Method for acquiring urban rail transit operation state information

[0092] It will be understood by those skilled in the art that the singular forms "a", "an", "the" and "the" as used herein can include the plural forms as well, unless expressly stated otherwise. It should be further understood that the word "comprising" used in the description of the present invention refers to the presence of stated features, integers, steps, operations, elements and/or modules, but does not exclude the presence or addition of one or more other features, Integers, steps, operations, elements, modules and/or groups thereof.

[0093] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art and, unless defined as herein, are not to be taken in an idealized or overly formal sense. explain.

[0094] In order to facilitate the understanding of the embodiments of the present invention, the following will take specific embodiments as examples for further explanation and description in conjunction with the accompanying drawings, and the embodiments do not constitute limitations to the embodiments of the present invention. Those skilled in the art should understand that the accompanying drawings are only schematic diagrams of the embodiments, and the components or processes in the accompanying drawings are not necessarily necessary to implement the present invention.

[0095] like Figure 1 to Figure 9 As shown, a method for obtaining operation status information of urban rail transit according to an embodiment of the present invention includes the following method steps:

[0096] S110: Search for a feasible initial physical path according to the travel time constraint condition recorded in the AFC data of automatic fare collection;

[0097] S120: constructing a time-expanded network for each of the initial physical paths, and searching the time-expanded network for a feasible path that satisfies a time constraint;

[0098] S130: Calculate the matching degree between the feasible path and the comprehensive time;

[0099] S140: Search for a ride plan with the highest matching degree based on a train capacity constraint;

[0100] S150: Acquire urban rail transit operation state information according to the ride plan, where the operation state information includes the ride plan, the passenger capacity of the train in the section, the number of people getting on and off at the station, and the number of times of passenger retention.

[0101] In a specific embodiment of the present invention, as Figure 10 As shown, the method for searching for a feasible initial physical path according to the travel time constraints recorded in the AFC data of automatic fare collection includes:

[0102] Initialize the AFC data and clear the candidate path set R optional set of data and initial physical paths for The data;

[0103] Using Dijkstra's algorithm to generate the shortest physical path r from the origin o to the destination d in the physical topology network G(N,A) shortest (o,d), if r shortest The time of (o,d) is greater than the travel time, i.e. then there is no feasible physical path, if r shortest (o,d) is less than or equal to the travel time, then r shortest (o,d) is replaced by the current path r current , while r current Join the initial physical path set

[0104] traverse r in order current the next node in the middle, and mark the current node as j∈r current , if j is the end point d, check the candidate path set R optional ,like then returns the initial physical path set If there is a candidate path set R optional , then for R optional path r in optional (k)∈R optional To make a validity judgment, if r optional The time cost of (k) is not satisfied then delete the path, if r optional The number of transfers of (k) does not satisfy the formula then delete the path, if r optional If there are duplicate nodes in (k), delete the path; if r optional If the node corresponding to the same transfer station appears repeatedly in (k), the path will be deleted. If r optional There is a transfer in (k), and the node of the same line appears again after the transfer, then delete the path; if j is not the end point d, traverse the set j of all connected nodes of the j node next , mark the current node as m∈j next , if m is in the current path, that is, m∈r current , then delete the arc a(j,m);

[0105] Use Dijkstra's algorithm to search for the shortest path r from node j to d shortest (j,d), if r shortest (j,d) does not exist, restore all deleted arcs, and traverse r current the next node in the middle, and mark the current node as j∈r current; if r shortest (j,d) exists, then judge r shortest Whether (j, d) already exists in the candidate path set, i.e. r shortest (j,d)∈R optional Whether it is true, if true, restore all deleted arcs and traverse r current the next node in the middle, and mark the current node as j∈r current , if not, get r current path r from o to j in current (o,j)∈r current , with r shortest (j,d) joins into a new path r connect (o,d), set r connect (o,d) is added to the candidate path set R optional =R optional ∪r connect (o,d), restore all deleted arcs, update the minimum number of transfers nt min;

[0106] extract R optional The shortest path in r optional_shortest , make a travel time constraint judgment, if the time cost of the shortest path is not greater than the travel time put R optional The shortest path in joins the initial physical path set At the same time, r optional_shortest set to the current path r current.

