Local railway passenger operation scheme decision-making method, device, equipment and product

Abstract

The invention discloses a local railway passenger operation scheme decision-making method, device, equipment and product, and relates to the technical field of railway operation. The method comprises the following steps of: firstly, taking whole-course ticket checking quantity time sequence data of a target local railway as whole-course passenger volume time sequence data, and predicting to obtain whole-course passenger volume time sequence prediction data of the target local railway in a plurality of continuous natural days in the future according to the whole-course passenger volume time sequence data; the method comprises the following steps: optimizing a passenger transport operation scheme of a target local railway in a plurality of continuous natural days in the future based on an optimization algorithm to obtain a passenger transport operation scheme with a shortest cosine distance, and finally taking the passenger transport operation scheme as an optimal passenger transport operation scheme of the target local railway in the plurality of continuous natural days in the future and outputting the optimal passenger transport operation scheme. Therefore, the method can be used for deciding a local railway passenger operation scheme, and the ticket supply quantity is matched with the passenger demand quantity to the maximum extent, so that the passenger operation attendance rate is improved, and local railway operation profit is facilitated.

Description

Technical Field

[0001] This invention belongs to the field of railway operation technology, specifically relating to a method, device, equipment, and product for making decisions on local railway passenger transport operation plans. Background Technology

[0002] Local railways, also known as local railways, refer to railway systems whose construction, operation, maintenance, and management are primarily undertaken by local government departments or enterprises, and are often used as branch lines.

[0003] In the management of local railway passenger transport, passenger volume is characterized by numerous influencing factors and significant fluctuations (e.g., holidays and weather conditions can greatly impact passenger volume). Furthermore, local railway passenger transport operation plans require advance decision-making and are characterized by stability but limited flexibility in adjustment. This easily leads to large fluctuations in passenger occupancy rates, which directly affect the profitability of local railway operations. Therefore, how to determine local railway passenger transport operation plans to maximize the match between ticket supply and passenger demand, thereby improving passenger occupancy rates, is a crucial research topic for those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide a method, apparatus, computer equipment, computer-readable storage product, and computer program product for making decisions on local railway passenger transport operation plans, in order to solve the problems of mismatch between ticket supply and passenger demand and large fluctuations in passenger occupancy rates in existing local railway passenger transport management plans.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] Firstly, a decision-making method for local railway passenger transport operation plans is provided, including:

[0007] Obtain time-series data of the total ticket checking volume of the target local railway. The time-series data of the total ticket checking volume contains multiple intraday unit time periods that are sequentially consecutive in time, as well as the historical ticket checking volume of each intraday unit time period within the multiple intraday unit time periods for the total ticket. The total ticket refers to a ticket from the passenger originating station to the passenger destination station.

[0008] The time-series data of the total ticket check volume is used as the time-series data of the total passenger volume. Based on the time-series data of the total passenger volume, the time-series prediction data of the total passenger volume of the target local railway for the next several consecutive natural days is obtained. The time-series data of the total passenger volume contains the historical passenger volume for the total ticket within a unit time period in each day. The time-series prediction data of the total passenger volume contains the predicted passenger volume for the total ticket within a unit time period in each of the next several consecutive natural days.

[0009] Using the time-series forecast data of the total passenger volume, an optimization algorithm is used to optimize the passenger operation plan of the target local railway for several consecutive future natural days, resulting in a passenger operation plan with the shortest cosine distance. The passenger operation plan includes the planned number of departures from the originating station to the terminating station within each unit time period of each future day, the planned passenger carriage type for each departure, and the planned number of carriages for each carriage type. The shortest cosine distance refers to the minimum cosine distance between the time-series forecast data of the total passenger volume and the time-series data of the total ticket supply calculated based on the passenger operation plan. The time-series data of the total ticket supply contains the supply of total tickets for each unit time period of each future day, and the supply is calculated according to the following formula:

[0010]

[0011] In the formula, x represents a positive integer, and P x Y represents the supply of the full-journey ticket within a unit time period in the xth future day. x This represents the total number of planned departures within a unit time period during the x-th future day, from the passenger originating station to the passenger terminating station, where y represents less than or equal to Y. x positive integers, Z x,y This represents the total number of planned passenger carriage types for the y-th scheduled departure within a unit time period in the x-th future day, where z represents less than or equal to Z. x,y positive integers, This represents the number of planned passenger carriages of the z-th type for the y-th planned departure during a unit time period within the x-th future day. This indicates the number of full-journey tickets that can be provided for each type of planned passenger carriage (z-th type).

[0012] The passenger transport operation plan with the shortest cosine distance is output as the optimal passenger transport operation plan for the target local railway for the next several consecutive natural days.

[0013] Based on the above-mentioned invention, a novel scheme for optimizing future passenger transport operation plans based on time-series data of total ticket checking volume is provided. First, the time-series data of total ticket checking volume of the target local railway is used as the time-series data of total passenger volume. Then, based on this time-series data, the time-series forecast data of total passenger volume for the target local railway over several consecutive natural days is obtained. Next, using this time-series forecast data, an optimization algorithm is applied to optimize the passenger transport operation plan for the target local railway over several consecutive natural days, resulting in a passenger transport operation plan with the shortest cosine distance. Finally, this passenger transport operation plan is output as the optimal passenger transport operation plan for the target local railway over several consecutive natural days. This can be used to decide on local railway passenger transport operation plans, maximizing the match between ticket supply and passenger demand, thereby improving passenger occupancy rates, increasing the profitability of local railway operations, and making local railway operations sustainable. This approach is convenient for practical application and promotion.

[0014] In one possible design, the time-series data of the total ticket checking volume is used as the time-series data of the total passenger volume, including:

[0015] Obtain the time-series data of the full-journey candidate cancellation volume of the target local railway, wherein the time-series data of the full-journey candidate cancellation volume contains the historical candidate cancellation volume for the full-journey tickets that depart within the unit time period of each day and were automatically cancelled due to non-payment within a preset time period before departure;

[0016] For each intraday unit time period, the corresponding historical ticket checking volume in the full-journey ticket checking volume time series data is added one-to-one to the corresponding historical candidate cancellation volume in the full-journey candidate cancellation volume time series data multiplied by a preset coefficient and rounded down to obtain the corresponding historical passenger volume for the full-journey ticket.

[0017] By summing up all the historical passenger volumes, the time-series data of the total passenger volume is obtained.

[0018] In one possible design, if there are intermediate passenger stations between the passenger departure station and the passenger destination station, then the time-series data of the total ticket checking volume is used as the time-series data of the total passenger volume, including:

[0019] Obtain time-series data of non-full-journey ticket checking volume for each non-full-journey section of the target local railway. Each non-full-journey section includes a section from the passenger originating station to each passenger intermediate station, a section from each passenger intermediate station to the passenger terminal station, and a one-way section from each passenger intermediate station to another passenger intermediate station. The time-series data of non-full-journey ticket checking volume contains historical ticket checking volume for non-full-journey tickets for trains departing from each intraday unit time period. The non-full-journey ticket refers to the ticket for the corresponding non-full-journey section.

[0020] For each of the non-full-journey segments, the historical ticket checking volume in the corresponding non-full-journey ticket checking volume time series data will be used as the historical non-full-journey passenger volume and broken down into historical short-distance passenger volumes corresponding to the corresponding intraday unit time period and between each pair of adjacent stations.

[0021] For each intraday unit time period and each pair of adjacent stations, the corresponding historical short-distance passenger volume is accumulated to obtain the corresponding historical short-distance passenger volume.

[0022] For each intraday unit time period, the corresponding historical ticket checking volume in the time series data of the total ticket checking volume is added one by one to the average, median or maximum value of the historical short-distance passenger volume of each pair of adjacent stations to obtain the corresponding historical passenger volume for the total ticket.

[0023] By summing up all the historical passenger volumes, the time-series data of the total passenger volume is obtained.

[0024] In one possible design, the time-series prediction data of the total passenger volume of the target local railway for several consecutive natural days is obtained based on the time-series data of the total passenger volume, including:

[0025] The total passenger volume time series data is divided into multiple sets of total passenger volume time series data that correspond one-to-one with multiple time windows that are sequentially consecutive in time series.

[0026] For the last time window among the multiple time windows, at least one set of full-journey passenger volume time-series data corresponding to at least one preceding time window is used as the corresponding training dataset, and one set of full-journey passenger volume time-series data corresponding to the corresponding window is used as the corresponding validation dataset.

[0027] A real-time dynamic prediction model for total passenger volume is established based on the Prophet model, expressed by the following formula:

[0028]

[0029] In the formula, t represents time, A(t) represents the total passenger volume at time t, g(t) represents the trend term at time t, s(t) represents the seasonal term at time t, h(t) represents the holiday term at time t, ε(t) represents the error noise term at time t, Q represents a positive integer, q represents a positive integer less than or equal to Q, and l q (t) represents the linear regression term at time t that corresponds to the qth linear influence factor among the Q linear influence factors;

[0030] The growth pattern of the trend term is set to linear, and natural weeks and natural days are added to make the seasonal term change periodically. The error noise term is also set to follow a mathematical expectation of μ and a variance of σ. 2 The normal distribution;

[0031] Add an influence weight adjustment factor to the holiday item to obtain a first model parameter vector consisting of the weight adjustment parameters of all holiday intervals in the holiday item;

[0032] Time series data of all factors related to the total passenger volume are obtained, and then the linear influencing factors of the total passenger volume are dynamically screened based on the time series data so as to determine the factors related to the total passenger volume that meet the screening conditions as the linear influencing factors, and a second model parameter vector composed of the weight adjustment parameters in the linear regression terms of all the linear influencing factors is obtained.

