the subway station according to a topographic map of the subway station, generates a transfer path set, and integrates the transfer paths in the subway station passing through the same escalator to generate a path aggregation point set based on the escalator;

[0056] The transfer path processing module further includes a path generation unit and a path aggregation unit.

[0057] The path generation unit is configured to compile a station transfer site waiting floor area code, and record any one station transfer site waiting floor area code as I i , where i represents the sequence number of the station transfer site waiting floor area; draw a transfer path in the subway station, which is all paths between one station transfer site waiting floor area and another station transfer site waiting floor area, and each transfer path in the subway station is a path passing through an escalator; obtain all transfer paths in the subway station drawn between the station transfer site waiting floor area I i and the station transfer site waiting floor area I j , and generate a transfer path set, recorded as TP(I i →I j )={tp x |x∈[1,a]},where j represents the sequence number of the station transfer site waiting floor area, and i≠j, tp x represents the xth transfer path in the subway station, and a represents the sequence number of the transfer path in the subway station.

[0058] The path aggregation unit is configured to uniformly code all escalators existing in the transfer path in the subway station, and record any one escalator as E e , where e represents the code number of the escalator; integrate the transfer paths in the subway station passing through the same escalator E e based on the transfer path in the subway station, and generate a path aggregation point set, recorded as CP(E e )={tp x |x∈[1,b]},where b represents the sequence number of the transfer path in the subway station.

[0059] The data acquisition module is configured to construct an escalator evacuation importance analysis model, calculate the evacuation importance of the escalator, divide time segments, monitor the people flow data passing through the escalator in the time segment, and take the evacuation importance of the escalator as a weight coefficient of the people flow data to generate a people flow dynamic data set.

[0060] The data acquisition module further includes an evacuation analysis unit and a dynamic perception unit.

[0061] The evacuation analysis unit constructs an escalator evacuation importance analysis model based on the transfer path set and the path aggregation point set, calculates the evacuation importance of the escalator, and the specific calculation formula is as follows:

[0062]

[0063] wherein, EI(E e ) represents the importance degree of the evacuation of the escalator E e , NUM[TP(I i →I j )] represents the total number of the in-station transfer paths contained in the transfer path set TP(I i →I j ), NUM[CP(E e )∩TP(I i →I j )] represents the total number of the in-station transfer paths contained in the intersection set between the path aggregation point set CP(E e ) and the transfer path set TP(I i →I j ), and N represents the maximum value of the serial number of the in-station transfer paths;

[0064] a dynamic perception unit, configured to divide a day into k continuous time segments in a time cycle period of a day, and any one time segment is recorded as T u , wherein u represents the serial number of the time segment and u≤k; a fixed flow observation point is arranged on each escalator, and the flow data passing through each escalator in the time segment T u is observed in real time, the flow data being the number of pedestrians passing through the escalator in each time segment; a flow dynamic data set is generated according to the in-station transfer paths, and is recorded as FD(T u |t)={EI(E e )×fd(E e )|e∈[1,c]},wherein FD(T u |t) represents the corresponding flow dynamic data set generated in the time segment T u in the tth time cycle period, fd(E e ) represents the flow data passing through the escalator E u in the time segment T e in the tth time cycle period, and c represents the total number of the escalators existing in the in-station transfer paths;

[0065] wherein the data cluster analysis module further comprises a data updating unit and a data correlation processing unit;

[0066] the data updating unit, configured to adjust the test running speed of the escalators and make the test running speeds of the escalators the same, record the dynamic changes of the flow dynamic data set, and update and mark the flow dynamic data set; and the running speed obtained by adjusting the escalators for the yth time is recorded as Vy， The updated and labeled dynamic dataset of pedestrian flow is then denoted as the test data cluster FD(T). u |t,V y );

[0067] The data association processing unit, based on the test data cluster FD(T) u |t,V y Within a time cycle t, the test running speed V is measured. y A pedestrian flow management test was conducted, and the test data cluster FD(T) was obtained based on the Euclidean distance formula. u |t,V y ) and test data cluster FD(T u+1 |t,V y The average inter-cluster distance between ) is denoted as CS[(T) u →T u+1 )|t,V y Collect the average distance between clusters and generate the evaluation data cluster, denoted as AD(t, V). y )={CS[(T u →T u+1 )|t,V y |u∈[1,k-1]};

[0068] The intelligent evaluation module dynamically changes the centroid position based on the evaluation data cluster, analyzes and obtains the maximum cluster distance and the centroid position when the cluster distance is maximum; within different time cycles, as the test operation data of the escalator is adjusted, it intelligently analyzes and optimizes the test operation speed of the escalator during different time segments.

