Spillover passenger flow estimation method, and electronic device and storage medium

Abstract

Provided in the present invention are a spillover passenger flow estimation method, and an electronic device and a storage medium. The method comprises: on the basis of metro line network data, constructing a line-network directed graph, wherein vertexes in the line-network directed graph are stations, and edges in the line-network directed graph comprise interval edges between stations and transfer edges in transfer stations; acquiring historical transactions and train timetables at M similar dates closest to the current date, and acquiring current line network anomaly data; on the basis of the train timetables, calculating the weights of the interval edges in the line-network directed graph; on the basis of the historical transactions and the train timetables, calculating the weights of the transfer edges in the line-network directed graph; on the basis of the line network anomaly data, deleting from the line-network directed graph vertexes and edges corresponding to abnormal stations, and constructing a line-network topology graph; on the basis of the line-network topology graph, an online map, and the weights of the interval edges and transfer edges in the line-network directed graph, constructing a line-network anomaly topology graph, and determining the weights of edges in the line-network anomaly topology graph, wherein the edges in the line-network anomaly topology graph comprise metro edges and external edges; and on the basis of the historical transactions, the line-network topology graph and the line-network anomaly topology graph, determining a spillover passenger flow generated due to a metro emergency. The present invention is used for accurately estimating a spillover passenger flow when a local fault occurs in a line network.

Description

Overflow passenger flow estimation method, electronic device and storage medium TECHNICAL FIELD

[0001] The present application relates to the field of urban rail transit passenger flow management, in particular to an overflow passenger flow estimation method, an electronic device and a storage medium. BACKGROUND

[0002] At present, the subways in many cities in China have formed large networks, with daily passenger flow ranging from several million to tens of millions, and the passenger flow has obvious peak-valley characteristics. The subway operation is usually stable and reliable, but unexpected events such as station water inflow and extreme weather, or major activities requiring some stations and sections to be suspended according to the plan, will cause partial operation interruption of the network. After the partial operation interruption of the subway, the passenger flow in the subway will overflow outside the subway, and thus needs to be guided by other transportation modes. In order to reasonably arrange the guiding capacity, it is necessary to obtain the distribution of the overflow passenger flow in time and space.

[0003] There are currently two types of passenger flow distribution algorithms. The first type is to distribute transactions in proportion to a plurality of possible paths according to a fixed proportion, and then calculate the generation time of passenger flow at each node according to the time weight of each edge in the path. The second type is to infer a train sequence that a passenger is most likely to take according to the passenger's entry and exit times and the train timetable, and calculate the generation time of passenger flow at each node according to the train sequence. The above two algorithms are based on normal network and infer the path according to the completed entry and exit behavior, and cannot directly calculate the overflow passenger flow when the network is partially faulty. In addition, in the process of inferring the spatial and temporal distribution of passenger flow, it is necessary to manually configure the time weight of each section in the network, the transfer time weight of the transfer station and other parameters, which requires a large amount of configuration and the calculation result has a large subjectivity. TECHNICAL PROBLEM

[0004] The purpose of the present application is to provide an overflow passenger flow estimation method, an electronic device and a storage medium for partial interruption of subway operation, for accurately estimating the overflow passenger flow when the network is partially faulty. TECHNICAL SOLUTION

[0005] In order to achieve the above purpose, the present application realizes the following technical scheme:

[0006] An overflow passenger flow estimation method, comprising the following steps:

[0007] constructing a network directed graph according to subway network data, wherein the vertices in the network directed graph are stations, and the edges in the network directed graph include section edges between stations and transfer edges in transfer stations;

[0008] acquire historical transactions of M days similar to the current date, train timetable, and acquire current line network abnormal data;

[0009] According to the train timetable, the weight of the interval edge in the line network directed graph is calculated.

[0010] According to the historical transactions and train timetable, the weight of the transfer edge in the line network directed graph is calculated.

[0011] According to the line network abnormal data, the vertex and edge corresponding to the abnormal station in the line network directed graph are deleted to construct a line network topology graph.

[0012] According to the line network topology graph and online map, and the weight of the interval edge and transfer edge in the line network directed graph, a line network abnormal topology graph is constructed, and the weight of each edge in the line network abnormal topology graph is determined, the edges in the line network abnormal topology graph include subway edges and external edges.

[0013] Based on the historical transactions, the overflow passenger flow generated by the subway incident is determined according to the line network topology graph and the line network abnormal topology graph.

[0014] Optionally, the subway line network data includes: a station list and an interval list.

[0015] The attributes of each record in the station list include: line name and station name, wherein the N-line transfer station has N records in the station list, the line names are different and the station names are the same in the N records.

[0016] The attributes of each record in the interval list include: starting line name, starting station name, ending line name, and ending station name.

[0017] Optionally, the line network directed graph is constructed according to the subway line network information, including:

[0018] Initialize the line network directed graph, and the vertex set and edge set are empty;

[0019] For each station in the station list, a corresponding vertex is added to the vertex set of the line network directed graph;

[0020] For each interval in the interval list, an edge is added to the edge set of the line network directed graph, and the type of the edge is interval edge.

[0021] For the N-line transfer station in the station list, a pair of directed edges is added between the corresponding N station vertices in the line network directed graph, and the type of the edge is transfer edge.

[0022] Optionally, the attribute of each transaction in the historical transactions comprises: an inbound line name, an inbound station name, an inbound time, an outbound line name, an outbound station name, and an outbound time.

[0023] The train timetable comprises a plurality of train services and a list of arrival-departure triplets corresponding to each train service, wherein the attribute of each train service comprises a line name and a train service number, and the attribute of each triplet in the list of arrival-departure triplets comprises a station name, an arrival time, and a departure time.

[0024] Optionally, the calculation of the weight of each interval edge in the line network directed graph according to the train timetable comprises:

[0025] initializing an empty first weight list for each interval edge in the line network directed graph;

[0026] determining the first weight list corresponding to each interval edge according to all train services and the list of arrival-departure triplets corresponding to each train service in the train timetable;

[0027] determining the weight of each interval edge in the line network directed graph according to the first weight list corresponding to each interval edge.