[0033] Using the training dataset and validation dataset corresponding to the last time window, the real-time dynamic prediction model for the entire passenger volume is trained and its parameters are optimized based on the particle swarm optimization algorithm to obtain the optimized search results of the model parameters that have passed the validation. The optimized search results of the model parameters include the optimized search results of the first model parameter vector, the second model parameter vector, the expected value, and the variance.

[0034] The optimized search results of the model parameters are input into the real-time dynamic prediction model of the total passenger volume, and then the real-time dynamic prediction model of the total passenger volume is applied to predict the passenger volume prediction value for the total ticket for each unit time period in each future day in multiple consecutive natural days.

[0035] By summing up all the passenger volume forecasts, the time-series forecast data of the total passenger volume of the target local railway for the next several consecutive natural days is obtained.

[0036] In one possible design, the time-series passenger volume forecast data is applied, and an optimization algorithm is used to optimize the passenger operation plan of the target local railway for the next several consecutive natural days, resulting in a passenger operation plan with the shortest cosine distance, including:

[0037] Using the time-series prediction data of the entire passenger volume, the passenger operation plan of the target local railway for the next several consecutive natural days is optimized based on the particle swarm optimization algorithm or the gray wolf optimization algorithm to obtain the passenger operation plan with the shortest cosine distance.

[0038] In one possible design, the time-series passenger volume forecast data is applied, and the passenger operation plan for the target local railway in the next several consecutive natural days is optimized based on the Grey Wolf optimization algorithm to obtain the passenger operation plan with the shortest cosine distance, including the following steps S31 to S38:

[0039] S31. Initialize the population: Set the number of gray wolves to N, and the number of iterations to t. max Next, and initialize the multi-parameter search range in the scheme to be optimized, and then execute step S32, where N represents a positive integer greater than or equal to 3, t max Representing a positive integer, the scheme to be optimized refers to the passenger operation scheme of the target local railway for the next several consecutive natural days. The passenger operation scheme includes the planned number of trains departing from the passenger originating station to the passenger terminal station within a unit time period in each future day, the planned passenger carriage type for each train, and the planned number of carriages for each carriage type. The multi-parameter search range includes the search range of planned trains and the search range of the planned number of carriages for each carriage type.

[0040] S32. Initialize the gray wolves: Randomly select three gray wolves from the N gray wolves as the initial α wolf, β wolf and δ wolf, and initialize the individual position vectors of each gray wolf in the N gray wolves within the multi-parameter search range. Then execute step S33, wherein the individual position vectors include the planned departure time search value and the planned number of carriages of various carriage types for each departure time search value.

[0041] S33. For each gray wolf, firstly, the corresponding current individual position vector is used as a passenger transport operation plan, and the corresponding time series data of the full-journey ticket supply is calculated based on the passenger transport operation plan. Then, the cosine distance between the time series prediction data of the full-journey passenger volume and the time series data of the full-journey ticket supply is calculated. Finally, the cosine distance is used as the corresponding individual fitness value, and then step S34 is executed. The time series data of the full-journey ticket supply contains the supply of the full-journey ticket for each unit time period in each future day. The supply is calculated according to the following formula:

[0042]

[0043] In the formula, x represents a positive integer, and P x Y represents the supply of the full-journey ticket within a unit time period in the xth future day. x This represents the total number of planned departures within a unit time period during the x-th future day, from the passenger originating station to the passenger terminating station, where y represents less than or equal to Y. x positive integers, Z x,yThis represents the total number of planned passenger carriage types for the y-th scheduled departure within a unit time period in the x-th future day, where z represents less than or equal to Z. x,y positive integers, This represents the number of planned passenger carriages of the z-th type for the y-th planned departure during a unit time period within the x-th future day. This indicates the number of full-journey tickets that can be provided for each type of planned passenger carriage (z-th type).

[0044] S34. Select the gray wolf with the smallest individual fitness value as the new α wolf, the gray wolf with the second smallest individual fitness value as the new β wolf, and the gray wolf with the third smallest individual fitness value as the new δ wolf, and determine whether the current iteration number has reached t. max If yes, proceed to step S38; otherwise, proceed to step S35.

[0045] S35. Calculate the convergence factors respectively. Collaborative Vectors and cooperative vectors Then step S36 is executed, wherein the convergence factor The cooperative vector and the cooperative vector The calculation formulas are as follows:

[0046]

[0047] In the formula, t represents the current iteration number, and tanh() represents the hyperbolic tangent function. and Let represent random vectors in the range [0,1].

[0048] S36. For each ω wolf, based on the current individual position vectors of the new α wolf, β wolf, and δ wolf, calculate the corresponding individual position vector in the (t+1)th iteration. Then step S37 is executed, wherein the individual position vector Calculated using the following formula:

[0049]

[0050] In the formula, This represents the current individual position vector of the new α wolf. This represents the current individual position vector of the new β wolf. This represents the current individual position vector of the new δ-wolf. This represents the individual position vector in the t-th iteration. and These represent the randomly calculated cooperative vectors, respectively. and These represent the randomly calculated cooperative vectors, respectively.

[0051] S37. Increment the iteration count by 1, then return to step S33;

[0052] S38. The current individual position vector corresponding to the new α wolf is used as the passenger transport operation scheme with the shortest cosine distance.

[0053] Secondly, a local railway passenger transport operation plan decision-making device is provided, which includes a historical data acquisition unit, a future passenger transport prediction unit, a passenger transport plan optimization unit, and an optimal plan output unit that are sequentially connected by communication.

[0054] The historical data acquisition unit is used to acquire time-series data of the total ticket checking volume of the target local railway. The time-series data of the total ticket checking volume contains multiple intraday unit time periods that are sequentially consecutive in time, as well as the historical ticket checking volume of each intraday unit time period within the multiple intraday unit time periods for the total ticket. The total ticket refers to a ticket from the passenger originating station to the passenger destination station.

[0055] The future passenger transport prediction unit is used to take the time series data of the total ticket check volume as the time series data of the total passenger transport volume, and to predict the time series data of the total passenger transport volume of the target local railway for the next several consecutive natural days based on the time series data of the total passenger transport volume. The time series data of the total passenger transport volume contains the historical passenger transport volume for the total ticket within each unit time period of each day, and the time series prediction data of the total passenger transport volume contains the predicted passenger transport volume for the total ticket within each unit time period of each future day in the next several consecutive natural days.

[0056] The passenger transport plan optimization unit is used to apply the time-series forecast data of the total passenger volume and optimize the passenger transport operation plan of the target local railway for the next several consecutive natural days based on an optimization algorithm to obtain a passenger transport operation plan with the shortest cosine distance. The passenger transport operation plan includes the planned number of departures from the passenger originating station to the passenger destination station within each unit time period of each future day, the planned passenger carriage type for each departure, and the planned number of carriages for each carriage type. The shortest cosine distance refers to the minimum value of the cosine distance between the time-series forecast data of the total passenger volume and the time-series data of the total ticket supply calculated based on the passenger transport plan. The time-series data of the total ticket supply contains the supply of total tickets for each unit time period of each future day, and the supply is calculated according to the following formula:

[0057]

[0058] In the formula, x represents a positive integer, and P x Y represents the supply of the full-journey ticket within a unit time period in the xth future day. x This represents the total number of planned departures within a unit time period during the x-th future day, from the passenger originating station to the passenger terminating station, where y represents less than or equal to Y. x positive integers, Z x,y This represents the total number of planned passenger carriage types for the y-th scheduled departure within a unit time period in the x-th future day, where z represents less than or equal to Z. x,y positive integers, This represents the number of planned passenger carriages of the z-th type for the y-th planned departure during a unit time period within the x-th future day. This indicates the number of full-journey tickets that can be provided for each type of planned passenger carriage (z-th type).

[0059] The optimal solution output unit is used to output the passenger operation plan with the shortest cosine distance as the optimal passenger operation plan for the target local railway for the next several consecutive natural days.

[0060] Thirdly, the present invention provides a computer device comprising a memory, a processor, and a transceiver connected in sequence, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the local railway passenger transport operation plan decision-making method as described in the first aspect or any possible design in the first aspect.

[0061] Fourthly, the present invention provides a computer-readable storage product storing instructions that, when executed on a computer, perform the local railway passenger transport operation plan decision-making method as described in the first aspect or any possible design in the first aspect.

[0062] Fifthly, the present invention provides a computer program product, including a computer program or instructions, which, when executed by a computer, implement the local railway passenger transport operation plan decision-making method as described in the first aspect or any possible design in the first aspect.