[0069] The intelligent evaluation module also includes a data filtering unit and an evaluation optimization unit;

[0070] The data filtering unit initializes the evaluation data cluster AD(t, V) based on the evaluation data cluster. y The centroid of ) is denoted as u: u→u+1, and the cluster element corresponding to the centroid position is CS[(T u →T u+1 )|t,V y Based on the centroid u: u→u+1, the evaluation data cluster AD(t, V) is calculated. y The cluster distance is calculated by adjusting the centroid position and iteratively calculating the cluster distance until the cluster distance is maximized. The iteration stops when the maximum cluster distance is reached, and the maximum cluster distance is denoted as \max[AD(t, V)]. y Let u be the centroid where the cluster distance is maximum. max ;

[0071] The evaluation optimization unit is used to set t = t + n, y = y + n, where n is an integer greater than or equal to 1, and u maxthe centroid of the evaluation data cluster AD(t+n, V y+n ) is evaluated, and the inter-cluster distance of the evaluation data cluster AD(t+n, V y+n ) is calculated, denoted as \[AD(t+n, V y+n )];the size of \[AD(t+n, V y+n )] and \max[AD(t, V y )] is compared, if there is any \[AD(t+n, V y+n )<\max[AD(t, V y )], the centroid u max and the test running speed V y of the escalator are output, otherwise the centroid u max and the test running speed V y+n of the escalator corresponding to the maximum value of \[AD(t+n, V y+n )] are output.

[0072] Please refer to Fig. 2 , in the second embodiment: a subway transfer data analysis method based on big data is provided, which comprises the following steps:

[0073] Step S100: according to the subway station topographic map, the station transfer path is drawn, the transfer path set is generated, and based on the escalator, the station transfer paths passing through the same escalator are integrated to generate the path aggregation point set;

[0074] Specifically, the station transfer station waiting floor area code is prepared, and any one station transfer station waiting floor area code is denoted as I i , wherein i represents the station transfer station waiting floor area serial number; the station transfer path is drawn, the station transfer path is the path between one station transfer station waiting floor area and another station transfer station waiting floor area, and each station transfer path is a path passing through an escalator; all station transfer paths drawn between station transfer station waiting floor area I i and station transfer station waiting floor area I j are obtained, and a transfer path set is generated, denoted as TP(I i →I j )={tp x |x∈[1,a]},wherein j represents the station transfer station waiting floor area serial number, and i≠j, tp x represents the xth station transfer path, and a represents the station transfer path serial number;

[0075] All escalators existing in the station transfer path are uniformly coded, and any one escalator is denoted as E e, wherein e represents the code number of the escalator; according to the station transfer path, the station transfer paths passing through the same escalator E e are integrated, and a path aggregation point set is generated, denoted as CP(E e )={tp x | x∈[1, b]}, wherein b represents the serial number of the station transfer path;

[0076] Step S200: An escalator evacuation importance analysis model is constructed, the evacuation importance of the escalator is calculated, time segments are divided, the passenger flow data passing through the escalator in the time segment is monitored, the evacuation importance of the escalator is taken as the weight coefficient of the passenger flow data, and a passenger flow dynamic data set is generated;

[0077] Specifically, based on the transfer path set and the path aggregation point set, an escalator evacuation importance analysis model is constructed, the evacuation importance of the escalator is calculated, and the specific calculation formula is as follows:

[0078]

[0079] , wherein EI(E e ) represents the evacuation importance of the escalator E e , NUM[TP(I i →I j )] represents the total number of station transfer paths contained in the transfer path set TP(I i →I j ), NUM[CP(E e )∩TP(I i →I j )] represents the total number of station transfer paths contained in the intersection set between the path aggregation point set CP(E e ) and the transfer path set TP(I i →I j ), and N represents the maximum value of the serial number of the station transfer path;

[0080] Taking a day as a time cycle unit, the time in a day is divided into k continuous time segments, and any one time segment is denoted as T u , wherein u represents the serial number of the time segment and u≤k;