[0028] Optionally, the determination of the first weight list corresponding to each interval edge according to all train services and the list of arrival-departure triplets corresponding to each train service in the train timetable comprises:

[0029] for each train service in the train timetable, the following operations are performed to determine the first weight list corresponding to each interval edge:

[0030] sorting all triplets in the list of arrival-departure triplets of the train service according to the arrival time;

[0031] for two adjacent triplets, searching for the corresponding interval edge in the line network directed graph according to the station names corresponding to the two triplets, and adding a first numerical value to the first weight list corresponding to the interval edge, wherein the first numerical value is equal to the arrival time corresponding to the latter triplet minus the arrival time corresponding to the former triplet.

[0032] Optionally, the determination of the weight of each interval edge in the line network directed graph according to the first weight list corresponding to each interval edge comprises:

[0033] determining the median of all numerical values in the first weight list corresponding to each interval edge, and setting the median as the weight of each interval edge in the line network directed graph.

[0034] Optionally, the calculation of the weight of each transfer edge in the line network directed graph according to the historical transactions and the train timetable comprises:

[0035] initialize an empty second weight list for each transfer edge in the line network directed graph;

[0036] search for transfer paths in the line network directed graph and construct a transfer path set; the transfer path is a path composed of four stations and three edges, the second edge in the transfer path is a transfer edge, and the first and third edges are section edges;

[0037] For each transfer path in the transfer path set, determine the second weight list corresponding to the transfer edge in the transfer path according to the historical transactions and the train timetable;

[0038] According to the second weight list corresponding to each transfer edge, determine the weight of each transfer edge in the line network directed graph.

[0039] Optionally, the second weight list corresponding to the transfer edge in the transfer path is determined according to the historical transactions and the train timetable, including:

[0040] For each transfer path in the transfer path set, the following operations are performed to determine the second weight list corresponding to the transfer edge in the transfer path:

[0041] Find all transactions from the historical transactions in which the entry station is the first station in the transfer path and the exit station is the fourth station in the transfer path, and generate a transaction set;

[0042] For each transaction in the transaction set, extract the data corresponding to the entry line, entry station, entry time, exit line, exit station, and exit time of the transaction, and find all pairs of train numbers that meet the preset conditions in the transaction day train timetable of the entry line and the exit line, wherein each pair of train numbers includes an entry train number on the entry line and an exit train number on the exit line.

[0043] For each pair of train numbers that meets the preset condition, subtract the arrival time of the exit train number at the third station in the transfer path from the departure time of the exit train number at the third station in the transfer path to obtain a second value, and add the second value to the second weight list of the transfer edge in the transfer path.

[0044] Optionally, the preset condition includes:

[0045] The entry train number passes through the first station and the second station in the transfer path in turn;

[0046] The exit train number passes through the third station and the fourth station in the transfer path in turn;

[0047] The departure time of the inbound train at the first station in the transfer path minus the inbound time in the transaction is greater than a first preset time length;

[0048] The outbound time in the transaction minus the arrival time of the outbound train at the fourth station in the transfer path is greater than a second preset time length.

[0049] Optionally, the determining of the weight of each transfer edge in the line network directed graph according to the second weight list corresponding to each transfer edge comprises:

[0050] The values in the second weight list of each transfer edge are sorted, and the average value obtained by deleting a certain number of minimum values and a certain number of maximum values is taken as the weight of the transfer edge.

[0051] Optionally, the line network abnormal data comprises a blocked station list and a blocked section list, and the attributes of each blocked station in the blocked station list comprise a line name and a station name, and the attributes of each blocked section in the blocked section list comprise a section start station and a section end station. For stations on different lines at a transfer station, whether to be closed can be set separately.

[0052] Optionally, the constructing of the line network topology graph from the line network abnormal data by deleting the vertex and edge corresponding to the abnormal station in the line network directed graph comprises:

[0053] For each blocked section in the line network abnormal data, the section edge corresponding to the blocked section is deleted from the line network directed graph;

[0054] For each blocked station in the line network abnormal data, if the station has a transfer edge connected thereto in the line network directed graph, the transfer edge is deleted.

[0055] Optionally, the constructing of the line network abnormal topology graph according to the line network topology graph, the online map, and the weights of the section edges and the transfer edges in the line network directed graph, and the determining of the weights of the edges in the line network abnormal topology graph comprise:

[0056] A copy of the line network topology graph is copied, and all edges in the copy are deleted to obtain an initial line network abnormal topology graph;

[0057] For each pair of stations in the initial abnormal line network topology graph, the pair of stations includes a first station and a second station, if neither of the two stations is blocked, find a first shortest path from the first station to the second station in the line network topology graph; if the first shortest path is found, add an edge from the first station to the second station in the initial abnormal line network topology graph, the type of the edge is a subway edge, and the weight of the edge is equal to the sum of the weights of the edges on the first shortest path; if the first shortest path is not found, call an online map interface to find a driving time from the first station to the second station, and add an edge from the first station to the second station in the initial abnormal line network topology graph, the type of the edge is an external edge, and the weight of the edge is determined according to the driving time.

[0058] Merge the station vertices of the multi-line transfer stations in the initial abnormal line network topology graph into one vertex, and delete all transfer edges to obtain the abnormal line network topology graph.

[0059] Optionally, the overflow passenger flow generated by the subway incident is determined based on the historical transactions, the line network topology graph and the abnormal line network topology graph, and includes:

[0060] An empty list is initialized, and each element in the empty list has the following attributes: a starting station, an ending station, and a 24-hour hour number.

[0061] For each transaction in the historical transactions, each element in the empty list is determined based on the line network topology graph and the abnormal line network topology graph.

[0062] The empty list is taken as a key value in the form of a triple consisting of the three attributes of an element, the number of corresponding elements in the empty list is counted to obtain a transaction number, and each transaction number is divided by the number of days M to obtain an overflow passenger flow list, each element in the overflow passenger flow list has the following attributes: a starting station, an ending station, a 24-hour hour number, and a transaction number.

[0063] Optionally, the determination of each element in the empty list based on the line network topology graph and the abnormal line network topology graph for each transaction in the historical transactions includes:

[0064] For each transaction in the historical transactions of M days, the following operations are performed to determine each element in the empty list:

[0065] The entry station, the exit station and the entry time of the transaction are extracted;

[0066] A second shortest path is found in the line network topology graph with the entry station as the starting point and the exit station as the ending point; if the second shortest path is found, the subsequent operations are skipped, and the next transaction is processed.