[0063] The beneficial effects of the above scheme are:

[0064] (1) This invention creatively provides a new scheme for optimizing future passenger transport operation schemes based on time series data of full-journey ticket checking volume. That is, firstly, the time series data of full-journey ticket checking volume of the target local railway is used as the time series data of full-journey passenger volume. Then, based on the time series data of full-journey passenger volume, the time series prediction data of full-journey passenger volume of the target local railway for several consecutive natural days is obtained. Then, the time series prediction data of full-journey passenger volume is applied to optimize the passenger transport operation scheme of the target local railway for several consecutive natural days based on the optimization algorithm to obtain the passenger transport operation scheme with the shortest cosine distance. Finally, the passenger transport operation scheme is used as the optimal passenger transport operation scheme of the target local railway for several consecutive natural days and output. This can be used to make decisions on the passenger transport operation scheme of local railways and maximize the matching between ticket supply and passenger demand, thereby improving passenger occupancy rate, which is conducive to the profitability of local railway operation and makes local railway operation sustainable. This is convenient for practical application and promotion.

[0065] (2) Based on the time series data of the total ticket checking volume and the time series data of the total candidate cancellation volume of the target local railway, we can also combine them to obtain the time series data of the total passenger volume that better reflects the real passenger demand, which will further help us to obtain the optimal local railway passenger operation plan that maximizes the match between ticket supply and passenger demand.

[0066] (3) It can also be based on the time series data of the full-journey ticket inspection volume of the target local railway and the time series data of the non-full-journey ticket inspection volume of each non-full-journey section to obtain the time series data of the full-journey passenger volume that better reflects the real passenger demand, which is more conducive to obtaining the optimal local railway passenger operation plan that maximizes the match between ticket supply and passenger demand. Attached Figure Description

[0067] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0068] Figure 1 This is a flowchart illustrating the decision-making method for local railway passenger transport operation plans provided in this application embodiment.

[0069] Figure 2 This is a schematic diagram of the structure of the local railway passenger transport operation plan decision-making device provided in the embodiments of this application.

[0070] Figure 3 A schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0071] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in conjunction with the accompanying drawings and descriptions of the embodiments or the prior art. Obviously, the following description of the structure of the accompanying drawings is only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. It should be noted that the description of these embodiments is for the purpose of helping to understand the present invention, but does not constitute a limitation of the present invention.

[0072] It should be understood that although the terms "first" and "second", etc., may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another. For example, the first object may be referred to as the second object, and similarly, the second object may be referred to as the first object, without departing from the scope of the exemplary embodiments of the invention.

[0073] It should be understood that the term "and / or" that may appear in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, B exists alone, or A and B exist simultaneously. Another example is A, B and / or C, which can mean that any one of A, B, and C or any combination thereof exists. The term " / and" that may appear in this document describes another relationship between related objects, indicating that two relationships can exist. For example, A / and B can mean: A exists alone or A and B exist simultaneously. In addition, the character " / " that may appear in this document generally indicates that the related objects before and after it are in an "or" relationship.

[0074] Example:

[0075] like Figure 1 As shown, the local railway passenger transport operation plan decision-making method provided in the first aspect of this embodiment can be executed, but is not limited to, by computer equipment with certain computing resources, such as cloud servers, personal computers (PCs, referring to a type of multi-purpose computer suitable for personal use in terms of size, price, and performance; desktop computers, laptops, mini-laptops, tablets, and ultrabooks are all considered personal computers), smartphones, personal digital assistants (PDAs), or wearable devices. Figure 1 As shown, the decision-making method for local railway passenger transport operation plans may include, but is not limited to, the following steps S1 to S3.

[0076] S1. Obtain the time-series data of the total ticket checking volume of the target local railway, wherein the time-series data of the total ticket checking volume contains multiple intraday unit time periods that are sequentially consecutive in time, and the historical ticket checking volume for each intraday unit time period in the multiple intraday unit time periods, which refers to the ticket from the passenger originating station to the passenger destination station.

[0077] In step S1, the target local railway is the object to be managed for passenger transport. Besides the passenger departure station and the passenger destination station, it may also include several intermediate passenger stations. The daily unit time period can be, but is not limited to, one hour. This means that all historical ticket checking volumes recorded for the full-journey ticket are divided into sub-intervals along the time dimension: starting from midnight each day, each day is divided into 24 time periods of one hour each, and the ticket checking volume within each time period is counted. The ticket checking volume statistics from multiple days are then combined to form the time-series data of the full-journey ticket checking volume. The historical ticket checking volume can be the inbound ticket checking volume obtained at the passenger originating station for the full-journey ticket (this can eliminate some cases where people who bought full-journey tickets boarded at intermediate stations, resulting in a more accurate actual passenger volume), or the outbound ticket checking volume obtained at the passenger terminal station for the full-journey ticket (this can eliminate some cases where people who bought partial tickets at the originating station but then purchased full-journey tickets midway, also resulting in a more accurate actual passenger volume); in addition, since this embodiment aligns passenger travel demand with departure ticket checking time, the obtained outbound ticket checking volume needs to correspond to the departure time of the ticket. For example, if the departure time is 10:10 and the outbound ticket checking time is 11:00, then the ticket checking record needs to be assigned to the daily unit time period of 10:00 to 11:00), or it can be a comprehensive statistical result of the aforementioned inbound ticket checking volume and the aforementioned outbound ticket checking volume (for example, taking the average of the two to obtain a more accurate actual passenger volume). In addition, the historical ticket count can be obtained routinely through existing ticket checking equipment.

[0078] S2. The time series data of the total ticket check volume is used as the time series data of the total passenger volume, and the time series data of the total passenger volume of the target local railway is predicted based on the time series data of the total passenger volume for several consecutive natural days in the future. The time series data of the total passenger volume contains the historical passenger volume for the total ticket within each unit time period in each day, and the time series prediction data of the total passenger volume contains the predicted passenger volume for the total ticket within each unit time period in each consecutive natural day in the future.

[0079] In step S2, since the ticket checking volume is the actual passenger volume, the historical ticket checking volumes in the time series data of the entire journey can be directly used as the historical passenger volume. The future consecutive natural days can be specifically set in advance based on the shortest adjustment cycle of the local railway passenger transport operation plan and / or the ticket pre-sale period. For example, if the shortest adjustment cycle is 10 days and the ticket pre-sale period is 15 natural days, then the future consecutive natural days can be from the 16th to the 25th day of the future. The unit time period within the future day needs to be the same length as the unit time period within the day, for example, also 1 hour. Therefore, it is necessary to predict the passenger volume forecast values ​​for each hour from the 16th to the 25th day of the future for the entire journey ticket. That is, the time series forecast data of the entire journey passenger volume is a 10×24 passenger volume forecast matrix. The aforementioned process of predicting the time-series passenger volume forecast data for the target local railway over several consecutive natural days based on the time-series passenger volume data can be, but is not limited to, using a prediction model pre-trained based on a Long Short-Term Memory (LSTM) network (a type of recurrent neural network). To improve the accuracy of the prediction results, preferably, the process of predicting the time-series passenger volume forecast data for the target local railway over several consecutive natural days based on the time-series passenger volume data includes, but is not limited to, the following steps S21 to S29.

[0080] S21. Divide the total passenger volume time series data into multiple sets of total passenger volume time series data that correspond one-to-one with multiple time windows that are sequentially consecutive in time series.

[0081] In step S21, the time window may be, but is not limited to, a series of consecutive natural days, a ten-day period, half a month, or a month.

[0082] S22. For the last time window among the plurality of time windows, at least one set of full-journey passenger volume time-series data corresponding to at least one preceding time window is used as the corresponding training dataset, and one set of full-journey passenger volume time-series data corresponding to the corresponding window is used as the corresponding validation dataset.

[0083] In step S22, the last time window may be, but is not limited to, the current ten-day period or the current month; the at least one preceding time window may be, but is not limited to, the first ten-day period and the period before that, or the at least one preceding time window may be, but is not limited to, the previous month and the month before that, etc.

[0084] S23. Based on the Prophet model, establish a real-time dynamic prediction model for the entire passenger volume using the following formula:

[0085]

[0086] In the formula, t represents time, A(t) represents the total passenger volume at time t, g(t) represents the trend term at time t, s(t) represents the seasonal term at time t, h(t) represents the holiday term at time t, ε(t) represents the error noise term at time t, Q represents a positive integer, q represents a positive integer less than or equal to Q, and l q (t) represents the linear regression term at time t that corresponds to the qth linear influence factor among the Q linear influence factors.

[0087] In step S23, the Prophet model is an open-source time series forecasting tool developed by the Artificial Intelligence Research Institute. Based on the additive model, it can handle data with time and seasonal effects and is applicable to time series forecasting problems in various fields, including but not limited to financial market forecasting, sales data analysis, and disease incidence monitoring. It provides an easy-to-use and highly flexible framework, enabling even non-professional data analysts to effectively perform time series forecasting. Therefore, the real-time dynamic forecasting model for the entire passenger volume can be built based on the Prophet model with conventional modifications.

[0088] S24. Set the growth mode of the trend term to linear growth mode, add natural weeks and natural days to make the seasonal term change periodically, and also set the error noise term to follow a mathematical expectation of μ and a variance of σ. 2 It follows a normal distribution.

[0089] In step S24, the specific configuration details of the trend term, the seasonal term, and the error noise term can be derived conventionally by referring to the corresponding term configuration methods of the existing Prophet model, and will not be repeated here.

[0090] S25. Add an influence weight adjustment factor to the holiday item to obtain a first model parameter vector composed of the weight adjustment parameters of all holiday intervals in the holiday item.