[0107] In a specific embodiment of the present invention, a method for constructing a time-expanded network for each of the initial physical paths, and searching for a feasible path satisfying a time constraint in the time-expanded network includes:

[0108] Read the current AFC record The initial set of physical paths for The ith physical path in Read the timetable T and build a time expansion network

[0109] Determine the first boarding time expansion node Whether the station attribute of , is consistent with the station attribute of the inbound node, that is, Is it true, if not, construct a virtual inbound node Will pit arc at the same time convert to t ow =t 1 +w(o,s o ), and finally set to

[0110] Determine the last alighting time to expand the node Whether the station attribute of the node is consistent with the station attribute of the outbound node, that is Is it true, if not, build a virtual outbound node Also will outbound arc convert to t dw =t 2 -w(s d ,d), and finally the set to

[0111] get path The number of transfers k (if there is no transfer in the path, k=0), extract the key station in the path KS={ks a |a=1,2,...,k+2}, in the same line l s(m) Generate line segments between key stations of

[0112] search from arrive "Earliest time expansion path" and "latest time expansion path", if the earliest time expansion path or the latest time to expand the path the physical path There is no feasible time expansion path in , and the algorithm ends; otherwise,

[0113] extract and Pick-up time expansion node in each segment s(m) Traverse the boarding time expansion node of each section s(m) For nodes that satisfy equation (3-38), call the segment search algorithm to search for subpaths in s(m) Extend the feasible boarding time to the node and alighting time expansion node Stored in the set of feasible pick-up nodes respectively and get off node collection

[0114] with segment arcs Connect the corresponding pick-up and drop-off time expansion nodes in the segment s(m), and at the same time with the sub-path form a map;

[0115] Set the current iteration number i=1, search from departure, collection If there is an inbound arc whose weight is not ∞, select the inbound arc with the smallest weight. Get the corresponding boarding time expansion node in section s(i) set nb for the current pickup point current;

[0116] Search segment arc Get a drop off node Determine whether the current segment is the last segment in the network, that is, whether i=k+1 is established, if so, obtain the outbound node through outbound arc search Record the nodes in the arc mapping subpaths of each segment into the current time expansion path tp in chronological order current , and set tp current Stored in the set of feasible time expansion paths Simultaneously make and nb current Weights for connecting inbound arcs or the weights of the swap arcs If not, search from departure, with For the transfer arc connected by the middle node, if there is an incoming arc whose weight is not ∞, select the one with the smallest weight. Get the boarding time expansion node of the next section s(i+1) Will set to nb current , and let i=i+1;

[0117] Determine the physical path ride plan Is it true, if not, extract the time expansion path Passengers in the available ride options The extracted objects are all boarding and alighting time expansion nodes, and the station attributes, train number attributes and station attributes of the alighting nodes are extracted respectively according to the chronological order, and the Store in the set of valid physical paths Get the set of all available rides for the passenger and the set of valid physical paths

[0118] In a specific embodiment of the present invention, the method for calculating the matching degree between the feasible path and the comprehensive time includes:

[0119] dividing the feasible routes into different travel types, and classifying all automatic fare collection AFC data according to the travel types;

[0120] Calculate the degree of matching between the feasible route and the travel time, that is, the degree of difference between the reference travel time of the feasible route and the actual travel time recorded by the AFC;

[0121] Calculate the matching degree between the feasible path and the random time, that is, the matching degree between the random time of the feasible path and the probability mass function of the random time;

[0122] According to the matching degree of the feasible path with the travel time and the random time, respectively, the matching degree between the feasible path and the comprehensive time is calculated. The above-mentioned random time includes the passenger's arrival time, the passenger's transfer time, the passenger's departure time, and the like.

[0123] In a specific embodiment of the present invention, the method for classifying all AFC data comprises the following steps:

[0124] All AFC data are divided into single physical path group and multi-physical path group according to the number of valid physical paths, and the multi-physical path group is regarded as a travel type, and the single physical path group includes only 1 valid physical path. pieces of AFC data, the multi-physical path group includes AFC data with at least two valid physical paths;

[0125] Dividing the AFC data in the single physical path group into three subgroups of no transfer, one transfer, and at least two transfers according to the number of passenger transfers;

[0126] The AFC data of the three subgroups and the multi-physical path group are divided into first travel type AFC data, second travel type AFC data, third travel type AFC data, and fourth travel type AFC data according to the number of feasible paths , AFC data of the fifth travel type, AFC data of the sixth travel type, and AFC data of the seventh travel type.