[0081] A fixed passenger flow observation point is arranged on each escalator, and the passenger flow data passing through each escalator in the time segment T u is observed in real time, the passenger flow data is the number of pedestrians passing through the escalator in each time segment; according to the station transfer path, a passenger flow dynamic data set is generated, denoted as FD(T u |t)={EI(E e )×fd(E e)|e∈[1,c]}, where FD(T u |t) represents the time segment T during the t-th time cycle. u The corresponding dynamic dataset of pedestrian flow generated within the dataset, fd(E) e ) represents the time segment T during the t-th time cycle. u Inside, via escalator E e The passenger flow data, where c represents the total number of escalators in the transfer path within the station;

[0082] Step S300: Adjust the test operation data of the escalator, update and label the dynamic dataset of pedestrian flow, and generate test data clusters; conduct pedestrian flow management tests using the same test operation data in different time cycles, obtain the average inter-cluster distance between test data clusters, and generate evaluation data clusters;

[0083] Specifically, the test running speed of the escalator is adjusted and made consistent, the dynamic changes of the pedestrian flow data set are recorded, and the data set is updated and labeled; the running speed obtained at the y-th adjustment of the escalator is denoted as V. y The updated and labeled dynamic dataset of pedestrian flow will then be denoted as the test data cluster FD(T). u |t,V y );

[0084] Based on the test data cluster FD(T) u |t,V y Within a time cycle t, the test running speed V is measured. y A pedestrian flow management test was conducted, and the test data cluster FD(T) was obtained based on the Euclidean distance formula. u |t,V y ) and test data cluster FD(T u+1 |t,V y The average inter-cluster distance between ) is denoted as CS[(T) u →T u+1 )|t,V y Collect the average distance between clusters and generate the evaluation data cluster, denoted as AD(t, V). y )={CS[(T u →T u+1 )|t,V y |u∈[1,k-1]};

[0085] Step S400: Based on the evaluation data cluster, dynamically change the centroid position, analyze and obtain the maximum cluster distance, and the centroid position when the cluster distance is maximum; within different time cycles, as the test operation data of the escalator is adjusted, intelligently analyze and optimize the test operation speed of the escalator during different time segments.

[0086] Specifically, based on the evaluation data cluster, the centroid of the evaluation data cluster AD(t, V y ) is initialized, denoted as u: u→u+1, and the cluster element corresponding to the centroid position is CS[(T u →T u+1 )|t, V y ], based on the centroid u: u→u+1, the intra-cluster distance of the evaluation data cluster AD(t, V y ) is calculated; the centroid position is adjusted, and the intra-cluster distance is iteratively calculated until the intra-cluster distance is maximum, and the iteration calculation stops, the maximum intra-cluster distance is denoted as \max[AD(t, V y )], and the centroid at the maximum intra-cluster distance is denoted as u max ;

[0087] Let t=t+n, y=y+n, n is an integer greater than or equal to 1, and u max is taken as the centroid of the evaluation data cluster AD(t+n, V y+n ), and the intra-cluster distance of the evaluation data cluster AD(t+n, V y+n ) is calculated, denoted as \[AD(t+n, V y+n )]; the size of \[AD(t+n, V y+n )] and \max[AD(t, V y )] is compared, if there is any \[AD(t+n, V y+n )<\max[AD(t, V y )], the centroid u max and the test running speed V y of the escalator are output, otherwise the centroid u max and the maximum value of \[AD(t+n, V y+n )] corresponding to the test running speed V y+n of the escalator are output.

[0088] For example, the maximum intra-cluster distance is denoted as \max[AD(1, V1)], and the centroid at the maximum intra-cluster distance is denoted as 1:1→2; if there is \[AD(2, V2)] greater than or equal to \max[AD(1, V1)], 1:1→2, V2 is output, and during the time segments 1 and 2, the test running speed V2 is preferentially used.

[0089] It is to be noted that, in the present text, relational terms such as first and second and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0090] Finally, it should be noted that the above-mentioned only constitutes preferred embodiments of the present application and is not intended to limit the present application, and although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still make modifications to the technical solutions described in the foregoing embodiments or make equivalent replacements to some of the technical features, all of which shall fall within the scope of protection of the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the scope of protection of the present application.