[0067] If the second shortest path is not found in the line network topology graph, a third shortest path is found in the line network abnormal topology graph, starting from the inbound station and ending at the outbound station;

[0068] For each external edge on the third shortest path, the external edge is pointed to by a third station and pointed to a fourth station, a time T corresponding to the external edge is calculated, a record R is generated, attributes of the record R include the third station, the fourth station, and a time T corresponding to the third station, and the record R is added to the empty list;

[0069] The time T is equal to the inbound time of the transaction plus a sum of time consumptions of all edges of the second shortest path from the inbound station to the third station, wherein the time consumption of a subway edge is equal to the weight of the edge, and the time consumption of an external edge is equal to the driving time corresponding to the external edge.

[0070] An electronic device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete communication with each other through the communication bus;

[0071] The memory is used to store a computer program;

[0072] The processor is used to execute the program stored on the memory, and implement the method steps of any one of the preceding method steps.

[0073] A computer readable storage medium, the computer readable storage medium stores a computer program, the computer program is executed by a processor to implement the method steps of any one of the preceding method steps. Advantages

[0074] Compared with the prior art, the present application has the following advantages:

[0075] The present application calculates the transfer time of each transfer station according to historical passenger flow and a timetable; constructs a line network abnormal topology graph, divides edges in the line network abnormal topology graph into subway edges and external edges, and calculates the weight of the subway edges and the external edges by using a shortest path algorithm and an online map; simulates the walking of passengers in the abnormal line network graph by using historical passenger flow and the line network abnormal topology graph, takes the external edges as channels for overflow passenger flow, and outputs a specific time attribute for each overflow passenger flow. The scheme of the present application is based on historical passenger flow, a train timetable, a line network abnormal topology graph and real-time traffic data of an online map, calculates overflow passenger flow, and the generated data has high accuracy, real-time performance and interpretability in space and time, can support the rescheduling of the external traffic system of the subway in a finer granularity, and can effectively evacuate overflow passenger flow. BRIEF DESCRIPTION OF DRAWINGS

[0076] In order to make the technical solutions of the present application clearer, the drawings needed to be used in the description will be briefly introduced as follows. Obviously, the drawings in the following description are one embodiment of the present application, and other drawings can also be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0077] Fig. 1 is a flow chart of an overflow passenger flow estimation method provided by one embodiment of the present application;

[0078] Fig. 2 is a partial structure diagram of a conventional line network diagram in an example;

[0079] Fig. 3 is a line network directed graph constructed according to the line network diagram shown in Fig. 2;

[0080] Fig. 4 is a line network topology diagram constructed according to the line network directed graph shown in Fig. 3;

[0081] Fig. 5 is a line network abnormal topology diagram constructed according to the line network topology diagram shown in Fig. 4;

[0082] Fig. 6 is a structure diagram of an electronic device provided by one embodiment of the present application. Best Mode for Carrying Out the Invention

[0083] The solutions proposed by the present application will be further described in detail in combination with the drawings and specific embodiments. The advantages and features of the present application will be clearer according to the following description. It should be noted that the drawings are very simplified and all use non-precise proportions, which are only used to facilitate and clarify the purpose of assisting the description of the embodiments of the present application.

[0084] One embodiment of the present application provides an overflow passenger flow estimation method, as shown in Fig. 1, which includes the following steps S1-S7.

[0085] In step S1, a line network directed graph is constructed according to metro line network data, the vertices in the line network directed graph are stations, and the edges in the line network directed graph include interval edges between stations and transfer edges in transfer stations.

[0086] Specifically, metro line network data can be obtained from a line network control center database, the metro line network data includes a station list and an interval list, the attributes of each record in the station list include line names and station names, wherein there are N records of the same station name but different line names in the station list for an N-line transfer station, and the attributes of each record in the interval list include starting line names, starting station names, terminal line names and terminal station names.

[0087] The step S1 of constructing a line network directed graph according to metro line network information can specifically include:

[0088] Step S11, initialize the line network directed graph, both the vertex set and the edge set are empty;

[0089] Step S12, for each station in the station list, add a corresponding vertex to the vertex set of the line network directed graph;

[0090] Step S13, for each section in the section list, add an edge to the edge set of the line network directed graph, and the type of the edge is a section edge;

[0091] Step S14, for each N-line transfer station in the station list, add a pair of directed edges between the corresponding N station vertices in the line network directed graph, and the type of the edge is a transfer edge.

[0092] In step S14, according to the station list, the number of lines under the station name is counted. If the number N corresponding to a station name is greater than 1, it indicates that a transfer station is found, and thus a pair of directed edges is added between the vertices corresponding to the transfer station in the line network directed graph, and the type of the directed edge is a transfer edge.

[0093] Step S2, obtain historical transactions, train schedules of M days similar to the current date, and obtain current line network abnormal data.

[0094] Specifically, according to the week number of the current date (i.e. the date for which the overflow passenger flow needs to be estimated), the latest M similar historical dates are selected. For example, if the current date is Monday, the latest M Mondays are selected. M is a global hyperparameter.

[0095] The historical transactions of the selected M days can be obtained from the line network control center database, and the attributes of each transaction in the historical transactions include: inbound line name, inbound station name, inbound time, outbound line name, outbound station name, and outbound time.

[0096] The train schedule can be obtained from the line network control center database. The train schedule includes a plurality of train numbers and a corresponding list of arrival-departure triplets for each train number. The attributes of the train number include: line name and train number. The attributes of each triplet in the list of arrival-departure triplets include: station name, arrival time, and departure time.

[0097] The line network abnormal data can be obtained from a configuration file, and the line network abnormal data includes: a list of blocked stations and a list of blocked sections. The attributes of each blocked station in the list of blocked stations include: line name and station name. The attributes of each blocked section in the list of blocked sections include: section start station and section end station. For stations under different lines of a transfer station, whether to be blocked can be set separately.

[0098] Step S3, calculating the weight of the section edge in the line network directed graph according to the train timetable.

[0099] The step S3 can specifically include:

[0100] Step S31, initializing an empty first weight list for each section edge in the line network directed graph;

[0101] Step S32, determining the first weight list corresponding to each section edge according to all train services and their corresponding arrival-departure triplets in the train timetable;

[0102] Step S33, determining the weight of each section edge in the line network directed graph according to the first weight list corresponding to each section edge.