[0127] In a specific embodiment of the present invention, the AFC data of the first travel type is a single physical route, no transfer, and a single feasible route type, and the AFC data set of the first travel type is Q type1 Represents; the second travel type AFC data is a single physical route, no transfer, and multiple feasible route types, and the second travel type AFC data set uses Q type2 Represents; the third travel type AFC data is a single physical route, one transfer, and a single feasible route type, and the third travel type AFC data set uses Q type3 Represents; the fourth travel type AFC data is a single physical route, one transfer, multiple feasible route types, and the fourth travel type AFC data set uses Q type4 Indicates; the fifth travel type AFC data is a single physical route, at least two transfers, and a single feasible route type, and the fifth travel type AFC data set uses Q type5 Represents; the sixth travel type AFC data is a single physical route, at least two transfers, multiple feasible route types, and the sixth travel type AFC data set uses Q type6 Representation; the seventh travel type AFC data is multi-physical path, multi-feasible path type, and the seventh travel type AFC data set uses Q type7 express.

[0128] In a specific embodiment of the present invention, the method for calculating the matching degree between the feasible path and the travel time includes:

[0129] The formula for calculating the difference between the reference travel time of the feasible route and the actual travel time recorded by the AFC is:

[0130]

[0131] Among them, T min Represents the minimum value of the reference travel time in the ride plan, T travel -T min Indicates the offset range of the reference travel time, T travel -T reference (i) represents the offset value between the reference travel time and the actual travel time, obviously θ(i)∈[0,1], if T min =T travel , then the reference travel time of all feasible paths is consistent with the actual travel time recorded by AFC, and the degree of difference is 0; the relationship between the benefit value S and the degree of deviation θ(i) of the reference travel time is described by a normal distribution, and the normal distribution The probability density function of is as follows

[0132]

[0133] Among them, μ represents the expected value of the normal distribution, the present invention takes μ=0, the relationship between the benefit value S and the degree of difference only needs to take the part of the normal distribution θ≥μ, σ=0.5, and the benefit value is calculated based on the normal function. The function of S and the degree of difference is corrected as Feasible Path Set BP Q The sum of the median benefit value S is The matching degree of each set of feasible paths and travel time is

[0134] In a specific embodiment of the present invention, the method for calculating the matching degree between the feasible path and random time includes:

[0135] bp the feasible path Q (i) The degree of matching with random time is defined as the premise of considering only random time, bp Q (i) is the conditional probability of the actual path of the passenger

[0136]

[0137] Feasible paths for the second trip type The match to random time is:

[0138]

[0139] Feasible paths for the fourth trip type How well it matches a random time:

[0140]

[0141] Feasible paths for the sixth trip type How well it matches a random time:

[0142]

[0143] Feasible paths for the seventh trip type How well it matches a random time:

[0144]

[0145] In a specific embodiment of the present invention, the formula for calculating the matching degree between the feasible path and the comprehensive time is:

[0146]

[0147] In a specific embodiment of the present invention, the search and calculation method for the ride plan with the highest matching degree based on the train capacity constraint includes:

[0148] Step 1: Define the Train Fragment

[0149] TrainSeg={trainseg i (trnum,s d ,t d ,s a ,t a ,load max ,load current )|i=1,2,...,N T},

[0150] where trnum ∈ L, s d ,s a ∈S, t d ,t a ∈T. trnum represents the number of trains, L represents the set of all trains in all lines, s d and s a Respectively represent the train operating interval (s d ,s a ), S represents the set of stations, t d and t a respectively represent the train at s d and s a The departure time and arrival time of , T represents the set of train schedules, load max and load current Respectively represent the maximum passenger capacity and the current passenger capacity of the train; the maximum passenger capacity load max =N c ×ω max , where N c Indicates the train capacity, ω max Represents the maximum full load rate of the train; train capacity is defined as the difference between the maximum passenger capacity of the train and the current passenger capacity, namely T c =load max -load current;

[0151] Step 2: Read all AFC data in the study period, search for the passenger's ride plan and valid physical route corresponding to each set of AFC data, and divide each set of AFC data into each travel type Q. type1 , Q type2 , Q type3 , Q type4 , Q type5 , Q type6 , Q type7 middle;

[0152] Step 3: For the travel type Q type1 , Q type2 , Q type3 , Q type4 , Q type5 , Q type6 , Q type7 The AFC data in are sorted according to the entry time from morning to night;