[0091] In step S25, specifically, an influence weight adjustment factor is added to the holiday item to obtain a first model parameter vector composed of the weight adjustment parameters of all holiday intervals in the holiday item, including but not limited to the following steps S251 to S252.

[0092] S251. Assume that there are C holiday intervals in the holiday item: H1, H2, ..., H k ,…,H CLet the holiday term h(t) = λ × Z(t) be denoted as h(t), where the holiday interval contains one natural day or at least two consecutive natural days in time sequence, C represents a positive integer, k represents a positive integer less than or equal to C, and H k Let λ represent the k-th holiday interval, where λ = (λ1, λ2, ..., λk). k ,…,λ C ), d k This represents the total number of days in the k-th holiday interval, where j represents less than or equal to d. k positive integers, λ k,j This represents the weight adjustment factor for the j-th day within the k-th holiday interval, and has... a k β represents the first weight adjustment parameter for the k-th holiday interval. k This represents the second weight adjustment parameter for the k-th holiday interval. This indicates that if t belongs to the k-th holiday interval H k If the value is 1, then the value is 1; otherwise, the value is zero.

[0093] S252. Summarize the first weight adjustment parameter and the second weight adjustment parameter of the C holiday intervals to obtain a first model parameter vector composed of the weight adjustment parameters of all holiday intervals in the holiday item.

[0094] S26. Obtain time series data of all influencing factors related to the total passenger volume, and then perform dynamic screening of linear influencing factors of the total passenger volume based on the time series data, so as to determine the influencing factors related to the total passenger volume that meet the screening conditions as the linear influencing factors, and obtain a second model parameter vector composed of the weight adjustment parameters in the linear regression terms of all the linear influencing factors.

[0095] In step S26, all factors related to total passenger volume include, but are not limited to, the resident population of the location of the passenger originating station and / or the passenger terminal station, the local GDP per capita (Gross Domestic Product), and local meteorological data, as well as any other possible relevant factors. Their time-series data can be routinely obtained. Specifically, based on the time-series data, all factors related to total passenger volume are dynamically screened for linear influencing factors of total passenger volume, so that factors meeting the screening criteria are identified as the linear influencing factors. This includes, but is not limited to, the following steps S261 to S264.

[0096] S261. Randomly select the historical total passenger volume of multiple natural days to form the first sample S1, and further select the data of the multiple natural days from the corresponding time series data as the second sample S2 for a certain total passenger volume related influencing factor among all the total passenger volume related influencing factors.

[0097] In step S261, the plurality of natural days can be, for example, greater than 60 days.

[0098] S262. Perform a normal distribution KS test on the first sample S1 to calculate a first test statistic p1, and also perform the normal distribution KS test on the second sample S2 to calculate a second test statistic p2.

[0099] S263. If both the first verification statistic p1 and the second verification statistic p2 are greater than the first preset threshold, then the correlation coefficient r between the first sample S1 and the sample S2 is calculated according to the following formula. q :

[0100]

[0101] In the formula, i represents a positive integer, and x i y represents the i-th sample value in the first sample S1. i This represents the i-th sample value in the second sample S2. This represents the sample mean of the first sample S1. This represents the sample mean of the second sample S2.

[0102] In step S263, the first preset threshold can be, for example, 0.05.

[0103] S264. If the correlation coefficient r q If the value is greater than the second preset threshold, then the relevant influencing factor of a certain total passenger volume is determined to meet the screening conditions, and the relevant influencing factor of a certain total passenger volume is determined as the linear influencing factor.

[0104] In step S264, the second preset threshold can be, for example, 0.3.

[0105] In step S26, for the certain total passenger volume related influencing factor that has been identified as the linear influencing factor, the corresponding linear regression term l can be specifically recorded. q (t)=r q ×θ q ×M q (t), where θ q M represents the corresponding weight adjustment parameter. q(t) represents the corresponding numerical value, thus obtaining the second model parameter vector composed of the weight adjustment parameters in the linear regression terms of all the linear influence factors.

[0106] S27. Using the training dataset and validation dataset corresponding to the last time window, the real-time dynamic prediction model for the entire passenger volume is trained and its parameters are optimized based on the particle swarm optimization algorithm to obtain the optimized search results of the model parameters that have passed the validation. The optimized search results of the model parameters include the optimized search results of the first model parameter vector, the second model parameter vector, the expected value, and the variance.

[0107] In step S27, the Particle Swarm Optimization (PSO) algorithm is a heuristic optimization algorithm that simulates the information sharing and cooperation mechanism in the foraging behavior of bird flocks, finding the optimal solution to the problem through interactions between individuals. To balance the ability to achieve local and global optima in parameter optimization and to improve the efficiency of parameter search while ensuring overall reliability, preferably, the training dataset and validation dataset corresponding to the final time window are used to train and optimize the parameters of the real-time dynamic prediction model for the entire passenger volume based on the PSO algorithm, obtaining the validated model parameter optimization search results, including but not limited to the following steps S2701 to S2712.

[0108] S2701. Initialize the particle swarm optimization algorithm with parameters including population size N, maximum number of iterations T', self-learning factor C1, and global learning factor C2, and randomly generate the initial positions of each particle in the particle population. Then, execute step S2702, wherein the initial positions of each particle are randomly generated based on the following formula:

[0109]

[0110] In the formula, i' represents a positive integer, d represents a positive integer less than or equal to D, D represents the number of dimensions of the model parameter vector to be optimized, and the model parameter vector to be optimized includes the first model parameter vector to be optimized, the second model parameter vector, the expected value, and the variance. u represents the initial position of the i'-th particle in the particle population along the d-th dimension of the parameter vector of the model to be optimized. c,d l represents the upper limit of the model parameter search space in the d-th dimension. c,d This represents the lower bound of the model parameter search space in the d-th dimension, and rand(0,1) represents a pure decimal random number generation function.

[0111] S2702. For each particle, the corresponding initial position is used as the search result for model parameters and input into the real-time dynamic prediction model of total passenger volume. Then, the training dataset corresponding to the last time window is used to train the real-time dynamic prediction model of total passenger volume, and the corresponding fitness F is calculated. Then, step S2703 is executed, wherein the fitness F is calculated according to the following formula:

[0112]

[0113] In the formula, n represents the total number of samples in the training dataset, and i represents a positive integer. This represents the predicted output term calculated after importing the model input term of the i-th sample data in the training dataset into the real-time dynamic prediction model for total passenger volume. (i) This represents the actual output item of the i-th sample data.

[0114] In step S2702, the specific process of model training is existing technology.

[0115] S2703. The initial position of a particle with the minimum fitness is taken as the initial global optimal position G. ini Let P be the local optimal position reached by the i'-th particle during the iteration process that corresponds to the minimum fitness. i’ Let G be the globally optimal position reached by the particle population during the iteration process that corresponds to the minimum fitness, and let V be the moving speed of the i'th particle. i’ Finally, the local optimal position P is initialized. i’ =G=G ini And initialize the movement speed to V. i’ The value is zero, and the particle movement inertia coefficient is also set to be in the range of [T] during the iteration process. min ,T max The process initializes the period T of the periodic decrease of the inertia coefficient to a randomly selected value within the range of the particle movement inertia coefficient, and also initializes the particle movement inertia coefficient during the iteration process to a value greater than zero, and also initializes the current iteration number t' to 0, and then executes step S2704.

[0116] S2704. Update the particle inertia coefficient of each particle, and then execute step S2705, wherein the update formula for the particle inertia coefficient is as follows:

[0117]

[0118] In the formula, Let represent the particle inertia coefficient obtained after the (t'+1)th iteration, and the particle's motion in the d-th dimension. Let represent the particle inertia coefficient of the i'th particle in the d-th dimension before the (t'+1)-th iteration, and t'%T represent the remainder of the previous iteration number t' with respect to the period T. If the value is less than 0, then a new value is randomly selected within the range of the particle's inertia coefficient as the new period T, and the calculation is recalculated. until Not less than 0.

[0119] S2705. Update the movement speed of each particle, and then execute step S2706, wherein the update formula for the movement speed is as follows:

[0120]

[0121] In the formula, This represents the velocity of the i'th particle in the d-th dimension, obtained after the (t'+1)-th iteration. Let represent the velocity of the i'th particle in the d-th dimension before the (t'+1)-th iteration, and γ1 represent the first random number uniformly distributed in (0,1). This represents the local optimal position reached by the i'-th particle in the d-th dimension during the first t' iterations. Let represent the position of the i'th particle in the d-th dimension before the (t'+1)-th iteration, and let γ2 represent a second random number uniformly distributed within (0,1). This represents the globally optimal position reached by the particle population in the d-th dimension during the first t' iterations.

[0122] S2706. Update the position of each particle, and then execute step S2707, wherein the position update formula is as follows:

[0123] In the formula, This represents the position of the i'th particle in the d-th dimension obtained after the t'+1th iteration.

[0124] S2707. For each particle, the corresponding position is used as the search result of the model parameter and input into the real-time dynamic prediction model of the total passenger volume. Then, the training dataset corresponding to the last time window is applied again to train the real-time dynamic prediction model of the total passenger volume, and the corresponding fitness F is calculated again. Then, step S2708 is executed.

[0125] S2708. Determine the relationship with the local optimal position P. i’ If the corresponding fitness is greater than the current fitness of the i'th particle, then the local optimal position P is set. i’ Update the position to the current position of the i'th particle, and then execute step S2709; otherwise, execute step S2709 directly.