Claims

1. A metro transfer data analysis method based on big data, characterized in that, The method comprises the following steps: Step S100: according to a subway station topographic map, drawing a station transfer path, generating a transfer path set, and based on an escalator, integrating the station transfer paths passing through the same escalator to generate a path aggregation point set; Step S200: constructing an escalator evacuation importance analysis model, calculating the evacuation importance of the escalator; dividing time segments, monitoring the people flow data passing through the escalator in the time segments, and taking the evacuation importance of the escalator as a weight coefficient of the people flow data to generate a people flow dynamic data set; Step S300: adjusting the test running data of the escalator, updating and marking the people flow dynamic data set to generate a test data cluster; in different time cycle periods, using the same test running data to test the people flow evacuation, obtaining the inter-cluster average distance between the test data clusters, and generating an evaluation data cluster; Step S400: based on the evaluation data cluster, dynamically changing the centroid position, analyzing the maximum inter-cluster distance, and the centroid position when the inter-cluster distance is maximum; in different time cycle periods, intelligently analyzing and optimizing the test running speed of the escalator during different time segments along with the adjustment of the test running data of the escalator. 2.The metro transfer data analysis method based on big data according to claim 1, wherein, The specific implementation process of the step S100 comprises: Step S101: compile the station metro transfer site waiting floor area code, and code any one station metro transfer site waiting floor area as I i , wherein i represents the station metro transfer site waiting floor area serial number; draw the station transfer path, which is the path between any two station metro transfer site waiting floor areas, and each station transfer path is the path through escalator; obtain all station transfer paths between station metro transfer site waiting floor areas I i and I j , and generate a transfer path set, denoted as TP(I i →I j )={tp x |x∈[1,a]},wherein j represents the station metro transfer site waiting floor area serial number, i≠j, and tp x represents the xth station transfer path, and a represents the station transfer path serial number; Step S102: encode all escalators existing in the station transfer path uniformly, and record any one escalator as E e , wherein e represents the code number of the escalator; according to the station transfer path, integrate the station transfer paths passing through the same escalator E e , and generate a path aggregation point set, recorded as CP(E e )={tp x |x∈[1,b]},wherein b represents the serial number of the station transfer path. 3.The metro transfer data analysis method based on big data according to claim 2, characterized in that, The specific implementation process of the step S200 comprises: Step S201: based on the transfer path set and the path aggregation point set, constructing an escalator evacuation importance analysis model, calculating the evacuation importance of the escalator, and the specific calculation formula is as follows: wherein, EI(E e ) represents the importance degree of the evacuation of the escalator E e , NUM[TP(I i →I j )] represents the total number of the in-station transfer paths included in the transfer path set TP(I i →I j ), NUM[CP(E e )∩TP(I i →I j )] represents the total number of the in-station transfer paths included in the intersection set between the path aggregation point set CP(E e ) and the transfer path set TP(I i →I j ), and N represents the maximum value of the serial number of the in-station transfer paths. Step S202: divide the time in a day into k continuous time segments with a time cycle unit of day, and any one time segment is denoted as T u wherein u represents the serial number of the time segment and u≤k; A fixed pedestrian flow observation point is set up on each escalator, and the time segment T is observed in real time. u The system collects pedestrian flow data for each escalator within the station, whereby the pedestrian flow data refers to the number of pedestrians passing through the escalator in each time segment. Based on the transfer routes within the station, a dynamic pedestrian flow dataset, denoted as FD(T), is generated. u |t)={EI(E e )×fd(E e )|e∈[1,c]}, where FD(T u |t) represents the time segment T during the t-th time cycle. u The corresponding dynamic dataset of pedestrian flow generated within the dataset, fd(E) e ) represents the time segment T during the t-th time cycle. u Inside, via escalator E e The passenger flow data, where c represents the total number of escalators in the transfer routes within the station.

4. The metro transfer data analysis method based on big data according to claim 3, characterized in that, The specific implementation process of the step S300 comprises: Step S301: adjusting the test running speed of the escalator and making the test running speed of the escalator the same, recording the dynamic change of the people flow dynamic data set, and updating and marking the people flow dynamic data set; the running speed obtained by adjusting the escalator for the yth time is recorded as V y , the updated and marked people flow dynamic data set is recorded as test data cluster FD(T u |t, V y ); Step S302: According to the test data cluster FD(T u |t, V y ), the crowd flow test is carried out at the test running speed V y in the time cycle t, and the inter-cluster average distance between the test data cluster FD(T u |t, V y ) and the test data cluster FD(T u+1 |t, V y ) is obtained based on the Euclidean distance formula, recorded as CS[(T u →T u+1 )|t, V y ], the inter-cluster average distance is collected, and the evaluation data cluster is generated, recorded as AD(t, V y )={CS[(T u →T u+1 )|t, V y ]|u∈[1, k-1]}.

