A Dynamic Scheduling Method for Ticket Checking Tasks at Railway Passenger Stations Based on Ant Colony Algorithm

Abstract

The present invention provides a dynamic scheduling method for ticket checking tasks at railway passenger stations based on the ant colony algorithm, including compiling relevant information according to the scheduling plan of ticket inspectors at railway stations, and establishing the working time segments of each ticket gate at the railway station; constructing a consecutive spatio-temporal network of ticket inspectors and the working time segments of ticket gates; establishing an optimization model for the dynamic scheduling plan of ticket inspectors at railway stations; using the ant colony search algorithm to solve the optimization model to obtain the optimal dynamic scheduling. The present invention aims at the dynamic post scheduling and layout mode of ticket gates at large railway stations, and realizes the rapid generation of the dynamic scheduling and layout plan of ticket inspectors; the present invention constructs a consecutive spatio-temporal network of ticket inspectors and the working time segments of ticket gates to facilitate the establishment of the optimization model; the present invention establishes an objective function through the minimization of comprehensive costs, takes into account the constraint conditions at the same time, and searches for the optimal dynamic scheduling scheme through the ant colony search algorithm, greatly improving the efficiency of ticket checking tasks.

Description

Technical Field

[0001] The present invention relates to the technical field of railway ticket checking scheduling, and in particular to a dynamic scheduling method for railway passenger station ticket checking tasks based on the ant colony algorithm. Background Art

[0002] The railway station ticket checking scheduling plan is an important part of the passenger transport management of railway passenger stations. Whenever a train departs from the station, the ticket checking gate corresponding to the train will arrange ticket inspectors to organize and guide passengers into the platform during the corresponding time period (usually 15 minutes to 5 minutes before the train departs). The preparation effect of the ticket inspector scheduling and posting plan is closely related to factors such as passenger service efficiency, station operation safety, and economic benefits.

[0003] At present, there are significant differences in the scheduling and posting management modes of ticket inspectors in stations under the jurisdiction of each railway administration. Most stations adopt the "fixed post" mode in the scheduling and posting of ticket inspectors, that is, each ticket checking gate of the station is assigned to a fixed ticket inspector, and the ticket inspector performs ticket checking tasks at this ticket checking gate throughout the day. The advantage of this mode is clear organization and simple implementation. The biggest disadvantage is that for stations with low train departure volumes, there will be a situation where there is no departing train at a certain ticket checking gate for a long time, resulting in a waste of human resources where ticket inspectors have nothing to do. To address this deficiency, a few railway stations implementing intelligent operation management have tried to implement a ticket inspector scheduling and posting method in the "dynamic post" mode. Under this mode, ticket inspectors are no longer fixedly arranged to work at a certain ticket checking gate, but flexibly arrange ticket inspectors to perform duties at different ticket checking gates according to the occupancy information of ticket checking gates at the train departure time. The advantage of this method is that it can effectively reduce the situation of ticket inspector idleness, effectively improve the work efficiency of ticket inspectors, and further reduce the number of ticket inspectors configured to reduce the station operation cost.

[0004] At present, there are very few studies on ticket checking scheduling methods under the "dynamic post" mode at home and abroad. Existing methods usually rely on the manual experience of the duty station master in charge of passenger transport at the station, that is, a ticket inspector scheduling and posting plan is formulated based on the train departure information of the station before station operation. However, the method adjusted based on manual experience is usually inefficient and brings great work pressure to the duty station master.

[0005] The problem of the railway station ticket inspector scheduling and posting plan under the "dynamic post" mode is a typical NP-Hard problem. Finding an exact optimal solution based on traditional operations research algorithms requires unacceptable time and cost and is not feasible in practice. Therefore, scholars in this field often strive to use approximate algorithms or heuristic intelligent optimization methods to quickly obtain solutions, although there are certain deviations in the quality of the obtained solutions from the optimal solution. At the same time, the existing auxiliary dispatching software mainly targets crew scheduling and cannot effectively assist in the scheduling and posting management decision-making of station ticket inspectors. Summary of the Invention

[0006] Aiming at the deficiencies of the prior art, the present invention provides a dynamic scheduling method for ticket checking tasks at railway passenger stations based on the ant colony algorithm, which realizes the rapid generation of a dynamic scheduling plan for ticket inspectors for the dynamic post scheduling layout mode at the ticket checking gates of large railway stations.