[0126] S2709. Determine whether the fitness corresponding to the global optimal position G is greater than the minimum value among the current fitness of each particle. If so, update the global optimal position G to the current position of any particle with the minimum value, and then execute step S2710. Otherwise, directly execute step S2710.

[0127] S2710. Increment the current iteration number t' by 1, and determine whether the current iteration number t' has reached the maximum iteration number T'. If yes, proceed to step S2711; otherwise, return to step S2704.

[0128] S2711. Input the global optimal position G as the model parameter search result into the real-time dynamic prediction model of the total passenger volume, then apply the real-time dynamic prediction model of the total passenger volume and the validation dataset corresponding to the last time window to predict the total passenger volume, and calculate the new fitness F. new Then proceed to step S2712.

[0129] S2712. Determine F new <ρ×F min If the condition is met, the globally optimal position G is used as the verified model parameter optimization search result; otherwise, after adjusting the search space of the model parameter vector to be optimized, the process returns to step S2701, where ρ represents a preset error verification coefficient greater than 1, and F... min The fitness corresponding to the global optimal position G is represented by the model parameter optimization search result, which includes the first model parameter vector, the second model parameter vector, the expected value, and the variance optimization search result.

[0130] S28. Input the optimized search results of the model parameters into the real-time dynamic prediction model of the total passenger volume, and then apply the real-time dynamic prediction model of the total passenger volume to predict the passenger volume prediction value for the total ticket for each unit time period in each future day in multiple consecutive natural days.

[0131] In step S28, the specific prediction application process of the real-time dynamic prediction model for the entire passenger volume can be derived conventionally based on existing prediction model application methods, and will not be elaborated here.

[0132] S29. Summarize all the passenger volume forecast values ​​to obtain the time-series forecast data of the total passenger volume of the target local railway for the next several consecutive natural days.

[0133] Based on the above steps S21 to S29, the time-series prediction data of the entire passenger volume can be made to have the characteristics of high real-time prediction, high prediction accuracy and stronger model generalization ability, which further facilitates practical application and promotion.

[0134] S3. Applying the time-series forecast data of the total passenger volume, the passenger operation plan of the target local railway for the next several consecutive natural days is optimized based on an optimization algorithm to obtain a passenger operation plan with the shortest cosine distance. The passenger operation plan includes, but is not limited to, the planned departure times from the originating station to the terminating station within each unit time period of the next future day, the planned passenger carriage types for each departure, and the planned number of carriages for each carriage type. The shortest cosine distance refers to the minimum cosine distance between the time-series forecast data of the total passenger volume and the time-series data of the total ticket supply calculated based on the passenger operation plan. The time-series data of the total ticket supply contains the supply of total tickets for each unit time period of the next future day, and the supply is calculated according to the following formula:

[0135]

[0136] In the formula, x represents a positive integer, and P x Y represents the supply of the full-journey ticket within a unit time period in the xth future day. x This represents the total number of planned departures within a unit time period during the x-th future day, from the passenger originating station to the passenger terminating station, where y represents less than or equal to Y. x positive integers, Z x,y This represents the total number of planned passenger carriage types for the y-th scheduled departure within a unit time period in the x-th future day, where z represents less than or equal to Z. x,y positive integers, This represents the number of planned passenger carriages of the z-th type for the y-th planned departure during a unit time period within the x-th future day. This indicates the number of full-journey tickets that can be provided for each type of planned passenger carriage (z).

[0137] In step S3, the planned departure schedule may include, but is not limited to, information such as a unique train number and departure time. Therefore, the total number of planned departures from the passenger originating station to the passenger terminal station within each unit time period of each future day can be calculated based on the departure time. The planned passenger carriage type may include, but is not limited to, hard seat carriages, first-class carriages, second-class carriages, hard sleeper carriages, and / or soft sleeper carriages. Different carriage types provide different numbers of seats (i.e., different full-journey ticket supply). Therefore, the supply of full-journey tickets within each unit time period of each future day can be calculated based on the above formula. Specifically, using the time-series passenger volume forecast data, an optimization algorithm is used to optimize the passenger operation plan of the target local railway for the next several consecutive natural days, resulting in a passenger operation plan with the shortest cosine distance. This includes, but is not limited to, using the time-series passenger volume forecast data and a particle swarm optimization algorithm or a grey wolf optimization algorithm to optimize the passenger operation plan of the target local railway for the next several consecutive natural days, resulting in a passenger operation plan with the shortest cosine distance. The grey wolf optimization algorithm (GWO) is an optimization algorithm inspired by the hunting patterns of grey wolves in their natural environment (it mainly imitates the hunting communication mechanism of wolf packs and the social status among wolves, which are reflected as hunting and hierarchy, respectively), and has the following algorithm principle:

[0138] Suppose a wolf pack has four different rank wolves, from top to bottom: α wolf, β wolf, δ wolf, and ω wolf. The higher-ranking wolf gives instructions to the lower-ranking wolves to first surround the prey. This part of the algorithm is represented as follows:

[0139]

[0140] In the formula, The distance between the individual gray wolf and its prey is represented by t, where t represents the current iteration number. Represents the position vector of the prey. This represents the position vector of an individual gray wolf. and Let each represent a cooperative vector, calculated as follows:

[0141]

[0142] In the formula, This represents the convergence factor, which decreases linearly from 2 to 0 during the iteration process. and Let represent random vectors in the range [0,1]. Next, a hunting process is conducted, where α, β, and δ wolves lead ω wolf in the hunt. That is, α, β, and δ wolves remain stationary while ω wolf iterates. The algorithm is as follows:

[0143]

[0144] In the formula, and Let these represent the position vectors of wolf α, wolf β, and wolf δ in this iteration. and These represent the randomly calculated cooperative vectors, respectively. and These represent the randomly calculated cooperative vectors, respectively. and These represent the distances between other individuals in the group and the α, β, and δ wolves, respectively. This represents the individual position vector in the t-th iteration. This represents the individual position vector in the (t+1)th iteration. Therefore, the Grey Wolf optimization algorithm can be applied to this embodiment to iteratively optimize the passenger transport operation plan of the target local railway for the next several consecutive natural days. Specifically, using the time-series prediction data of the total passenger volume, the Grey Wolf optimization algorithm is used to optimize the passenger transport operation plan of the target local railway for the next several consecutive natural days, obtaining the passenger transport operation plan with the shortest cosine distance, including but not limited to the following steps S31 to S38.

[0145] S31. Initialize the population: Set the number of gray wolves to N, and the number of iterations to t. max Next, and initialize the multi-parameter search range in the scheme to be optimized, and then execute step S32, where N represents a positive integer greater than or equal to 3, t max Representing a positive integer, the scheme to be optimized refers to the passenger operation scheme of the target local railway for the next several consecutive natural days. The passenger operation scheme includes, but is not limited to, the planned departure times from the passenger originating station to the passenger terminating station within each unit time period of the next future day, the planned passenger carriage types for each departure, and the planned number of carriages for each carriage type. The multi-parameter search range includes the search range of planned departure times and the search range of the planned number of carriages for each carriage type.

[0146] In step S31, the search range for planned departure times needs to ensure that the departure times of the planned departure times fall within the corresponding future intraday time period; and considering factors such as departure costs, the search range for the planned number of carriages needs to be an interval value, for example, [8, 16]. Furthermore, for example, N can be 10, t max For example, 200.

[0147] S32. Initialize the gray wolves: Randomly select three gray wolves from the N gray wolves as the initial α wolf, β wolf and δ wolf, and initialize the individual position vectors of each gray wolf in the N gray wolves within the multi-parameter search range. Then execute step S33, wherein the individual position vectors include, but are not limited to, the search values ​​of planned departure times and the search values ​​of the planned number of carriages of various carriage types for each departure.

[0148] In step S32, the planned departure time search value includes at least the departure time search value of the planned departure time.

[0149] S33. For each gray wolf, firstly, the corresponding current individual position vector is used as a passenger transport operation plan, and the corresponding time series data of the full-journey ticket supply is calculated based on the passenger transport operation plan. Then, the cosine distance between the time series prediction data of the full-journey passenger volume and the time series data of the full-journey ticket supply is calculated. Finally, the cosine distance is used as the corresponding individual fitness value, and then step S34 is executed. The time series data of the full-journey ticket supply contains the supply of the full-journey ticket for each unit time period in each future day. The supply is calculated according to the following formula:

[0150]

[0151] In the formula, x represents a positive integer, and P x Y represents the supply of the full-journey ticket within a unit time period in the xth future day. x This represents the total number of planned departures within a unit time period during the x-th future day, from the passenger originating station to the passenger terminating station, where y represents less than or equal to Y. x positive integers, Z x,y This represents the total number of planned passenger carriage types for the y-th scheduled departure within a unit time period in the x-th future day, where z represents less than or equal to Z. x,y positive integers, This represents the number of planned passenger carriages of the z-th type for the y-th planned departure during a unit time period within the x-th future day. This indicates the number of full-journey tickets that can be provided for each type of planned passenger carriage (z).