[0153] Step 4: According to Q type1 , Q type3 and Q type5 Feasible path bp for each set of AFC data Q in Q the number of trips trnum Q , search all train segment sets TrainSeg that satisfy trnum=trnum Q collection of train fragments Effective physical path FR according to Q QGet the interval set Se that passengers pass through in the physical topology network Q ={se Q (i)=se Q (s i ,s i+1 )s i ,s i+1 ∈fr Q (k),fr Q (k)∈FR Q}, search train fragment collection all satisfy s d =s i ,s a =s i+1 The train segment collection TrainSeg final The identification result of the ride plan is obtained, which is expressed as update the train segment collection TrainSeg final All train fragments in trainseg final (u) current passenger capacity, i.e. load current =load current +1;

[0154] Step 5: Establish outbound time distribution, according to Q type1 , Q type3 and Q type5 The identification results of the ride plan in , for each station s e , establishing each time period T i from l i Departure time distribution of alighting

[0155] Step 6: Read Q in chronological order type2 For each set of AFC data, Q 2 ∈Q type2 , for each set of feasible paths First calculate the matching degree P with the travel time travel (i), then according to D egress Calculate the degree of match P with a random time rand om(i), and then calculate the matching degree P of each set of feasible paths and the comprehensive time time (i);

[0156] to the Q type2 All feasible paths in are sorted according to the comprehensive time matching degree According to the order of sorting, read the current feasible path The "key station" KS, get the passenger's pick-up station S from the key station board ={s board (v)v=1,...,N v}, while obtaining feasible paths at each station s board (v) the number trnum(v) of the boarding train;

[0157] Satisfy all of the TrainSeg trnum=trnum(v),s d =s board (v) train segment trainseg, determine its train capacity constraint T c 0 is true, if all train segments meet the train capacity constraints, update the passenger capacity of the corresponding train segment and obtain the identification result of the ride plan Otherwise, judge whether the current feasible path is the last set of feasible paths in ; if The last set of feasible paths in Otherwise read the next set of feasible paths;

[0158] Step 7: Build the pit time distribution, according to Q type2 The identification results of the ride plan in , for each station s a , establishing each time period T i from l i The distribution of pit stop times for boarding

[0159] Step 8: Read Q in chronological order type4 Each set of AFC records in Q 4 ∈Q type4 , for each set of feasible paths First calculate the matching degree P with the travel time travel (i), then according to D access and D egress Calculate feasible paths The degree of match P with a random time random (i), and then calculate the comprehensive time matching degree P of each group of feasible paths time (i);

[0160] to the Q type4 All feasible paths in are sorted according to the comprehensive time matching degree According to the order of sorting, the train capacity constraint judgment is made, and the identification result of the ride plan is obtained. And update the passenger capacity of the corresponding train segment;

[0161] Step 9: Establish the transfer time distribution, using Q type4 The identification results of the ride scheme in , for each transfer station s t , establishing each time period T i from l i to l j The transfer time distribution of

[0162] Step 10: Read Q in chronological order type6 Each set of AFC records in Q 6 ∈Q type6 , for each set of feasible paths First calculate the matching degree P with the travel time travel (i), then according to D access , D egress and D transfer Calculate feasible paths The degree of match P with a random time random (i), and then calculate the comprehensive time matching degree P of each group of feasible paths time (i);

[0163] to the Q type6 All feasible paths in are sorted according to the comprehensive time matching degree According to the order of sorting, the train capacity constraint judgment is made, and the identification result of the ride plan is obtained. And update the passenger capacity of the corresponding train segment;

[0164] Step 11: Read Q in chronological order type7 Each set of AFC records in Q 7 ∈Q type7 , for each set of feasible paths First calculate the matching degree P with the travel time travel (i), then according to D access , D egress and D transfer Calculate feasible paths The degree of match P with a random time random (i), and then calculate the comprehensive time matching degree P of each group of feasible paths time (i);

[0165] to the Q type7 All feasible paths in are sorted according to the comprehensive time matching degree According to the order of sorting, the train capacity constraint judgment is made, and the identification result of the ride plan is obtained. And update the passenger capacity of the corresponding train segment.

[0166] The method for identifying the passenger boarding plan of the urban rail transit network proposed by the present invention can obtain the full load rate of the corresponding trains in each section, the number of people getting on and off at each station, and the distribution of the number of passengers at the station, etc., which can describe the evaluation of the operation status of urban rail transit. index. Based on the identification results of the corresponding passenger travel plans in the AFC records of the morning rush hour on this weekday, taking the Beijing Metro Line 5 during the morning rush hour (7:00 to 9:00 in the morning) as an example, this method is applied to the line operation status evaluation , the operation status evaluation indicators used are mainly the full load rate of the train in each section and the retention rate of the station.