[0007] The technical solution of the present invention is: a dynamic scheduling method for ticket checking tasks at railway passenger stations based on the ant colony algorithm, including the following steps:

[0008] S1). According to the relevant information compiled in the ticket inspector scheduling plan of the railway station, determine the working time segments of each ticket checking gate of the railway station;

[0009] S2). Construct a continuous space-time network of ticket inspectors and working time segments of ticket checking gates;

[0010] S3). Establish an optimization model for the dynamic scheduling plan of ticket inspectors at railway stations;

[0011] S4). Use the ant colony search algorithm to solve the optimization model to obtain the optimal dynamic scheduling.

[0012] Preferably, in step S1), the relevant information compiled according to the ticket inspector scheduling plan of the railway station includes:

[0013] The set L of departure train information of the railway station, the set T of train departure time information, the set E of ticket checking gate information, and the set S of ticket inspector information; among them, the departure time of each train is , and the ticket checking gate for the departure of each train is .

[0014] Preferably, in step S1), the working time segments of each ticket checking gate of the railway station are based on the departure time and departure ticket checking gate of each train, and the working time segments of each ticket checking gate are expressed as:

[0015] ;

[0016] Among them, is the earliest advance passing time through the ticket checking gate before the train departs, is the latest advance passing time through the ticket checking gate before the train departs; is the departure time of each train

[0017] Obtain the set of the working time segments of the ticket checking gates of the railway station, where the duration of each working time segment is .

[0018] Preferably, in step S1), if it is the same ticket checking gate, if the working time segments determined by two adjacent departure trains If there is an overlapping time segment, then the two working time segments are merged into one working time segment .

[0019] Preferably, in step S2), a continuous spatio-temporal network of ticket inspectors and ticket gate working time segments is constructed based on the operations research graph theory method , where the spatio-temporal nodes represent the source / sink points formed by the ticket gate working time segments, and the arc segments represent the connection time between two consecutive ticket gate working time segments

[0020] Preferably, in step S2), the connection time between two consecutive ticket gate working time segments is calculated as follows

[0021] When the end time of the previous ticket gate working time segment is less than the start time of the next ticket gate working time segment, the connection time is the difference between the latter and the former

[0022] When the end time of the previous ticket gate working time segment is greater than or equal to the start time of the next ticket gate working time segment, the connection time is infinity; that is

[0023] ;

[0024] In the formula represents the connection time between the i th ticket gate and the j th ticket gate is the end time of the working time segment ; is the start time of the working time segment ; is the set threshold value is a sufficiently large positive integer

[0025] Preferably, in step S3), an optimization model for the dynamic scheduling plan of railway station ticket inspectors is established, which specifically includes the following steps

[0026] S31), On the constructed continuous spatio-temporal network of ticket inspectors and ticket gate working time segments, according to the actual requirements of the ticket inspector scheduling and posting at the station, a mixed integer linear integer optimization model is established

[0027] S32), Construct the constraint conditions of the optimization model

[0028] Preferably, in step S31), based on minimizing the comprehensive cost Z, the objective function of the optimization model is

[0029] ;

[0030] The objective function requires minimizing the number of ticket inspectors and the duration of ticket inspectors in working time segments;

[0031] In the formula, and is a 0-1 decision variable, Ticket inspector Whether to undertake work time fragments The ticket checking task, Indicates the ticket inspector Whether it is divided into working time segments Continuation of working time segment ; It is the conversion factor from the number of ticket inspectors to comprehensive fees; It is the conversion coefficient of the connecting time of the working time segment of the ticket gate to the comprehensive fee, and Z is the comprehensive fee.

[0032] Preferably, in step S32), the constraint condition is:

[0033] Constraint 1: Uniqueness constraint of the working time segment of the ticket gate

[0034] , ;

[0035] Constraint 2: Feasibility constraint of the continuation of working time segments at the ticket gate

[0036] , .

[0037] Preferably, in step S4), the optimization model is solved by using an ant colony search algorithm, which specifically includes the following steps:

[0038] S41), setting ant colony algorithm parameters;

[0039] S42), placing all ants in the virtual center, each ant randomly selects the next task node according to the transfer rule, and obtains their own complete path after traversing all time and space nodes;

[0040] S43), update the pheromones on each path, update the current number of iterations, determine whether the algorithm is finished, and output the optimal solution.