5. The metro transfer data analysis method based on big data according to claim 4, characterized in that, The specific implementation process of the step S400 comprises: Step S401: based on the evaluation data cluster, initialize the centroid of the evaluation data cluster AD(t, V y ), denoted as u: u→u+1, and the cluster elements corresponding to the centroid position are CS[(T u →T u+1 )|t, V y ], based on the centroid u: u→u+1, the intra-cluster distance of the evaluation data cluster AD(t, V y ) is calculated; adjust the centroid position, and iteratively calculate the intra-cluster distance until the intra-cluster distance is maximum, and the iteration calculation stops, the maximum intra-cluster distance is denoted as \max[AD(t, V y )], and the centroid when the intra-cluster distance is maximum is denoted as u max ; Step S402: let t=t+n, y=y+n, n is an integer greater than or equal to 1, to u max As the centroid of the evaluation data cluster AD(t+n, V y+n ), and the cluster distance of the evaluation data cluster AD(t+n, V y+n ) is calculated, denoted as \[AD(t+n, V y+n )];Compare \[AD(t+n, V y+n )] with \max[AD(t, V y )], if there is any \[AD(t+n, V y+n )<\max[AD(t, V y )], output the centroid u max and the test running speed V y of the escalator, otherwise output the centroid u max and the maximum value of \[AD(t+n, V y+n )] corresponding to the test running speed V y+n of the escalator. 6.A metro transfer data analysis system based on big data, characterized in that, The system comprises a transfer path processing module, a data acquisition module, a data cluster analysis module, and an intelligent evaluation module; The transfer path processing module draws a station transfer path according to a subway station topographic map, generates a transfer path set, and based on an escalator, integrates the station transfer paths passing through the same escalator to generate a path aggregation point set; The data acquisition module is configured to construct an escalator evacuation importance analysis model, calculate the evacuation importance of the escalator, divide time segments, monitor the people flow data passing through the escalator in the time segments, and take the evacuation importance of the escalator as a weight coefficient of the people flow data to generate a people flow dynamic data set; The data cluster analysis module is configured to adjust the test running data of the escalator, update and mark the people flow dynamic data set to generate a test data cluster; in different time cycle periods, using the same test running data to test the people flow evacuation, obtaining the inter-cluster average distance between the test data clusters, and generating an evaluation data cluster; The intelligent evaluation module is configured to, based on the evaluation data cluster, dynamically change the centroid position, analyze the maximum inter-cluster distance, and the centroid position when the inter-cluster distance is maximum; in different time cycle periods, intelligently analyze and optimize the test running speed of the escalator during different time segments along with the adjustment of the test running data of the escalator.

7. The big data-based metro transfer data analysis system according to claim 6, characterized in that: The transfer path processing module further comprises a path generation unit and a path aggregation unit; The path generating unit is configured to compile a station-to-station transfer site waiting floor area code, and code any one station-to-station transfer site waiting floor area as I i , wherein i represents a station-to-station transfer site waiting floor area serial number; draw a station transfer path, the station transfer path being all paths between one station-to-station transfer site waiting floor area and another station-to-station transfer site waiting floor area, and each station transfer path being a path passing through an escalator; obtain all station transfer paths drawn between station-to-station transfer site waiting floor areas I i and I j , and generate a transfer path set, denoted as TP(I i →I j ) = {tp x |x∈[1,a]}, wherein j represents a station-to-station transfer site waiting floor area serial number, i≠j, tp x represents the xth station transfer path, and a represents a station transfer path serial number; The path aggregation unit is configured to uniformly encode all escalators existing in the station transfer path, and record any one escalator as E e . Wherein, e represents the code number of the escalator; according to the station transfer path, the station transfer paths passing through the same escalator E e are integrated, and a path aggregation point set is generated, recorded as CP(E e )={tp x |x∈[1,b]}. Wherein, b represents the serial number of the station transfer path.