[0103] The step S32 can be implemented in the following manner: for each train service in the train timetable, the following operations are performed to determine the first weight list corresponding to each section edge:

[0104] sorting all triplets in the arrival-departure triplet list of the train service according to the arrival time;

[0105] for two adjacent triplets, finding the corresponding section edge in the line network directed graph according to the station names corresponding to the two triplets, and adding a first value to the first weight list corresponding to the section edge, the first value being equal to the arrival time corresponding to the latter triplet minus the arrival time corresponding to the former triplet.

[0106] For example, the two adjacent triplets are (station name 1, arrival time 1, departure time 1) and (station name 2, arrival time 2, departure time 2), the corresponding section edge is found in the line network directed graph according to (station name 1, station name 2), and a first value X1 is added to the first weight list L1 corresponding to the section edge, X1 = arrival time 2 - arrival time 1.

[0107] The step S33 can be: determining the median of all values in the first weight list corresponding to each section edge, and setting the median as the weight of each section edge in the line network directed graph. In other embodiments, other ways can also be used to determine the weight of each section edge in the line network directed graph, for example, the average of all weights in the first weight list corresponding to each section edge is taken as the weight of each section edge. The present application does not limit this.

[0108] It should be noted that in computer theory, the weight of a graph refers to a numerical value or a measure assigned to the edges of the graph, which is used to represent the relationship or connection distance, strength, cost, etc. between vertices. In this embodiment, the time consumption of passing between two vertices is taken as the weight of the edge between the two vertices. Each number in the first weight list L1 is a time length, such as seconds. Assuming that the first weight list L1 of an interval edge is [20, 25, 26, 28, 40], the median is 26, and 26 is the weight of the interval edge.

[0109] Step S4, according to the historical transactions and the train timetable, calculating the weight of the transfer edge in the line network directed graph.

[0110] The step S4 can specifically include:

[0111] Step S41, initializing an empty second weight list for each transfer edge in the line network directed graph;

[0112] Step S42, searching for transfer paths in the line network directed graph and constructing a transfer path set; the transfer path is a path composed of four stations and three edges, the second edge in the transfer path is a transfer edge, and the first and third edges are interval edges;

[0113] Step S43, for each transfer path in the transfer path set, determining the second weight list corresponding to the transfer edge in the transfer path according to the historical transactions and the train timetable;

[0114] Step S44, determining the weight of each transfer edge in the line network directed graph according to the second weight list corresponding to each transfer edge.

[0115] It should be noted that step S42 searches for all paths that meet the following characteristics in the line network directed graph as transfer paths: the path (represented as S1->S2->S3->S4) only contains four stations (S1, S2, S3, S4) and three edges (S1->S2, S2->S3, S3->S4 in turn), wherein the first edge S1->S2 and the third edge S3->S4 are interval edges, and the second edge S2->S3 is a transfer edge. All the searched transfer paths are combined to form a transfer path set.

[0116] Among them, the step S43 can be implemented in the following way: for each transfer path in the transfer path set, the following operations are performed to determine the second weight list corresponding to the transfer edge in the transfer path:

[0117] find all transactions from the historical transactions in which the entry station is the first station in the transfer path and the exit station is the fourth station in the transfer path, and generate a transaction set;

[0118] For each transaction in the transaction set, extract the data of the entry line, the entry station, the entry time, the exit line, the exit station, and the exit time corresponding to the transaction, and find all pairs of train numbers that meet the preset conditions in the transaction day train timetable of the entry line and the exit line, wherein each pair of train numbers includes an entry train number on the entry line and an exit train number on the exit line.

[0119] For each pair of train numbers that meets the preset conditions, subtract the arrival time of the entry train number at the second station in the transfer path from the departure time of the exit train number at the third station in the transfer path to obtain a second value, and add the second value to the second weight list of the transfer edge in the transfer path.

[0120] The preset conditions include:

[0121] The entry train number passes through the first station and the second station in the transfer path in sequence;

[0122] The exit train number passes through the third station and the fourth station in the transfer path in sequence;

[0123] The departure time of the entry train number at the first station in the transfer path minus the entry time in the transaction is greater than a first preset time length;

[0124] The exit time in the transaction minus the arrival time of the exit train number at the fourth station in the transfer path is greater than a second preset time length.

[0125] The following takes the transfer path S1->S2->S3->S4 as an example to introduce how to determine the second weight list L2 corresponding to the transfer edge S2->S3 in the transfer path S1->S2->S3->S4.

[0126] Find all transactions in which the entry station is equal to S1 and the exit station is equal to S4 from the historical transactions, and generate a transaction set TransSet1.

[0127] For each transaction Trans0 in the transaction set TransSet1, extract the entry line Lo, the entry station S1, the entry time To, the exit line Ld, the exit station S4, and the exit time Td, and then find a pair of train numbers (TrainO, TrainD) that meet the following conditions in the transaction day train timetable of the entry line Lo and the exit line Ld:

[0128] 1) TrainO passes through S1 and S2 in sequence;

[0129] 2) TrainD passes through S3 and S4 in sequence;

[0130] 3) TrainO's departure time at S1 - arrival time To > first preset duration (e.g., 10 seconds);

[0131] 4) Departure time Td - Arrival time of TrainD at S4 > first preset duration (e.g., 10 seconds);

[0132] For any (TrainO, TrainD) that satisfies the above four conditions, calculate the second value X2 = TrainD's departure time in S3 - TrainO's arrival time in S2, and add the second value X2 to the second weight set L2 of the S2->S3 transfer edge.

[0133] In step S44, the values ​​in the second weight list for each transfer edge can be sorted, and the smallest and largest values ​​can be removed. The average value is then taken as the weight of that transfer edge. For example, 30% of the smaller values ​​and 30% of the larger values ​​can be removed, and the average of the remaining 60% can be taken. It is understandable that real transaction data contains many abnormal data points that deviate significantly from the reasonable range. Removing the highest and lowest values, while potentially shifting the average value, can generally improve the accuracy of the estimation.

[0134] Step S5: Based on the network anomaly data, delete the vertices and edges corresponding to the stations that have anomalies from the directed graph of the network, and construct a network topology graph.

[0135] Step S5 may specifically include:

[0136] For each blocked interval in the network anomaly data, delete the interval edge corresponding to the blocked interval from the directed graph of the network;

[0137] For each blocked station in the network anomaly data, if the station has a transfer edge connected to it in the directed graph of the network, then the transfer edge is deleted.