[0041] As a preferred embodiment, in step S41), the ant colony algorithm parameters set include the number of ants , pheromone importance parameters , parameters of the importance of heuristic factors , pheromone evaporation coefficient , Pheromone increase intensity coefficient , Maximum number of algorithm iterations .

[0042] Preferably, in step S42), it is specifically as follows:

[0043] S421), Place all ants at the virtual starting point. Each ant selects the ticket-checking gate working time segment with the earliest start time as the first spatio-temporal node, and randomly selects the next spatio-temporal node according to the following formula transfer rule. After all spatio-temporal nodes are selected and included in the ant search path, all ants continue to move towards the virtual end point, thereby obtaining their respective completion paths;

[0044] ;

[0045] Among them, represents the ant transferring from node to node probability, represents the ant at node set of feasible selection nodes; is the pheromone of arc segment ; is the heuristic information of arc segment ; is a parameter representing the importance degree of pheromone; is a parameter representing the importance degree of heuristic information.

[0046] Preferably, in step S43), it specifically includes the following steps:

[0047] S431), Update the pheromone on each path searched by the ants according to the following formula, that is:

[0048] ;

[0049] ;

[0050] Among them, is the pheromone concentration on path after the -th iteration; is the pheromone concentration on path after the -th iteration; is the pheromone evaporation coefficient; is the pheromone increment after this iteration, is the objective function value of the optimal path searched by the ant in this round, is the average objective function value of the paths searched by all ants in this round;

[0051] S432), update the current iteration count , if the current iteration count is less than or equal to the set maximum iteration count , then continue to repeat S21), otherwise the algorithm ends and outputs the optimal solution.

[0052] The beneficial effects of the present invention are as follows:

[0053] 1. The present invention aims at the dynamic post scheduling layout mode of ticket inspection gates in large railway stations, and realizes the rapid generation of the dynamic scheduling plan for ticket inspectors;

[0054] 2. The present invention constructs a continuous spatio-temporal network by combining ticket inspectors with the working time segments of ticket inspection gates to facilitate the establishment of an optimization model;

[0055] 3. The present invention establishes an objective function by minimizing the comprehensive cost, takes into account the constraint conditions at the same time, and searches for the optimal dynamic scheduling plan through the ant colony search algorithm, which greatly improves the efficiency of ticket inspection tasks. Description of the Drawings

[0056] Figure 1 is the overall flowchart of the embodiment of the present invention;

[0057] Figure 2 is the schematic diagram of the topological structure of the railway station in the embodiment of the present invention;

[0058] Figure 3 is the example diagram of the working time segments of all ticket inspection gates in the station in the embodiment of the present invention;

[0059] Figure 4 is the schematic diagram of the continuous spatio-temporal network of the working time segments of the ticket inspection gate in the embodiment of the present invention;

[0060] Figure 5 is the schematic diagram of the working time segments borne by a single ticket inspector in the embodiment of the present invention;

[0061] Figure 6 is the schematic diagram of the comparison post of the working hours of ticket inspectors in the embodiment of the present invention. Detailed Embodiment

[0062] The following further describes the detailed embodiment of the present invention with reference to the drawings:

[0063] As Figure 1 shown, this embodiment provides a dynamic scheduling method for ticket inspection tasks in railway passenger stations based on the ant colony algorithm. This embodiment uses the topological structure information of a certain railway station under the Guangzhou Railway Group and relevant train information (the trains departing from 20:00 to 8:00 on the night shift on May 15, 2024) as data to illustrate the present invention. The schematic diagram of the topological structure of this railway station is asFigure 2 As shown, this station is equipped with 13 ticket gates for passengers to board trains. According to the station conditions, the earliest and latest times to pass through the ticket gates before the train departs are respectively .

[0064] The method of this implementation specifically includes the following steps:

[0065] S1), respectively count the departing trains for the 13 ticket gates and form the working time segments of the ticket gates , and each working time segment of the ticket gate contains the train number and the corresponding ticket checking working period information; the specific method is as follows:

[0066] Determine the working time segment of the ticket gate according to the departure time of each train and the departure ticket gate , where is the earliest time to pass through the ticket gate in advance before the train departs, is the latest time to pass through the ticket gate in advance before the train departs. It should be noted that at the same ticket gate, if the working time segments and determined by two adjacent departing trains have overlapping time segments, for example , then the above two working time segments are merged into one working time segment .