[0152] In step S33, cosine distance is a method for measuring the similarity between two vectors. It is calculated based on the complement of cosine similarity (its value ranges from 0 to 2, the closer to 0, the more similar the two vectors are, and the closer to 2, the less similar they are). Therefore, the time series prediction data of the entire passenger volume and the time series data of the entire ticket supply can be respectively used as two vectors to calculate the cosine distance and use it as a matching index to measure the matching between ticket supply and passenger demand (the smaller the index value, the better the matching). In addition, the cosine distance can be used as the individual fitness value.

[0153] S34. Select the gray wolf with the smallest individual fitness value as the new α wolf, the gray wolf with the second smallest individual fitness value as the new β wolf, and the gray wolf with the third smallest individual fitness value as the new δ wolf, and determine whether the current iteration number has reached t. max If yes, proceed to step S38; otherwise, proceed to step S35.

[0154] In step S34, the second smallest individual fitness value refers to an individual fitness value that is only smaller than the smallest individual fitness value, and the next smallest individual fitness value refers to an individual fitness value that is only smaller than both the smallest and the second smallest individual fitness values. For example, if the fitness values ​​of the first four individuals are 0.02, 0.06, 0.12, and 0.21 respectively, then the gray wolf with a fitness value of 0.02 can be designated as the new α wolf, the gray wolf with a fitness value of 0.06 as the new β wolf, and the gray wolf with a fitness value of 0.12 as the new δ wolf.

[0155] S35. Calculate the convergence factors respectively. Collaborative Vectors and cooperative vectors Then step S36 is executed, wherein the convergence factor The cooperative vector and the cooperative vector The calculation formulas are as follows:

[0156]

[0157] In the formula, t represents the current iteration number, and tanh() represents the hyperbolic tangent function. and Let represent random vectors in the range [0, 1].

[0158] In step S35, considering the purely linear defect of the convergence factor in the traditional Grey Wolf algorithm, in order to make the convergence factor nonlinear and facilitate the algorithm to achieve global optimization, this embodiment is influenced by the tanh activation function image in the neural network (i.e., selecting the image of the function in the range of [-3,3] and performing transformation operations such as scaling, symmetry and translation successively), and substituting the number of iterations into the function, thus improving the above-mentioned convergence factor expression.

[0159] S36. For each ω wolf, based on the current individual position vectors of the new α wolf, β wolf, and δ wolf, calculate the corresponding individual position vector in the (t+1)th iteration. Then step S37 is executed, wherein the individual position vector Calculated using the following formula:

[0160]

[0161] In the formula, This represents the current individual position vector of the new α wolf. This represents the current individual position vector of the new β wolf. This represents the current individual position vector of the new δ-wolf. This represents the individual position vector in the t-th iteration. and These represent the randomly calculated cooperative vectors, respectively. and These represent the randomly calculated cooperative vectors, respectively.

[0162] In step S36, this embodiment also performs weighted assignment on the gray wolf position update strategy, that is, calculates weight coefficients w1, w2 and w3 respectively, so as to form the improvement points of the gray wolf optimization algorithm together with the new convergence factor expression. Through the performance under 10 commonly used international standard test functions, it is found that it has good accuracy and convergence speed.

[0163] S37. Increment the iteration count by 1, then return to step S33.

[0164] S38. The current individual position vector corresponding to the new α wolf is used as the passenger transport operation scheme with the shortest cosine distance.

[0165] In step S38, according to the principle of the gray wolf algorithm, the current individual position vector corresponding to the new α wolf is the search value of the planned departure times from the passenger origin station to the passenger destination station in each future day unit time period, as well as the search value of the planned number of carriages of various carriage types for each departure, which can be used as the optimal passenger operation scheme.

[0166] S4. The passenger operation plan with the shortest cosine distance is taken as the optimal passenger operation plan for the target local railway in the next several consecutive natural days and output.

[0167] Therefore, based on the local railway passenger transport operation plan decision-making method described in steps S1 to S4 above, a new scheme is provided for optimizing future passenger transport operation plans based on the time series data of the entire journey ticket checking volume. First, the time series data of the entire journey ticket checking volume of the target local railway is used as the time series data of the entire journey passenger volume. Then, based on the time series data of the entire journey passenger volume, the time series prediction data of the entire journey passenger volume of the target local railway for the next several consecutive natural days is obtained. Next, the time series prediction data of the entire journey passenger volume is applied, and an optimization algorithm is used to optimize the passenger transport operation plan of the target local railway for the next several consecutive natural days, obtaining the passenger transport operation plan with the shortest cosine distance. Finally, this passenger transport operation plan is taken as the optimal passenger transport operation plan of the target local railway for the next several consecutive natural days and output. This can be used to decide on local railway passenger transport operation plans, maximizing the match between ticket supply and passenger demand, thereby improving passenger occupancy rates, benefiting the profitability of local railway operations, and making local railway operations sustainable. This approach is convenient for practical application and promotion.

[0168] Based on the technical solution of the first aspect mentioned above, this embodiment also provides a possible design for how to accurately obtain the actual passenger demand when considering candidate ticket booking situations, that is, to use the time series data of the total ticket checking volume as the time series data of the total passenger volume, including but not limited to the following steps S2011 to S2013.

[0169] S2011. Obtain the time-series data of the total number of candidate cancellations for the target local railway, wherein the time-series data of the total number of candidate cancellations includes, but is not limited to, the historical number of candidate cancellations for the total tickets that depart within the unit time period of each day and were automatically cancelled due to non-payment within a preset time period before departure.

[0170] S2012. For each intraday unit time period, add the corresponding historical ticket checking volume in the full-journey ticket checking volume time series data to the corresponding historical candidate cancellation volume in the full-journey candidate cancellation volume time series data multiplied by the preset coefficient and rounded down to obtain the corresponding historical passenger volume for the full-journey ticket.

[0171] S2013. Summarize all the historical passenger volumes to obtain the time-series data of the entire passenger volume.

[0172] Therefore, based on the aforementioned possible design one, we can also combine the time-series data of the total ticket checking volume and the time-series data of the total candidate cancellation volume of the target local railway to obtain the time-series data of the total passenger volume that better reflects the real passenger demand. This will further facilitate the subsequent acquisition of the optimal local railway passenger operation plan that maximizes the match between ticket supply and passenger demand.

[0173] Based on the technical solution of the first aspect mentioned above, this embodiment also provides a possible design two for accurately obtaining the actual passenger demand when considering boarding and alighting at intermediate stations. That is, if there are intermediate passenger stations between the passenger departure station and the passenger destination station, the time series data of the total ticket checking volume is used as the time series data of the total passenger volume, including but not limited to the following steps S2021 to S2025.

[0174] S2021. Obtain time-series data of non-full-journey ticket checking volume for each non-full-journey section in the target local railway, wherein each non-full-journey section includes a section from the passenger originating station to each passenger intermediate station, a section from each passenger intermediate station to the passenger terminal station, and a one-way section from each passenger intermediate station to another passenger intermediate station. The time-series data of non-full-journey ticket checking volume contains historical ticket checking volume for non-full-journey tickets for trains departing from each intraday unit time period, wherein a non-full-journey ticket refers to a ticket for the corresponding non-full-journey section.

[0175] S2022. For each of the non-full-journey segments, the historical ticket checking volume in the corresponding non-full-journey ticket checking volume time series data will be used as the historical non-full-journey passenger volume and broken down into historical short-distance passenger volumes corresponding to the corresponding intraday unit time period and between each pair of adjacent stations.

[0176] S2023. For each intraday unit time period and each pair of adjacent stations, accumulate all corresponding historical short-distance passenger volumes to obtain the corresponding historical short-distance passenger volume.

[0177] S2024. For each intraday unit time period, add the corresponding historical ticket checking volume in the time series data of the total ticket checking volume to the average, median or maximum value of the historical short-distance passenger volume of each pair of adjacent stations to obtain the corresponding historical passenger volume for the total ticket.

[0178] S2025. Summarize all the historical passenger volumes to obtain the time-series data of the entire passenger volume.

[0179] Therefore, based on the aforementioned possible design two, we can also combine the time-series data of the total ticket checking volume of the target local railway with the time-series data of the non-total ticket checking volume of each non-total segment to obtain the time-series data of the total passenger volume that better reflects the real passenger demand. This will further facilitate the subsequent acquisition of the optimal local railway passenger operation plan that maximizes the match between ticket supply and passenger demand.

[0180] like Figure 2 As shown, the second aspect of this embodiment provides a virtual device for implementing the local railway passenger transport operation plan decision-making method described in the first aspect, possible design one or possible design two, including a historical data acquisition unit, a future passenger transport prediction unit, a passenger transport plan optimization unit and an optimal plan output unit that are sequentially connected by communication.

[0181] The historical data acquisition unit is used to acquire time-series data of the total ticket checking volume of the target local railway. The time-series data of the total ticket checking volume contains multiple intraday unit time periods that are sequentially consecutive in time, as well as the historical ticket checking volume of each intraday unit time period within the multiple intraday unit time periods for the total ticket. The total ticket refers to a ticket from the passenger originating station to the passenger destination station.

[0182] The future passenger transport prediction unit is used to take the time series data of the total ticket check volume as the time series data of the total passenger transport volume, and to predict the time series data of the total passenger transport volume of the target local railway for the next several consecutive natural days based on the time series data of the total passenger transport volume. The time series data of the total passenger transport volume contains the historical passenger transport volume for the total ticket within each unit time period of each day, and the time series prediction data of the total passenger transport volume contains the predicted passenger transport volume for the total ticket within each unit time period of each future day in the next several consecutive natural days.