[0138] It should be noted that, assuming station X is an interchange station for Line 1 and Line 2, in the event of a network anomaly, only station X on Line 1, only station X on Line 2, or the entire interchange station X can be closed. If only station X on Line 1 is closed, the interchange side between station X on Line 1 and station X on Line 2 must be deleted.

[0139] In practical applications, the line network directed graph can be copied to obtain a line network directed graph copy, and then the vertices and edges in the line network directed graph copy are deleted according to the line network abnormal data to obtain a line network topology graph.

[0140] In step S6, a line network abnormal topology graph is constructed according to the line network topology graph, the online map, and the weights of the interval edges and transfer edges in the line network directed graph, and the weights of the edges in the line network abnormal topology graph are determined, wherein the edges in the line network abnormal topology graph include subway edges and external edges.

[0141] The step S6 can specifically include the following steps.

[0142] In step S61, a copy of the line network topology graph is obtained, and all edges in the copy are deleted to obtain an initial line network abnormal topology graph.

[0143] In step S62, for each pair of stations in the initial line network abnormal topology graph, if neither of the two stations is blocked, a first shortest path from a first station to a second station in the line network topology graph is searched, if the first shortest path is found, a new edge from the first station to the second station is added to the initial line network abnormal topology graph, the type of the edge is a subway edge, and the weight of the edge is equal to the sum of the weights of the edges on the first shortest path, if the first shortest path is not found, an online map interface is called to search for a driving time from the first station to the second station, and a new edge from the first station to the second station is added to the initial line network abnormal topology graph, the type of the edge is an external edge, and the weight of the edge is determined according to the driving time.

[0144] In step S63, the station vertices of the multi-line transfer stations in the initial line network abnormal topology graph are merged into one vertex, and all transfer edges are deleted to obtain the line network abnormal topology graph.

[0145] In step S62, for each pair of stations (O, D) in the initial line network abnormal topology graph, if neither of the two stations is blocked, a shortest path from O to D in the line network topology graph is searched by using the Dijkstra algorithm. If a shortest path PathMin1 is found, a new edge is added to the initial line network abnormal topology graph, the type of the edge is a subway edge, and the weight of the edge is the sum of the weights of the edges on PathMin1. If the shortest path is not found, an online map API (interface) is called to search for a driving time from O to D, and X=k*driving time+d is calculated as the weight of the new edge, and the type attribute of the edge is “external edge”. Wherein, k and d represent a magnification coefficient and an offset, respectively.

[0146] It can be understood that the weight of the path represents the willingness of the passenger to select the path, and the smaller the weight, the greater the selection willingness. For example, from A to B, the subway is 30 minutes, and the car is 10 minutes. However, the car is more expensive, and the general passenger selection willingness will decrease. At this time, the weight of the external edge of the car needs to be enlarged. Therefore, the driving time is transformed by k*x+d in the embodiment, and is used as the weight of the external edge.

[0147] Step S7, based on the historical transactions, determining the overflow passenger flow generated by the subway incident according to the line network topology graph and the line network abnormal topology graph.

[0148] Specifically, the step S7 includes:

[0149] Step S71, initializing an empty list, each element in the empty list including: starting station, ending station, and 24-hour hour number;

[0150] Step S72, for each transaction in the historical transactions, determining each element in the empty list according to the line network topology graph and the line network abnormal topology graph;

[0151] Step S73, taking the triplets composed of the three attributes of the elements in the empty list as the key values, counting the number of corresponding elements in the empty list to obtain the transaction number, and dividing each transaction number by the number of days M to obtain an overflow passenger flow list, each element in the overflow passenger flow list including: starting station, ending station, 24-hour hour number, and transaction number.

[0152] The overflow passenger flow list is a set of overflow passenger flows generated by the subway incident.

[0153] The "24-hour hour number" refers to the hour number, ranging from 0 to 23. The application finally uses the 24 numbers from 0 to 23 as the aggregation dimension of the passenger flow. For example, if the passenger flow occurs at 13:24:57, the passenger flow is aggregated in the time period from 13:00 to 14:00, which is represented by 13.

[0154] In the step S72, for each transaction in the historical transactions of M days, the following operations are performed to determine each element in the empty list:

[0155] Extracting the entry station, exit station, and entry time of the transaction;

[0156] Taking the entry station as the starting point and the exit station as the ending point, searching for the second shortest path in the line network topology graph; if the second shortest path is found, skipping the subsequent operations and continuing to process the next transaction;

[0157] If the second shortest path is not found in the line network topology graph, a third shortest path is found in the line network abnormal topology graph, starting from the inbound station and ending at the outbound station;

[0158] For each external edge on the third shortest path, the external edge is pointed to by a third station and points to a fourth station, a time T corresponding to the external edge is calculated, a record R is generated, attributes of the record R include the third station, the fourth station, and a time T corresponding to the third station, and the record R is added to the empty list.

[0159] The time T is equal to the inbound time of the transaction plus the sum of the time consumption of all edges of the second shortest path from the inbound station to the third station, wherein the time consumption of the subway edge is equal to the weight of the edge, and the time consumption of the external edge is equal to the driving time corresponding to the external edge.

[0160] The time T represents the time when the passenger enters the external edge, that is, the actual start time of the overflow passenger flow.

[0161] The application will be described in detail below in combination with the specific embodiments shown in the accompanying drawings 2-5.

[0162] First, referring to FIG. 2, a local structure of a conventional subway line network graph is shown, which includes two stations B and E of line 2, one station J of line 3, and a transfer station F of line 2 and line 3. Other areas of the subway line network (i.e., line network east, line network south, line network west, and line network north areas) are represented by dashed lines.

[0163] Referring to FIG. 3, a line network directed graph G0 is shown. The line network directed graph G0 includes all stations in FIG. 2, wherein the 2 / 3 line transfer station F corresponds to two vertices L2F and L3F of the line network directed graph G0. Non-transfer stations L2B, L2E, and L3J each correspond to a vertex. There is a pair of directed edges between the vertices. Among them, the type of the edge between L2F and L3F is a transfer edge, represented by a dashed line; the types of the remaining edges are interval edges, represented by solid lines.

[0164] Referring to FIG. 3, taking the interval edge L2B->L2E as an example, the calculation method of the weight of the interval edge is described in detail.

[0165] Step 301: initialize an empty first weight list ListC1 for the interval edge L2B->L2E.

[0166] Step 302: for each train of line 2, perform the following operations: sort the arrival and departure triplets list according to the arrival time.