[0067] Taking the No. 2 ticket gate as an example, the trains, train departure times and deduced ticket checking working periods it contains are shown in Table 1:

[0068] Table 1 Information related to the ticket checking working time period of the No. 2 ticket gate

[0069]

[0070] Form the initial working time segment of the No. 2 ticket gate as follows:

[0071] 19:58~20:09, 20:18~20:29, 20:30~20:41, 20:53~21:04, 21:31~21:42, 21:47~21:58; 22:14~22:25, 1:34~1:45, 4:37~4:48, 6:38~6:49, 6:59~7:18, 7:24~7:50;

[0072] Set the threshold , it can be found that in the working time segment of Ticket Check-in Gate 2, from 20:18 to 20:29 and from 20:30 to 20:41, there is only a 1-minute time difference between the end time of the previous working time segment and the start time of the next working time segment. Therefore, the working time segments from 20:18 to 20:29 and from 20:30 to 20:41 are merged into 20:18 to 20:41, and the working time segments from 21:31 to 21:42 and from 21:47 to 21:58 are merged into 21:31 to 21:58. Thus, the adjusted working time segments of Ticket Check-in Gate 2 are as follows, and the reduction in the number of working time segments reduces the problem scale.

[0073] 19:58~20:09, 20:18~20:41, 20:53~21:04, 21:31~21:58, 22:14~22:25;

[0074] 1:34~1:45, 4:37~4:48, 6:38~6:49, 6:59~7:18, 7:24~7:50;

[0075] Similarly, the working time segments of the other 12 ticket check-in gates are obtained, as Figure 3 shown, where the bold black ones are the adjusted working time segments.

[0076] The working time segments of Ticket Check-in Gate 1 are as follows:

[0077] 20:17~20:28, 21:57~22:08, 22:19~22:30, 23:01~23:12;

[0078] 0:30~0:41, 1:58~2:09, 6:24~6:50, 7:34~7:45;

[0079] The working time segments of Ticket Check-in Gate 3 are as follows:

[0080] 19:53~20:04, 20:13~20:24, 20:56~21:22, 22:47~22:58, 0:12~0:23;

[0081] 1:53~2:04, 3:05~3:16, 6:04~6:30, 7:12~7:33, 7:50~8:01;

[0082] The working time segments of Ticket Check-in Gate 4 are as follows:

[0083] 20:03~20:14, 20:23~20:34, 20:50~21:17, 21:26~21:52, 22:02~22:13, 22:52~23:17;

[0084] 0:07 to 0:18, 0:45 to 0:56, 1:09 to 1:20, 1:55 to 2:22, 5:53 to 6:04, 7:12 to 7:56;

[0085] The working time segments of Ticket Check-in Gate No. 5 are as follows:

[0086] 20:08 to 20:19, 20:50 to 21:01, 21:08 to 21:19, 21:36 to 21:47, 22:08 to 22:19, 22:41 to 22:52;

[0087] 23:14 to 23:25, 23:42 to 23:53, 0:25 to 0:36, 3:20 to 3:44, 6:48 to 7:11, 7:17 to 8:06;

[0088] The working time segments of Ticket Check-in Gate No. 6 are as follows:

[0089] 20:45 to 21:12, 21:52 to 22:03, 5:58 to 6:09, 6:33 to 6:54, 7:21 to 7:32, 7:56 to 8:07;

[0090] The working time segments of Ticket Check-in Gate No. 7 are as follows:

[0091] 20:36 to 20:47, 21:01 to 21:28, 21:48 to 22:11, 6:22 to 7:00;

[0092] The working time segments of Ticket Check-in Gate No. 8 are as follows:

[0093] 20:21 to 20:32, 20:40 to 21:36, 22:26 to 23:02, 5:48 to 5:59;

[0094] 6:16 to 6:27, 6:38 to 6:49, 6:55 to 7:06, 7:50 to 8:01;

[0095] The working time segments of Ticket Check-in Gate No. 9 are as follows:

[0096] 20:49 to 21:00, 21:14 to 21:45, 21:54 to 22:17, 22:24 to 22:35;

[0097] 22:45 to 22:56, 23:20 to 23:31, 6:27 to 6:38, 7:18 to 7:57;

[0098] The working time segments of Ticket Check-in Gate No. 10 are as follows:

[0099] 20:26 - 20:37, 21:01 - 21:17, 7:07 - 7:18;

[0100] The working time segments of Ticket Check-in Gate No. 11 are as follows:

[0101] 20:15 - 20:26, 20:42 - 20:53, 21:12 - 21:23, 21:35 - 21:46, 22:19 - 22:30;

[0102] 5:50 - 6:01, 6:32 - 6:58, 7:12 - 7:33, 7:57 - 8:08;

[0103] The working time segments of Ticket Check-in Gate No. 12 are as follows:

[0104] 6:15 - 6:26, 7:52 - 8:03;

[0105] The working time segments of Ticket Check-in Gate No. 13 are as follows:

[0106] 20:01 - 20:12, 21:57 - 22:08, 5:55 - 6:06, 7:42 - 7:53;

[0107] S2) Based on the graph theory method of operations research, establish the successive spatio-temporal network of the working time segments of the ticket check-in gates As Figure 4 shown, where the spatio-temporal nodes represent the source / sink points formed by the working time segments of the ticket check-in gates. The number of working time segments of the ticket check-in gates obtained in step S1) is 96. Therefore, the number of nodes in the spatio-temporal network is 96, and the arcs represent the successive time between two adjacent working time segments, where:

[0108] ;

[0109] Taking the working time segment 20:17 - 20:28 of Ticket Check-in Gate No. 1 as an example, its successive time with the working time segment 20:48 - 20:59 is 20 min, and its successive time with the working time segment 19:58 - 20:09 is .

[0110] In this way, the successive time matrix between all pairs of working time segments is obtained. It should be noted that the successive spatio-temporal network of the working time segments In the figure, the working time segment connection time matrix is not a symmetric matrix. Taking two working time segments from 19:58 to 20:09 and from 20:17 to 20:28 as an example, the connection time between 19:58 to 20:09 and 20:17 to 20:28 is 20 minutes. On the contrary, the end time of 20:17 to 20:28 and 19:58 to 20:09 is .

[0111] S3), Establish an optimization model for the dynamic scheduling plan of railway station ticket inspectors, specifically:

[0112] Based on minimizing the comprehensive cost Z, construct the objective function and constraint conditions of the optimization model, that is:

[0113] ;

[0114] The objective function requires minimizing the number of ticket inspectors and the connection time of ticket inspectors in working time segments;

[0115] In the formula, and are 0-1 decision variables, represents whether ticket inspector undertakes the ticket inspection task of working time segment , represents whether ticket inspector is connected from working time segment to working time segment ; is the conversion coefficient of the connection time of the working time segment of the ticket gate to the comprehensive cost, Z is the comprehensive cost, F Set of working time segments of the ticket gate; S is the set of ticket inspector information.

[0116] S4), Use the ant colony search algorithm to solve the optimization model, specifically including the following steps:

[0117] S41), Set the algorithm parameters, where the number of ants , the importance parameter of pheromone , the parameter of the importance degree of the heuristic factor , the pheromone evaporation coefficient , the pheromone increase intensity coefficient , the maximum number of iterations of the algorithm , set the current number of iterations .

[0118] S42), Place all ants at the virtual starting point , each ant selects the working time segment of the ticket gate with the earliest start time as the first spatio-temporal node, and randomly selects the next spatio-temporal node according to the following formula movement rule. After all spatio-temporal nodes are selected and included in the ant search path, all ants continue to move towards the virtual end point, thus obtaining their respective completion paths;

[0119] ;

[0120] Among them, is the pheromone of arc segment ; is the heuristic information of arc segment ; is the parameter representing the importance of pheromone; is the parameter representing the importance of heuristic information;

[0121] S43), update the pheromone on each path searched by the ants according to the following formula,

[0122] ;

[0123] ;

[0124] Among them, is the pheromone increment after this iteration, is the objective function value of the optimal path searched by the ant in this round, is the average objective function value of the paths searched by all ants in this round;

[0125] Update the current iteration number , if the current iteration number is less than or equal to the set maximum iteration number , then continue to repeat S2.1, otherwise the algorithm ends and outputs the optimal solution.