[0167] According to the identification results of the passenger boarding plan, the full load rate of some trains on Line 5 in each section and the number of people getting on and off at each station are obtained. like Figure 5 As shown, through the analysis of all trains, the stations with a large number of trains in the downward direction include Tiantongyuan North, Tiantongyuan, and Lishuiqiao. Datun East Road", "Datun Road East - South Exit of Huixin West Street" and "South Exit of Huixin West Street - North Exit of Huixin West Street"; in the downward direction, the sections with higher train full load rate are distributed in "Tiantongyuan South - Lama Temple" section, the section with the highest full load rate is the section "Huixin West Street North Exit - Huixin West Street South Exit", the maximum train full load rate is 1.21, and the full load rate of trains departing from Tiantongyuan North from 7:30 to 8:10 higher than other periods. like Image 6 As shown, the station with more trains in the upward direction is Songjiazhuang, and the sections with higher full load rate include "Liujiayao-Puhuangyu", "Puhuangyu-East Gate of the Temple of Heaven", "East Gate of the Temple of Heaven-Ciqikou", "Ciqikou" -Chongwenmen" and "Chongwenmen-Dongdan"; for the upward direction, the sections with higher train full load rate are distributed in the section "Songjiazhuang-Dengshikou", and the section with the highest full load rate is "Ciqikou-Chongwenmen" In the interval, the maximum train full load rate is 1.28, and the full load rate of trains departing from Songjiazhuang from 7:50 to 8:15 is higher than other time periods. It can be seen that the number of people getting on the train is generally larger at the stations with larger inbound volume (Tiantongyuan North, Tiantongyuan) and transfer volume (Lishuiqiao, Songjiazhuang).

[0168] like Figure 7-8 The average passenger retention rate in the up and down directions of all stations on Line 5 with a granularity of half an hour. It can be seen that the stations with a higher retention rate are mainly the stations with a large number of boarding people and the stations at both ends of the interval with a high train full load rate. . From the point of view of operation status, the number of passengers retained at each time period can better reflect the operational status than the number of individual passengers. The passenger retention rate can not only reflect the supply relationship between transportation capacity and passenger flow demand, but also reflect the number of people gathered on the platform, thereby reflecting the operational safety status, and providing a reference for the implementation of the station’s morning peak passenger flow control measures.

[0169] Through the evaluation of the operation status from the two aspects of train full load rate and station retention rate, during the morning rush hour, the operating pressure in the downward direction of Beijing Metro Line 5 is relatively large. For the upward direction, the transfer volume at Songjiazhuang Station is too large, and the full load rate of the upward trains departing from this station is relatively high. In this direction, the full load rate of the trains from Songjiazhuang to Dongdan is relatively high, causing the stations from Songjiazhuang to Chongwenmen to stay. The traffic rate is relatively high, and the operation risk is relatively high; for the downward direction, the number of incoming stations at Tiantongyuan North Station and Tiantongyuan Station and the transfer volume of Lishui Bridge are relatively large. The passenger retention rate of these stations at the North Exit of Huixin West Street is relatively high, and the operation risk is relatively high. From the point of view of train full load rate and retention rate, although Line 5 is under great operating pressure, the overall operation is in good condition, and the impact of large passenger flow in the morning peak on urban rail transit operations has been well limited.

[0170] By collecting the passenger capacity of all trains in each section in the corresponding time period, the section throughput of the section in this time period is obtained. The comparison of the cross-section (15-minute granularity) of Line 5 based on the operational statistics and the identification results of the ride plan is as follows: Figure 9 As shown in the figure, the change trend of the two is basically the same, but the difference between the two is large in the range of high section volume. There are many reasons for the difference. For example, setting the upper limit of the train's full load rate to 130% may be different from the actual situation, or it may be the difference between the travel time distribution and the actual situation, resulting in the identification result of the passenger travel plan and the actual existence. differences, which in turn lead to large differences in train full load ratios.

[0171]To sum up, the embodiment of the present invention can not only identify and obtain the travel plan corresponding to each set of AFC data, but also obtain the passenger capacity of the train in each section, the number of people getting on and off the train at each station, and the number of passengers who stay on board. The urban rail transit operation evaluation index improves the refined management level and operational efficiency of urban rail transit operations.

[0172] From the description of the above embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus a necessary general hardware platform. Based on this understanding, the technical solutions of the present invention can be embodied in the form of software products in essence or the parts that make contributions to the prior art. The computer software products can be stored in storage media, such as ROM/RAM, magnetic disks, etc. , CD, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

[0173] The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited to this. Substitutions should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.