[0167] Step 303: For each train of line 2, do the following: traverse the arrival-departure triplets of each train, if two adjacent triplets (L2B, arrival time 1, departure time 1) and (L2E, arrival time 2, departure time 2) are encountered, calculate the value t = arrival time 2 - arrival time 1, the unit of t is second, then put t into ListC1.

[0168] Suppose there are 200 trains passing through the interval L2B -> L2E in the train timetable, then ListC1 will finally have 200 values.

[0169] Step 304: Calculate the median of the values in ListC1, and take the median as the weight of the interval edge L2B -> L2E.

[0170] Referring to FIG. 3, taking the transfer edge L3F -> L2F as an example, the calculation method of the weight of the transfer edge is described in detail.

[0171] Step 401: For the transfer edge L3F -> L2F, initialize an empty second weight list ListC2.

[0172] Step 402: Search for a path in G0 that satisfies the following conditions:

[0173] 1) The path only contains four stations, S1 -> L3F -> L2F -> S4.

[0174] 2) S1 -> L3F and L2F -> S4 are interval edges.

[0175] The above path that meets the conditions is constructed as a transfer path set PathSet.

[0176] Referring to FIG. 3, it can be seen that the path L3J -> L3F -> L2F -> L2E is in the set PathSet, denoted as Path1. The following steps take Path1 as an example to describe in detail how Path1 affects the weight of the transfer edge L3F -> L2F.

[0177] Step 403: Find all transactions from historical transactions where the entry station is equal to L3J and the exit station is equal to L2E, and generate a transaction set TransSet1.

[0178] For each transaction Trans0 in TransSet1, find a pair of trains (Train3, Train2) in the transaction day timetable of line 3 and line 2 that satisfy the following conditions:

[0179] 1) Train3 is a train of line 3, which passes through L3J and L3F in turn.

[0180] 2) Train2 is the train number of line 2, which successively passes through L2F and L2E.

[0181] 3) Train3 is the departure time of Train3 at L3J - the passenger arrival time at the station > 10 seconds.

[0182] 4) The passenger departure time - the arrival time of Train2 at L2E > 10 seconds.

[0183] For any pair of trains satisfying all the above conditions, a second value X2 = Train2 departure time at L2F - Train3 arrival time at L3F is calculated, the unit of the second value X2 is second, and then the second value X2 is added to ListC2.

[0184] Step 404: Sort the values in ListC2, delete the data with the smallest 30% value and the data with the highest 30% value, and take the average of the remaining values. The average value is the weight of the transfer edge L3F->L2F.

[0185] Fig. 4 is a diagram of the line network in which the interval between L3J and L3F is out of order, and the bidirectional interval is blocked. At this time, the directed edges L3J->L3F and L3F->L3J corresponding to the interval are deleted from the line network directed graph G0, and the line network topology graph G1 is obtained.

[0186] Referring to Figs. 4 and 5, the generation method of the line network abnormal topology graph G2 is described in detail.

[0187] Step 601, copy the line network topology graph G1 to obtain a line network abnormal topology graph G2, and delete all edges from the line network abnormal topology graph G2.

[0188] Step 602, for each pair of stations (O, D) in the line network abnormal topology graph G2, if both stations O and D are not blocked, and the shortest path O->D is found in G1 by Dijkstra algorithm. If a path PathMin1 is found, a new edge is added to the line network abnormal topology graph G2, the type of the edge is subway edge, and the weight of the edge = the sum of the weights of the edges on PathMin1. Otherwise, step 603 is performed. Those skilled in the art can easily understand the method of calling Dijkstra algorithm to find the shortest path in the directed graph.

[0189] In Fig. 4, the station pair (L2B, L3F) satisfies the above conditions, and the shortest path is L2B->L2E->L2F->L3F. Therefore, a new edge is added to the line network abnormal topology graph G2, the type of the edge is subway edge, and the weight of the edge = the weight of (L2B->L2E) + the weight of (L2E->L2F) + the weight of (L2F+L3F).

[0190] Step 603, call the online map service provider API to find the driving time t from O to D, and take t * k + d as the weight of the new edge, and the type attribute of the edge is "external edge". k and d represent the magnification coefficient and offset respectively, and a set of optional values of k and d are: k = 1.8, d = 300 seconds.

[0191] In FIG. 4, the station pair (L3J, L2B) does not satisfy the condition of step 602, so a new edge (L3J -> L2B) is added to the online network abnormal topology graph G2, and the type of the edge is external edge. Then call the online map service provider API to find the driving time t from L3J to L2B, and the unit of t is second, and take t * 1.8 + 300 as the weight of the edge (L3J -> L2B).

[0192] Step 604, merge the station vertices of the multi-line transfer station in the online network abnormal topology graph G2 into one vertex, and delete all transfer edges to obtain the final online network abnormal topology graph G2. In FIG. 4, the vertices of L3F and L2F need to be merged, and (L2F -> L3F) and (L3F -> L2F) are deleted to obtain the final online network abnormal topology graph G2, that is, FIG. 5.

[0193] Referring to FIGS. 4 and 5, the method for generating overflow passenger flow according to the online network abnormal topology graph G2 is described in detail.

[0194] Step 701: initialize an empty list ListFlow, and each element in ListFlow has attributes including start station, end station, and hour number in 24-hour format.

[0195] Step 702: for each transaction in the M-day historical passenger flow, perform the following operations:

[0196] 1) extract the entry station, exit station, and entry time of the transaction. Assume that the entry station of the current transaction is L2E, and the exit station is L3J.

[0197] 2) find the second shortest path in the online network topology graph G1 with L2E as the start point and L3J as the end point. At this time, no path can be found in the online network topology graph G1.

[0198] 3) find the third shortest path in the online network abnormal topology graph G2 with E as the start point and J as the end point. At this time, the third shortest path E -> B -> J is found.

[0199] 4) there is an external edge B -> J on the path E -> B -> J. For the external edge B -> J, calculate the time T of the external edge = entry time + (E -> B) edge time weight.

[0200] 5) generate a record R, the attributes of R include the start station B, the end station J, the hour number corresponding to the time T, and add R to the empty list ListFlow.