[0126] According to the above process, the following station ticket inspector scheduling and posting plan is obtained:

[0127] Note: The value outside the parentheses is the time span of the working time segment, and the value inside the parentheses is the ticket gate.

[0128] The working time segment for ticket inspector 1 to start work is:

[0129] 20:17 - 20:28 (1), 20:40 - 21:36 (8), 21:48 - 22:11 (7), 22:24 - 22:35 (9), 22:45 - 22:56 (9), 23:20 - 23:31 (9), 23:42 - 23:53 (5), 0:07 - 0:18 (4), 0:30 - 0:41 (1), 1:09 - 1:20 (4), 1:34 - 1:45 (2), 1:55 - 2:22 (4), 3:05 - 3:16 (3), 4:37 - 4:48 (2), 5:48 - 5:59 (8), 6:16 - 6:27 (8), 6:38 - 6:49 (8), 6:59 - 7:18 (2), 7:34 - 7:45 (1), 7:56 - 8:07 (6)

[0130] The total working duration is 301 min.

[0131] Ticket checker 2

[0132] 20:45 - 20:12 (6), 21:26 - 21:52 (4), 22:02 - 22:13 (4), 22:26 - 23:02 (8), 23:14 - 23:25 (5), 0:25 - 0:36 (5), 1:53 - 2:04 (3), 3:20 - 3:44 (5), 6:48 - 7:11 (5), 7:21 - 7:32 (6), 7:42 - 7:53 (13)

[0133] The total working duration is 202 min.

[0134] Ticket checker 3

[0135] 20:01 - 20:12 (13), 20:23 - 20:34 (4), 20:50 - 21:17 (4), 21:31 - 21:58 (2), 22:08 - 22:19 (5), 22:41 - 22:52 (5), 0:12 - 0:24 (3), 0:45 - 0:56 (4), 1:58 - 2:09 (1), 6:24 - 6:50 (1), 7:07 - 7:18 (10), 7:50 - 8:01 (8)

[0136] The total working duration is 179 min.

[0137] Ticket checker 4

[0138] 20:21 - 20:32 (8), 20:42 - 20:53 (11), 21:08 - 21:19 (5), 21:36 - 21:47 (5), 21:57 - 22:08 (1), 22:19 - 22:30 (1), 23:01 - 23:12 (1), 5:55 - 6:06 (13), 6:22 - 7:00 (7), 7:12 - 7:56 (4)

[0139] The total working hours are 170 min.

[0140] Ticket inspector 5

[0141] 20:03 - 20:14 (4), 20:26 - 20:37 (10), 21:01 - 21:17 (10), 21:35 - 21:46 (11), 21:57 - 22:08 (13), 22:19 - 22:30 (11), 22:52 - 23:17 (4), 5:53 - 6:04 (4), 6:15 - 6:26 (12), 6:38 - 6:49 (2), 7:24 - 7:50 (2)

[0142] The total working hours are 155 min.

[0143] Ticket inspector 6

[0144] 19:58 - 20:09 (2), 20:53 - 21:04 (2), 21:14 - 21:45 (9), 22:14 - 22:25 (2), 22:47 - 22:58 (2), 6:04 - 6:30 (3), 7:12 - 7:33 (3), 7:50 - 8:01 (3)

[0145] The total working hours are 133 min.

[0146] Ticket inspector 7

[0147] 19:53 - 20:04 (3), 20:18 - 20:41 (2), 20:56 - 21:22 (3), 21:54 - 22:17 (9), 6:27 - 6:38 (9), 6:55 - 7:06 (8), 7:18 - 7:57 (9)

[0148] The total working hours are 144 min.

[0149] Ticket inspector 8

[0150] 20:13 - 20:24 (3), 20:36 - 20:47 (7), 21:01 - 21:28 (7), 21:52 - 22:03 (6), 6:33 - 6:54 (6), 7:17 - 8:06 (5)

[0151] The total working hours is 130 min.

[0152] Ticket inspector 9

[0153] 20:08 - 20:19 (5), 20:50 - 21:01 (5), 21:12 - 21:23 (11), 5:50 - 6:01 (11), 6:32 - 6:58 (11), 7:12 - 7:33 (11), 7:57 - 8:08 (11)

[0154] The total working hours is 102 min.