[0183] The passenger transport plan optimization unit is used to apply the time-series forecast data of the total passenger volume and optimize the passenger transport operation plan of the target local railway for the next several consecutive natural days based on an optimization algorithm to obtain a passenger transport operation plan with the shortest cosine distance. The passenger transport operation plan includes the planned number of departures from the passenger originating station to the passenger destination station within each unit time period of each future day, the planned passenger carriage type for each departure, and the planned number of carriages for each carriage type. The shortest cosine distance refers to the minimum value of the cosine distance between the time-series forecast data of the total passenger volume and the time-series data of the total ticket supply calculated based on the passenger transport plan. The time-series data of the total ticket supply contains the supply of total tickets for each unit time period of each future day, and the supply is calculated according to the following formula:

[0184]

[0185] In the formula, x represents a positive integer, and P x Y represents the supply of the full-journey ticket within a unit time period in the xth future day. x This represents the total number of planned departures within a unit time period during the x-th future day, from the passenger originating station to the passenger terminating station, where y represents less than or equal to Y.x positive integers, Z x,y This represents the total number of planned passenger carriage types for the y-th scheduled departure within a unit time period in the x-th future day, where z represents less than or equal to Z. x,y positive integers, This represents the number of planned passenger carriages of the z-th type for the y-th planned departure during a unit time period within the x-th future day. This indicates the number of full-journey tickets that can be provided for each type of planned passenger carriage (z-th type).

[0186] The optimal solution output unit is used to output the passenger operation plan with the shortest cosine distance as the optimal passenger operation plan for the target local railway for the next several consecutive natural days.

[0187] The working process, working details and technical effects of the aforementioned device provided in the second aspect of this embodiment can be found in the decision-making method for local railway passenger transport operation schemes described in the first aspect, possible design one or possible design two, and will not be repeated here.

[0188] like Figure 3 As shown, the third aspect of this embodiment provides a computer device for executing the local railway passenger transport operation plan decision-making method as described in the first aspect, possible design one, or possible design two. The device includes a memory, a processor, and a transceiver connected in sequence. The memory stores a computer program, the transceiver sends and receives messages, and the processor reads the computer program and executes the local railway passenger transport operation plan decision-making method as described in the first aspect, possible design one, or possible design two. Specifically, the memory may include, but is not limited to, random-access memory (RAM), read-only memory (ROM), flash memory, first-in-first-out (FIFO) memory, and / or first-in-last-out (FILO) memory, etc.; the processor may include, but is not limited to, an STM32F105 series microprocessor. Furthermore, the computer device may also include, but is not limited to, a power module, a display screen, and other necessary components.

[0189] The working process, working details and technical effects of the aforementioned computer equipment provided in the third aspect of this embodiment can be found in the decision-making method for local railway passenger transport operation schemes described in the first aspect, possible design one or possible design two, and will not be repeated here.

[0190] This fourth aspect of the embodiment provides a computer-readable storage product that stores instructions comprising a local railway passenger transport operation plan decision-making method as described in the first aspect, possible design one, or possible design two. Specifically, the computer-readable storage product stores instructions that, when executed on a computer, perform the local railway passenger transport operation plan decision-making method as described in the first aspect, possible design one, or possible design two. The computer-readable storage product refers to a data storage medium, which may include, but is not limited to, computer-readable storage media such as floppy disks, optical disks, hard disks, flash memory, USB flash drives, and / or Memory Sticks. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0191] The working process, working details and technical effects of the aforementioned computer-readable storage product provided in the fourth aspect of this embodiment can be found in the local railway passenger transport operation plan decision-making method as described in the first aspect, possible design one or possible design two, and will not be repeated here.

[0192] This fifth aspect of the embodiment provides a computer program product, including a computer program or instructions, which, when executed by a computer, implement the local railway passenger transport operation plan decision-making method as described in the first aspect, possible design one, or possible design two. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.

[0193] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A decision-making method for local railway passenger transport operation plans, characterized in that, include: Obtain time-series data of the total ticket checking volume of the target local railway. The time-series data of the total ticket checking volume contains multiple intraday unit time periods that are sequentially consecutive in time, as well as the historical ticket checking volume of each intraday unit time period within the multiple intraday unit time periods for the total ticket. The total ticket refers to a ticket from the passenger originating station to the passenger destination station. The time-series data of the total ticket check volume is used as the time-series data of the total passenger volume. Based on the time-series data of the total passenger volume, the time-series prediction data of the total passenger volume of the target local railway for the next several consecutive natural days is obtained. The time-series data of the total passenger volume contains the historical passenger volume for the total ticket within a unit time period in each day. The time-series prediction data of the total passenger volume contains the predicted passenger volume for the total ticket within a unit time period in each of the next several consecutive natural days. Using the time-series forecast data of the total passenger volume, an optimization algorithm is used to optimize the passenger operation plan of the target local railway for several consecutive future natural days, resulting in a passenger operation plan with the shortest cosine distance. The passenger operation plan includes the planned number of departures from the originating station to the terminating station within each unit time period of each future day, the planned passenger carriage type for each departure, and the planned number of carriages for each carriage type. The shortest cosine distance refers to the minimum cosine distance between the time-series forecast data of the total passenger volume and the time-series data of the total ticket supply calculated based on the passenger operation plan. The time-series data of the total ticket supply contains the supply of total tickets for each unit time period of each future day, and the supply is calculated according to the following formula: In the formula, x represents a positive integer, and P x Y represents the supply of the full-journey ticket within a unit time period in the xth future day. x This represents the total number of planned departures within a unit time period during the x-th future day, from the passenger originating station to the passenger terminating station, where y represents less than or equal to Y. x positive integers, Z x,y This represents the total number of planned passenger carriage types for the y-th scheduled departure within a unit time period in the x-th future day, where z represents less than or equal to Z. x,y positive integers, This represents the number of planned passenger carriages of the z-th type for the y-th planned departure during a unit time period within the x-th future day. This indicates the number of full-journey tickets that can be provided for each type of planned passenger carriage (z-th type). The passenger transport operation plan with the shortest cosine distance is output as the optimal passenger transport operation plan for the target local railway for the next several consecutive natural days.

2. The decision-making method for local railway passenger transport operation plans according to claim 1, characterized in that, The time-series data of the total ticket checking volume is used as the time-series data of the total passenger volume, including: Obtain the time-series data of the full-journey candidate cancellation volume of the target local railway, wherein the time-series data of the full-journey candidate cancellation volume contains the historical candidate cancellation volume for the full-journey tickets that depart within the unit time period of each day and were automatically cancelled due to non-payment within a preset time period before departure; For each intraday unit time period, the corresponding historical ticket checking volume in the full-journey ticket checking volume time series data is added one-to-one to the corresponding historical candidate cancellation volume in the full-journey candidate cancellation volume time series data multiplied by a preset coefficient and rounded down to obtain the corresponding historical passenger volume for the full-journey ticket. By summing up all the historical passenger volumes, the time-series data of the total passenger volume is obtained.

3. The decision-making method for local railway passenger transport operation plans according to claim 1, characterized in that, If there are intermediate passenger stations between the passenger departure station and the passenger destination station, then the time series data of the total ticket checking volume will be used as the time series data of the total passenger volume, including: Obtain time-series data of non-full-journey ticket checking volume for each non-full-journey section of the target local railway. Each non-full-journey section includes a section from the passenger originating station to each passenger intermediate station, a section from each passenger intermediate station to the passenger terminal station, and a one-way section from each passenger intermediate station to another passenger intermediate station. The time-series data of non-full-journey ticket checking volume contains historical ticket checking volume for non-full-journey tickets for trains departing from each intraday unit time period. The non-full-journey ticket refers to the ticket for the corresponding non-full-journey section. For each of the non-full-journey segments, the historical ticket checking volume in the corresponding non-full-journey ticket checking volume time series data will be used as the historical non-full-journey passenger volume and broken down into historical short-distance passenger volumes corresponding to the corresponding intraday unit time period and between each pair of adjacent stations. For each intraday unit time period and each pair of adjacent stations, the corresponding historical short-distance passenger volume is accumulated to obtain the corresponding historical short-distance passenger volume. For each intraday unit time period, the corresponding historical ticket checking volume in the time series data of the total ticket checking volume is added one by one to the average, median or maximum value of the historical short-distance passenger volume of each pair of adjacent stations to obtain the corresponding historical passenger volume for the total ticket. By summing up all the historical passenger volumes, the time-series data of the total passenger volume is obtained.