[0201] Step 703, count the number of corresponding elements in ListFlow (i.e. the transaction number) according to the key value of the empty list ListFlow as a three-tuple of (start station, end station, hour number in 24-hour system), and divide each number by the number of days M to obtain a new list LF, each element in the list LF is a four-tuple of (start station, end station, hour number in 24-hour system, transaction number). LF is the overflow passenger flow caused by the partial interruption of the subway operation bureau.

[0202] Based on the same inventive concept, an embodiment of the present application also provides an electronic device, as shown in Figure 6, comprising a processor 301, a communication interface 302, a memory 303 and a communication bus 304, wherein the processor 301, the communication interface 302 and the memory 303 complete communication with each other through the communication bus 304;

[0203] The memory 303 is used to store a computer program.

[0204] The processor 301 is used to execute the program stored on the memory 303, and realize the steps of the overflow passenger flow estimation method in the above-mentioned embodiment.

[0205] In another embodiment provided by the present application, a computer readable storage medium is also provided, the computer readable storage medium stores a computer program, and the computer program is executed by a processor to realize the steps of the overflow passenger flow estimation method in the above-mentioned embodiment.

[0206] In another embodiment provided by the present application, a computer program product containing instructions is also provided, which, when running on a computer, causes the computer to execute the steps of the overflow passenger flow estimation method in the above-mentioned embodiment.

[0207] Compared with the prior art, the technical scheme of the present application has the following innovative points:

[0208] Innovative point 1: according to the historical passenger flow and the timetable, the transfer time of each transfer station is calculated;

[0209] Innovative point 2: construct a line network abnormal topology graph G2, divide the edges in G2 into subway edges and external edges, and calculate the weight values of the subway edges and the external edges respectively by using the minimum distance algorithm and the online map;

[0210] Innovative point 3: simulate the walking of passengers in the abnormal line network graph by using the historical passenger flow and G2, take the external edges as the channels of the overflow passenger flow, and output a specific time attribute for each overflow passenger flow.

[0211] The computing scheme of the present application has the following beneficial effects: based on historical passenger flow, train timetable, abnormal topology graph of line network, real-time traffic data of online map, the overflow passenger flow is calculated, the generated data has high accuracy, real-time and interpretability in space and time, can support the rescheduling of the external traffic system of the subway in a finer granularity, and can more effectively evacuate the overflow passenger flow.

[0212] Although the present application has been described in detail by the above preferred embodiments, it should be recognized that the above description should not be considered as limiting the present application. Various modifications and alternatives of the present application will be apparent to those skilled in the art after reading the above description. Therefore, the protection scope of the present application should be defined by the appended claims.

Claims

1. A method for estimating overflow passenger flow, characterized in that, The method comprises the following steps: According to the subway line network data, a line network directed graph is constructed, wherein the vertices of the line network directed graph are stations, and the edges of the line network directed graph include interval edges between stations and transfer edges in transfer stations; Obtain historical transactions, train schedules of similar dates in the past M days closest to the current date, and obtain current line network abnormal data; According to the train schedule, the weight of the interval edge in the line network directed graph is calculated; According to the historical transactions and the train schedule, the weight of the transfer edge in the line network directed graph is calculated; According to the line network abnormal data, the vertices and edges corresponding to the abnormal stations in the line network directed graph are deleted to construct a line network topology graph; According to the line network topology graph, the online map, and the weights of the interval edges and the transfer edges in the line network directed graph, a line network abnormal topology graph is constructed, and the weight of each edge in the line network abnormal topology graph is determined, wherein the edges in the line network abnormal topology graph include subway edges and external edges; Based on the historical transactions, the line network topology graph and the line network abnormal topology graph are used to determine the overflow passenger flow generated by the subway incident.

2. The spill passenger flow estimation method of claim 1, wherein, The subway line network data includes a station list and an interval list; The attributes of each record in the station list include line name and station name, wherein there are N records of the N-line transfer station in the station list, and the line names are different and the station names are the same in the N records; The attributes of each record in the interval list include the starting line name, the starting station name, the terminal line name, and the terminal station name.

3. The spill passenger flow estimation method of claim 2, wherein, The line network directed graph is constructed according to the subway line network information, comprising: Initializing the line network directed graph, wherein the vertex set and the edge set are empty; For each station in the station list, a corresponding vertex is added to the vertex set of the line network directed graph; For each interval in the interval list, an edge is added to the edge set of the line network directed graph, and the type of the edge is an interval edge; For the N-line transfer stations in the station list, a pair of directed edges is added between the corresponding N station vertices in the line network directed graph, and the type of the edge is a transfer edge.

4. The spill passenger flow estimation method of claim 1, wherein, The attributes of each transaction in the historical transactions include the entering line name, the entering station name, the entering time, the exiting line name, the exiting station name, and the exiting time; The train schedule includes a plurality of train numbers and a corresponding arrival and departure triple list for each train number, wherein the attributes of the train number include the line name and the train number, and the attributes of each triple in the arrival and departure triple list include the station name, the arrival time, and the departure time.

5. The spill passenger flow estimation method of claim 4, wherein, The weight of the interval edge in the line network directed graph is calculated according to the train schedule, comprising: Initializing an empty first weight list for each interval edge in the line network directed graph; According to all train numbers and their corresponding arrival and departure triple lists in the train schedule, the first weight list corresponding to each interval edge is determined; According to the first weight list corresponding to each interval edge, the weight of each interval edge in the line network directed graph is determined.

6. The spill passenger flow estimation method of claim 5, wherein, According to all train numbers and their corresponding arrival and departure triple lists in the train schedule, the first weight list corresponding to each interval edge is determined, comprising: For each train in the train schedule, the following operations are performed to determine a first weight list corresponding to each interval edge: Sort all triplets in the arrival and departure triplet list of the train according to arrival time; For two adjacent triplets, find the corresponding interval edge in the line network directed graph according to the station names corresponding to the two triplets, and add a first value to the first weight list corresponding to the interval edge, the first value being equal to the arrival time corresponding to the latter triplet minus the arrival time corresponding to the former triplet.

7. The spill passenger flow estimation method of claim 5, wherein, The determination of the weight of each interval edge in the line network directed graph according to the first weight list corresponding to each interval edge comprises: Determine the median of all values in the first weight list corresponding to each interval edge, and set the median as the weight of each interval edge in the line network directed graph.