[0155] Ticket inspector 10

[0156] 20:15 - 20:26 (11), 20:49 - 21:00 (9), 5:58 - 6:09 (6), 7:52 - 8:03 (12)

[0157] The total working hours is 44 min.

[0158] It can be found that in the initial station ticket inspector scheduling plan, there is a large gap in the total working hours between ticket inspector 1 and ticket inspector 10, and between ticket inspector 2 and ticket inspector 9. Therefore, the adjustment is as follows:

[0159] The working time segments for ticket inspector 1 to start work are:

[0160] 20:17 - 20:28 (1), 20:40 - 21:36 (8), 21:48 - 22:11 (7), 23:42 - 23:53 (5), 5:48 - 5:59 (8), 6:16 - 6:27 (8), 6:38 - 6:49 (8), 6:59 - 7:18 (2), 7:34 - 7:45 (1), 7:56 - 8:07 (6)

[0161] The total working hours is 175 min.

[0162] Ticket inspector 2

[0163] 20:45 - 20:12 (6), 21:26 - 21:52 (4), 22:02 - 22:13 (4), 23:14 - 23:25 (5), 0:25 - 0:36 (5), 1:53 - 2:04 (3), 3:20 - 3:44 (5), 6:48 - 7:11 (5), 7:21 - 7:32 (6), 7:42 - 7:53 (13)

[0164] The total working hours is 166 min.

[0165] Ticket inspector 9

[0166] 20:08 - 20:19 (5), 20:50 - 21:01 (5), 21:12 - 21:23 (11), 22:26 - 23:02 (8), 5:50 - 6:01 (11), 6:32 - 6:58 (11), 7:12 - 7:33 (11), 7:57 - 8:08 (11)

[0167] The total working hours is 138 min.

[0168] Ticket inspector 10

[0169] 20:15 - 20:26 (11), 20:49 - 21:00 (9), 22:24 - 22:35 (9), 22:45 - 22:56 (9), 23:20 - 23:31 (9), 0:07 - 0:18 (4), 0:30 - 0:41 (1), 1:09 - 1:20 (4), 1:34 - 1:45 (2), 1:55 - 2:22 (4), 3:05 - 3:16 (3), 4:37 - 4:48 (2), 5:58 - 6:09 (6), 7:52 - 8:03 (12)

[0170] The total working hours is 170 min.

[0171] Taking ticket inspector 2 as an example, the time segments of the ticket inspection work undertaken by him are as Figure 5 shown.

[0172] The working hours of all ticket inspectors are as Figure 6 shown, and it can be found that the difference in the working hours of ticket inspectors does not exceed 20%.

[0173] The above embodiments and descriptions in the specification only illustrate the principles and the best embodiments of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed.

Claims

1. A dynamic scheduling method for ticket checking tasks at railway passenger stations based on the ant colony algorithm, characterized in that, It includes the following steps: S1), compile relevant information according to the ticket inspector scheduling plan of the railway station, and establish the working time segments of each ticket gate in the railway station; S2), construct the continuous space-time network of ticket inspectors and ticket gate working time segments; specifically as follows: Constructing a continuous spatio-temporal network of ticket inspectors' and ticket-checking gate working time segments based on the graph theory method of operational research , where spatio-temporal nodes represent the source or sink points formed by the working time segments of ticket-checking gates, and arcs represent the connection time between two consecutive working time segments of ticket-checking gates; Among them, the connection time of the working time segments of the two ticket gates is The calculation is: When the end time of the working time segment of the previous ticket gate is less than the start time of the working time segment of the next ticket gate, the continuous time is the difference between the latter and the former; When the end time of the working time segment of the previous ticket gate is greater than or equal to the start time of the working time segment of the next ticket gate, the continuous time is infinite; that is: ； In the formula, represents the connection time between the i th ticket gate and the j th ticket gate; is the end time of the working time segment ; is the start time of the working time segment ; is the set threshold; is a sufficiently large positive integer; S3), establish an optimization model for the dynamic scheduling plan of ticket inspectors in the railway station; specifically including the following steps: S31), on the continuous space-time network of ticket inspectors and ticket gate working time segments constructed, establish a mixed integer linear integer optimization model according to the actual requirements of the ticket inspector scheduling and post arrangement in the station; Based on minimizing the comprehensive cost Z, the objective function of the optimization model is constructed as: ； The objective function requires minimizing the number of ticket inspectors and the continuous time of ticket inspectors in the working time segments; wherein, and are 0-1 decision variables, indicating whether the ticket inspector undertakes the ticket inspection task during the working time segment ; indicating whether the ticket inspector is followed by the working time segment to continue the working time segment ; is the conversion coefficient of the number of ticket inspectors to the comprehensive cost; is the conversion coefficient of the connection time of the working time segment at the ticket gate to the comprehensive cost, and Z is the comprehensive cost; S32), construct the constraint conditions of the optimization model; Constraint 1: Uniqueness constraint of ticket gate working time segments ， ； Constraint 2: Feasibility constraint of ticket gate working time segment continuity ， ； In the formula, F is the set of working time segments of the ticket checking gate, and S is the set of ticket checker information; S4), use the ant colony search algorithm to solve the optimization model to obtain the optimal dynamic scheduling.