4. The decision-making method for local railway passenger transport operation plans according to claim 1, characterized in that, Based on the aforementioned time-series data of total passenger volume, the time-series prediction data of the total passenger volume of the target local railway for several consecutive natural days in the future is obtained, including: The total passenger volume time series data is divided into multiple sets of total passenger volume time series data that correspond one-to-one with multiple time windows that are sequentially consecutive in time series. For the last time window among the multiple time windows, at least one set of full-journey passenger volume time-series data corresponding to at least one preceding time window is used as the corresponding training dataset, and one set of full-journey passenger volume time-series data corresponding to the corresponding window is used as the corresponding validation dataset. A real-time dynamic prediction model for total passenger volume is established based on the Prophet model, expressed by the following formula: In the formula, t represents time, A(t) represents the total passenger volume at time t, g(t) represents the trend term at time t, s(t) represents the seasonal term at time t, h(t) represents the holiday term at time t, ε(t) represents the error noise term at time t, Q represents a positive integer, q represents a positive integer less than or equal to Q, and l q (t) represents the linear regression term at time t that corresponds to the qth linear influence factor among the Q linear influence factors; The growth pattern of the trend term is set to linear, and natural weeks and natural days are added to make the seasonal term change periodically. The error noise term is also set to follow a mathematical expectation of μ and a variance of σ. 2 The normal distribution; Add an influence weight adjustment factor to the holiday item to obtain a first model parameter vector consisting of the weight adjustment parameters of all holiday intervals in the holiday item; Time series data of all factors related to the total passenger volume are obtained, and then the linear influencing factors of the total passenger volume are dynamically screened based on the time series data so as to determine the factors related to the total passenger volume that meet the screening conditions as the linear influencing factors, and a second model parameter vector composed of the weight adjustment parameters in the linear regression terms of all the linear influencing factors is obtained. Using the training dataset and validation dataset corresponding to the last time window, the real-time dynamic prediction model for the entire passenger volume is trained and its parameters are optimized based on the particle swarm optimization algorithm to obtain the optimized search results of the model parameters that have passed the validation. The optimized search results of the model parameters include the optimized search results of the first model parameter vector, the second model parameter vector, the expected value, and the variance. The optimized search results of the model parameters are input into the real-time dynamic prediction model of the total passenger volume, and then the real-time dynamic prediction model of the total passenger volume is applied to predict the passenger volume prediction value for the total ticket for each unit time period in each future day in multiple consecutive natural days. By summing up all the passenger volume forecasts, the time-series forecast data of the total passenger volume of the target local railway for the next several consecutive natural days is obtained.

5. The decision-making method for local railway passenger transport operation plans according to claim 1, characterized in that, Using the aforementioned time-series passenger volume forecast data, an optimization algorithm is used to optimize the passenger operation plan for the target local railway over the next several consecutive natural days, resulting in a passenger operation plan with the shortest cosine distance, including: Using the time-series prediction data of the entire passenger volume, the passenger operation plan of the target local railway for the next several consecutive natural days is optimized based on the particle swarm optimization algorithm or the gray wolf optimization algorithm to obtain the passenger operation plan with the shortest cosine distance.

6. The decision-making method for local railway passenger transport operation schemes according to claim 5, characterized in that, Using the aforementioned full-journey passenger volume time-series forecast data, the passenger operation plan for the target local railway in the next several consecutive natural days is optimized based on the Grey Wolf optimization algorithm to obtain a passenger operation plan with the shortest cosine distance, including the following steps S31 to S38: S31. Initialize the population: Set the number of gray wolves to N, and the number of iterations to t. max Next, and initialize the multi-parameter search range in the scheme to be optimized, and then execute step S32, where N represents a positive integer greater than or equal to 3, t max Representing a positive integer, the scheme to be optimized refers to the passenger operation scheme of the target local railway for the next several consecutive natural days. The passenger operation scheme includes the planned number of trains departing from the passenger originating station to the passenger terminal station within a unit time period in each future day, the planned passenger carriage type for each train, and the planned number of carriages for each carriage type. The multi-parameter search range includes the search range of planned trains and the search range of the planned number of carriages for each carriage type. S32. Initialize the gray wolves: Randomly select three gray wolves from the N gray wolves as the initial α wolf, β wolf and δ wolf, and initialize the individual position vectors of each gray wolf in the N gray wolves within the multi-parameter search range. Then execute step S33, wherein the individual position vectors include the planned departure time search value and the planned number of carriages of various carriage types for each departure time search value. S33. For each gray wolf, firstly, the corresponding current individual position vector is used as a passenger transport operation plan, and the corresponding time series data of the full-journey ticket supply is calculated based on the passenger transport operation plan. Then, the cosine distance between the time series prediction data of the full-journey passenger volume and the time series data of the full-journey ticket supply is calculated. Finally, the cosine distance is used as the corresponding individual fitness value, and then step S34 is executed. The time series data of the full-journey ticket supply contains the supply of the full-journey ticket for each unit time period in each future day. The supply is calculated according to the following formula: In the formula, x represents a positive integer, and P x Y represents the supply of the full-journey ticket within a unit time period in the xth future day. x This represents the total number of planned departures within a unit time period during the x-th future day, from the passenger originating station to the passenger terminating station, where y represents less than or equal to Y. x positive integers, Z x,y This represents the total number of planned passenger carriage types for the y-th scheduled departure within a unit time period in the x-th future day, where z represents less than or equal to Z. x,y positive integers, This represents the number of planned passenger carriages of the z-th type for the y-th planned departure during a unit time period within the x-th future day. This indicates the number of full-journey tickets that can be provided for each type of planned passenger carriage (z-th type). S34. Select the gray wolf with the smallest individual fitness value as the new α wolf, the gray wolf with the second smallest individual fitness value as the new β wolf, and the gray wolf with the third smallest individual fitness value as the new δ wolf, and determine whether the current iteration number has reached t. max If yes, proceed to step S38; otherwise, proceed to step S35. S35. Calculate the convergence factors respectively. Collaborative Vectors and cooperative vectors Then step S36 is executed, wherein the convergence factor The cooperative vector and the cooperative vector The calculation formulas are as follows: In the formula, t represents the current iteration number, and tanh() represents the hyperbolic tangent function. and Let represent random vectors in the range [0,1]. S36. For each ω wolf, based on the current individual position vectors of the new α wolf, β wolf, and δ wolf, calculate the corresponding individual position vector in the (t+1)th iteration. Then step S37 is executed, wherein the individual position vector Calculated using the following formula: In the formula, This represents the current individual position vector of the new α wolf. This represents the current individual position vector of the new β wolf. This represents the current individual position vector of the new δ-wolf. This represents the individual position vector in the t-th iteration. and These represent the randomly calculated cooperative vectors, respectively. and These represent the randomly calculated cooperative vectors, respectively. S37. Increment the iteration count by 1, then return to step S33; S38. The current individual position vector corresponding to the new α wolf is taken as the passenger transport operation scheme with the shortest cosine distance.

7. A decision-making device for local railway passenger transport operation plans, characterized in that, It includes a historical data acquisition unit with sequential communication connections, a future passenger transport prediction unit, a passenger transport plan optimization unit, and an optimal plan output unit; The historical data acquisition unit is used to acquire time-series data of the total ticket checking volume of the target local railway. The time-series data of the total ticket checking volume contains multiple intraday unit time periods that are sequentially consecutive in time, as well as the historical ticket checking volume of each intraday unit time period within the multiple intraday unit time periods for the total ticket. The total ticket refers to a ticket from the passenger originating station to the passenger destination station. The future passenger transport prediction unit is used to take the time series data of the total ticket check volume as the time series data of the total passenger transport volume, and to predict the time series data of the total passenger transport volume of the target local railway for the next several consecutive natural days based on the time series data of the total passenger transport volume. The time series data of the total passenger transport volume contains the historical passenger transport volume for the total ticket within each unit time period of each day, and the time series prediction data of the total passenger transport volume contains the predicted passenger transport volume for the total ticket within each unit time period of each future day in the next several consecutive natural days. The passenger transport plan optimization unit is used to apply the time-series forecast data of the total passenger volume and optimize the passenger transport operation plan of the target local railway for the next several consecutive natural days based on an optimization algorithm to obtain a passenger transport operation plan with the shortest cosine distance. The passenger transport operation plan includes the planned number of departures from the passenger originating station to the passenger destination station within each unit time period of each future day, the planned passenger carriage type for each departure, and the planned number of carriages for each carriage type. The shortest cosine distance refers to the minimum value of the cosine distance between the time-series forecast data of the total passenger volume and the time-series data of the total ticket supply calculated based on the passenger transport plan. The time-series data of the total ticket supply contains the supply of total tickets for each unit time period of each future day, and the supply is calculated according to the following formula: In the formula, x represents a positive integer, and P x Y represents the supply of the full-journey ticket within a unit time period in the xth future day. x This represents the total number of planned departures within a unit time period during the x-th future day, from the passenger originating station to the passenger terminating station, where y represents less than or equal to Y. x positive integers, Z x,y This represents the total number of planned passenger carriage types for the y-th scheduled departure within a unit time period in the x-th future day, where z represents less than or equal to Z. x,y positive integers, This represents the number of planned passenger carriages of the z-th type for the y-th planned departure during a unit time period within the x-th future day. This indicates the number of full-journey tickets that can be provided for each type of planned passenger carriage (z-th type). The optimal solution output unit is used to output the passenger operation plan with the shortest cosine distance as the optimal passenger operation plan for the target local railway for the next several consecutive natural days.

8. A computer device, characterized in that, The system includes a memory, a processor, and a transceiver connected in sequence, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the local railway passenger transport operation plan decision-making method as described in any one of claims 1 to 6.

9. A computer-readable storage product, characterized in that... The computer-readable storage product stores instructions that, when executed on a computer, perform the local railway passenger transport operation plan decision-making method as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or the instructions are executed by the computer, they implement the local railway passenger transport operation plan decision-making method as described in any one of claims 1 to 6.

Images