8. The big data-based metro transfer data analysis system according to claim 7, characterized in that: The data collection module further comprises a guidance analysis unit and a dynamic perception unit; The guidance analysis unit constructs an escalator guidance importance analysis model based on the set of transfer paths and the set of path aggregation points, and calculates the guidance importance of the escalator, and the specific calculation formula is as follows: wherein, EI(E e ) represents the importance degree of the evacuation of the escalator E e , NUM[TP(I i →I j )] represents the total number of the in-station transfer paths included in the transfer path set TP(I i →I j ), NUM[CP(E e )∩TP(I i →I j )] represents the total number of the in-station transfer paths included in the intersection set between the path aggregation point set CP(E e ) and the transfer path set TP(I i →I j ), and N represents the maximum value of the serial number of the in-station transfer paths. The dynamic sensing unit is used to divide a day into k consecutive time segments, with a day as the time cycle unit, and denot any one time segment as T. u Where u represents the sequence number of the time segment and u≤k; a fixed pedestrian flow observation point is set on each escalator, and the time segment T is observed in real time. u The system collects pedestrian flow data for each escalator within the station, whereby the pedestrian flow data refers to the number of pedestrians passing through the escalator in each time segment. Based on the transfer routes within the station, a dynamic pedestrian flow dataset, denoted as FD(T), is generated. u |t)={EI(E e )×fd(E e )|e∈[1,c]}, where FD(T u |t) represents the time segment T during the t-th time cycle. u The corresponding dynamic dataset of pedestrian flow generated within the dataset, fd(E) e ) represents the time segment T during the t-th time cycle. u Inside, via escalator E e The passenger flow data, where c represents the total number of escalators in the transfer routes within the station.

9. The big data-based metro transfer data analysis system according to claim 8, characterized in that: The data cluster analysis module further comprises a data updating unit and a data correlation processing unit; The data updating unit is configured to adjust the test running speed of the escalator to be the same, record dynamic changes of the people flow dynamic data set, and update and mark the people flow dynamic data set; and record the running speed obtained when the escalator is adjusted for the yth time as V y The people flow dynamic data set updated and marked is recorded as a test data cluster FD(T u |t, V y ). The data correlation processing unit, according to the test data cluster FD(T u |t, V y ), carries out the crowd dispersing test at the test running speed V y in the time cycle t, and obtains the inter-cluster average distance between the test data cluster FD(T u |t, V y ) and the test data cluster FD(T u+1 |t, V y ) based on the Euclidean distance formula, recorded as CS[(T u →T u+1 )|t, V y ], collects the inter-cluster average distance, and generates the evaluation data cluster, recorded as AD(t, V y )={CS[(T u →T u+1 )|t, V y ]|u∈[1, k-1]}.

10. The big data-based metro transfer data analysis system according to claim 9, characterized in that: The intelligent evaluation module further comprises a data screening unit and an evaluation optimization unit; The data screening unit initializes the centroid of the evaluation data cluster AD(t, V y ) based on the evaluation data cluster, denoted as u: u→u+1, and the cluster elements corresponding to the centroid position are CS[(T u →T u+1 )|t, V y ], calculates the intra-cluster distance of the evaluation data cluster AD(t, V y ) based on the centroid u: u→u+1, adjusts the centroid position, and iteratively calculates the intra-cluster distance until the intra-cluster distance is maximum, at which point the iteration calculation stops, the maximum intra-cluster distance is denoted as \max[AD(t, V y )], and the centroid at the time when the intra-cluster distance is maximum is denoted as u max ; The evaluation optimization unit is configured to let t=t+n, y=y+n, n is an integer greater than or equal to 1, to u max is the centroid of the evaluation data cluster AD(t+n, V y+n ), and the intra-cluster distance of the evaluation data cluster AD(t+n, V y+n ) is calculated and denoted as \[AD(t+n, V y+n )]; the size of \[AD(t+n, V y+n )] and \max[AD(t, V y )] is compared, if there is any \[AD(t+n, V y+n )<\max[AD(t, V y )], the centroid u max and the test running speed V y of the escalator are output, otherwise the centroid u max and the test running speed V y+n of the escalator corresponding to the maximum value of \[AD(t+n, V y+n )] are output.

Images