2. The dynamic scheduling method for ticket checking tasks at railway passenger stations based on the ant colony algorithm according to claim 1 is characterized in that: In step S1), the working time segments of each ticket gate in the railway station are based on the departure time and departure ticket gate of each train, and the working time segments of each ticket gate are expressed as: ]； Among them, is the earliest time to pass through the ticket gate in advance before the train departs, is the latest time to pass through the ticket gate in advance before the train departs; is the departure time of each train; Obtain a set of working time segments of the ticket check gates at railway stations , where the duration of each working time segment is .

3. The dynamic scheduling method for ticket checking tasks at railway passenger stations based on the ant colony algorithm according to claim 2, characterized in that: In step S1), for the same ticket gate, if there is a time segment overlap between the working time segments determined for two adjacent departing trains and , the two working time segments are combined into one working time segment .

4. A dynamic scheduling method for ticket checking tasks at railway passenger stations based on the ant colony algorithm according to claim 1, characterized in that: In step S4), using the ant colony search algorithm to solve the optimization model specifically includes the following steps: S41), set the parameters of the ant colony algorithm; S42), place all ants at the virtual center, and each ant randomly selects the next task node according to the transfer rule. After traversing all space-time nodes, each ant obtains its own complete path; S43), update the pheromone on each path, update the current iteration number, judge whether the algorithm ends, and output the optimal solution.

5. A dynamic scheduling method for ticket checking tasks at railway passenger stations based on the ant colony algorithm according to claim 4, characterized in that: In step S41), the parameters of the ant colony algorithm set include the number of ants , the parameter of the importance degree of pheromone , the parameter of the importance degree of the heuristic factor , the pheromone evaporation coefficient , the pheromone increase intensity coefficient , the maximum number of iterations of the algorithm .

6. The dynamic scheduling method for ticket checking tasks at railway passenger stations based on the ant colony algorithm according to claim 5, wherein: Specifically in step S42) as follows: S421), place all ants at the virtual starting point, and each ant selects the working time segment of the ticket gate with the earliest start time as the first space-time node. According to the following formula of the transfer rule, randomly select the next space-time node. After all space-time nodes are selected and entered into the ant search path, all ants continue to move towards the virtual end point, so as to obtain their respective complete paths; Among them, represents an ant transferring from node to node with a probability of represents an ant in the set of feasible alternative nodes of node ; is the pheromone of path ; is the heuristic information of path ; is a parameter representing the importance degree of pheromone; is a parameter representing the importance degree of heuristic information.

7. A dynamic scheduling method for ticket checking tasks at railway passenger stations based on the ant colony algorithm according to claim 6, characterized in that: In step S43), it specifically includes the following steps: S431), update the pheromone on each path searched by the ants according to the following formula, that is: ； ； Among them, is the pheromone concentration on the path after the -th iteration; is the pheromone concentration on the path after the -th iteration; is the pheromone evaporation coefficient; is the pheromone increment after this iteration, is the objective function value of the optimal path found by the ant in this round of search, is the average objective function value of the paths searched by all ants in this round of search; S432), Update the current iteration count , if the current iteration count is less than or equal to the set maximum iteration count , then continue to repeat S421), otherwise the algorithm ends and outputs the optimal solution.

Images