8. The spill passenger flow estimation method of claim 4, wherein, The calculation of the weight of each transfer edge in the line network directed graph according to the historical transactions and the train schedule comprises: Initialize an empty second weight list for each transfer edge in the line network directed graph; Search for transfer paths in the line network directed graph and construct a transfer path set; the transfer path is a path composed of four stations and three edges, the second edge in the transfer path being a transfer edge, and the first and third edges being interval edges; For each transfer path in the transfer path set, determine the second weight list corresponding to the transfer edge in the transfer path according to the historical transactions and the train schedule; Determine the weight of each transfer edge in the line network directed graph according to the second weight list corresponding to each transfer edge.

9. The spill passenger flow estimation method of claim 8, wherein, The determination of the second weight list corresponding to the transfer edge in each transfer path in the transfer path set according to the historical transactions and the train schedule comprises: For each transfer path in the transfer path set, the following operations are performed to determine the second weight list corresponding to the transfer edge in the transfer path: Find all transactions from the historical transactions in which the entry station is the first station in the transfer path and the exit station is the fourth station in the transfer path, and generate a transaction set; For each transaction in the transaction set, extract the data of the entry line, entry station, entry time, exit line, exit station, and exit time corresponding to the transaction, and find all pairs of trains in the train schedule of the entry line and the exit line on the transaction day that meet a preset condition, wherein each pair of trains includes an entry train on the entry line and an exit train on the exit line; For each pair of trains that meet the preset condition, subtract the departure time of the exit train at the third station in the transfer path from the arrival time of the entry train at the second station in the transfer path to obtain a second value, and add the second value to the second weight list of the transfer edge in the transfer path.

10. The spill passenger flow estimation method of claim 9, wherein, The preset condition comprises: The entry train passes through the first station and the second station in the transfer path in sequence; The exit train passes through the third station and the fourth station in the transfer path in sequence; The departure time of the inbound train at the first station in the transfer path minus the inbound time in the transaction is greater than a first preset time length; The outbound time in the transaction minus the arrival time of the outbound train at the fourth station in the transfer path is greater than a second preset time length.

11. The spill passenger flow estimation method of claim 8, wherein, The determining of the weight of each transfer edge in the rail network directed graph according to the second weight list corresponding to each transfer edge comprises: The values in the second weight list of each transfer edge are sorted, and the average value obtained by deleting a certain number of minimum values and a certain number of maximum values is taken as the weight of the transfer edge.

12. The spill passenger flow estimation method of claim 1, wherein, The rail network abnormal data comprises a list of blocked stations and a list of blocked sections; the attributes of each blocked station in the list of blocked stations comprise a line name and a station name, and the attributes of each blocked section in the list of blocked sections comprise a section start station and a section end station; for stations of different lines at a transfer station, whether to be closed can be set separately.

13. The spill passenger flow estimation method of claim 12, wherein, The constructing of the rail network topology graph from the rail network abnormal data by deleting the vertex and edge corresponding to the abnormal station in the rail network directed graph comprises: For each blocked section in the rail network abnormal data, the section edge corresponding to the blocked section is deleted from the rail network directed graph; For each blocked station in the rail network abnormal data, if the station has a transfer edge connected thereto in the rail network directed graph, the transfer edge is deleted.

14. The spill passenger flow estimation method of claim 1, wherein, The constructing of the rail network abnormal topology graph according to the rail network topology graph, the online map, and the weights of the section edges and the transfer edges in the rail network directed graph, and the determining of the weights of the edges in the rail network abnormal topology graph comprise: A copy of the rail network topology graph is copied, and all edges in the copy are deleted to obtain an initial rail network abnormal topology graph; For each pair of stations in the initial rail network abnormal topology graph, if neither of the two stations is blocked, a first shortest path from the first station to the second station is found in the rail network topology graph; if the first shortest path is found, an edge is added in the initial rail network abnormal topology graph, the edge is directed from the first station to the second station, the type of the edge is a subway edge, and the weight of the edge is equal to the sum of the weights of the edges on the first shortest path; if the first shortest path is not found, a driving time from the first station to the second station is found by calling an online map interface, and an edge is added in the initial rail network abnormal topology graph, the edge is directed from the first station to the second station, the type of the edge is an external edge, and the weight of the edge is determined according to the driving time. The rail network abnormal topology graph is obtained by merging the station vertices of the multi-line transfer stations in the initial rail network abnormal topology graph into one vertex and deleting all transfer edges.

15. The spill passenger flow estimation method of claim 1, wherein, The determining of the overflow passenger flow generated by a subway emergency event according to the rail network topology graph and the rail network abnormal topology graph based on the historical transactions comprises: An empty list is initialized, and each element in the empty list comprises a start station, an end station, and a 24-hour hour number. For each transaction in the historical transactions, determine each element in the empty list according to the line network topology graph and the line network abnormal topology graph; The empty list is used as a key value in the form of a three-tuple of three attributes of the element, the number of transactions in the empty list is counted to obtain the number of transactions, and each transaction number is divided by the number of days M to obtain the overflow passenger flow list, and the attributes of each element in the overflow passenger flow list include: starting station, ending station, hour number in 24-hour system, and transaction number.

16. The spill passenger flow estimation method of claim 15, wherein, The method for determining each element in the empty list according to the line network topology graph and the line network abnormal topology graph for each transaction in the historical transactions comprises the following steps: For each transaction in the historical transactions of M days, the following operations are performed to determine each element in the empty list: Extract the inbound station, outbound station and inbound time of the transaction; Take the inbound station as the starting point and the outbound station as the ending point to find the second shortest path in the line network topology graph; if the second shortest path is found, skip the following operation and continue to process the next transaction; If the second shortest path is not found in the line network topology graph, take the inbound station as the starting point and the outbound station as the ending point to find the third shortest path in the line network abnormal topology graph; For each external edge on the third shortest path, the external edge is pointed to by a third station and a fourth station, calculate the time T corresponding to the external edge, generate a record R, the attributes of the record R include the third station, the fourth station and the hour number corresponding to the time T, and add the record R to the empty list; The time T is equal to the inbound time of the transaction plus the sum of the time consumption of all edges of the second shortest path from the inbound station to the third station, wherein the time consumption of the subway edge is equal to the weight of the edge, and the time consumption of the external edge is equal to the corresponding driving time.

17. An electronic device, comprising: The computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the method steps of any one of claims 1-16. The computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the method steps of any one of claims 1-16. The computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the method steps of any one of claims 1-16.

18. A computer-readable storage medium, characterized in that, ​

Images