Fixed-route generation method in demand-responsive transit (DRT) considering non-vulnerable area constraints and computer-readable recording medium storing program for executing fixed-route generation method

Abstract

The present invention relates to a fixed-route generation method in demand-responsive transit (DRT) considering non-vulnerable area constraints and to a computer-readable recording medium storing a program for executing the fixed-route generation method, wherein during operation of demand-responsive transit, the method resolves a swamp phenomenon in which a vehicle cannot easily leave a non-vulnerable area due to a large number of calls generated within the non-vulnerable area, and ensures mobility rights in transportation-vulnerable areas in demand-responsive transit via improved operational efficiency.

Description

A method for generating fixed routes for a demand-responsive transport (DRT) system considering limiting conditions in non-vulnerable areas, and a medium storing a computer-readable program for executing the fixed route generation method.

[0001] The present invention relates to a method for generating a fixed route for a demand-responsive transport (DRT) system considering restricted conditions in non-vulnerable areas, and a medium storing a computer-readable program for executing the fixed route generation method. More specifically, the invention relates to a method for generating a fixed route for a demand-responsive transport (DRT) system considering restricted conditions in non-vulnerable areas, and a medium storing a computer-readable program for executing the fixed route generation method, which can resolve the swamp phenomenon—a phenomenon in which vehicles cannot easily leave non-vulnerable areas due to a large number of calls occurring within non-vulnerable areas during the operation of a demand-responsive transport (DRT) system—and guarantee mobility rights in transportation-vulnerable areas within the demand-responsive transport system through operational efficiency, and secure the public interest and equity of the demand-responsive transport system.

[0002] Conventional public transportation services operate only on fixed times and routes, regardless of user demand.

[0003] In this general public transportation operation method, users cannot decide the boarding time or boarding and alighting stops, but rather they are determined by the route operation methods of local governments or transportation companies.

[0004] In addition, service satisfaction is not high due to the disadvantage of high transportation costs and the inability to use public transportation at desired times, as the system operates on designated routes repeatedly regardless of user demand.

[0005] Demand Responsive Transport (DRT) is a new public transportation service being introduced primarily in areas with poor transportation to address this issue. This DRT is a demand-responsive transportation service that allows users to call a multi-passenger vehicle to their desired time and place using smart devices or apps, and it refers to a means of transportation that enables users to board and reach their desired destination quickly compared to current public transportation services.

[0006] However, due to the aforementioned purpose, Demand-Responsive Transit (DRT) faces problems during actual operation. Specifically, the high volume of requests originating from transportation-insecure areas (the city center or surrounding areas) prevents citizens in transportation-insecure areas (outskirts) who actually need the service from using it.

[0007] Specifically, non-vulnerable areas are characterized by the availability of sufficient alternative public transportation options, short distances, and no mobility issues even without using Demand-Responsive Transit (DRT); a key feature of these regions is their high population density. In other words, if calls from non-vulnerable areas are allowed, high demand would result in limited short-distance transportation within and around the city center. Consequently, intermittent calls from vulnerable areas would be given lower priority due to operational efficiency concerns regarding vehicle routes, potentially leading to long waiting times for users in these areas or even being rejected by the system.

[0008] As such, due to the large number of calls occurring within non-vulnerable areas, a phenomenon occurs where vehicles cannot easily leave non-vulnerable areas. Here, this phenomenon is referred to as the "swamp phenomenon," likening it to a vehicle falling into a swamp.

[0009] The swamp phenomenon is a situation where demand-responsive public transport (DRT) vehicles are unable to leave high-density demand areas due to a large number of calls, making it impossible to accommodate calls from low-density demand areas that are transportation-vulnerable; in other words, it results in low priority for visits and high waiting times.

[0010] In other words, Demand-Responsive Transit (DRT) operates without fixed routes or timetables, creating efficient routes in real-time solely in response to user requests. This leads to a "swamp" phenomenon where vehicles become isolated in high-demand areas, causing users requesting services from low-demand outskirts to experience long waiting times. Furthermore, repeated requests from downtown areas result in lower dispatch priority for users in outskirts. Consequently, users in outskirts experience low service satisfaction due to vehicles arriving later than scheduled, which leads to user cancellations and the system refusing to dispatch vehicles.

[0011] The inventor of the present invention confirmed the phenomenon of swamp occurrence using actual demand-responsive public transport (DRT) operation data and explains this with reference to Fig. 1. Fig. 1 is a diagram illustrating an example of the phenomenon of swamp occurrence in demand-responsive public transport (DRT).

[0012] To verify the phenomenon of swamp occurring in demand-responsive public transportation (DRT), 14 months of demand-responsive public transportation (DRT) operation performance data from Area A (City A) were used.

[0013] We identified the top 50 ODs with high call success rates and the top 50 ODs with high call failure rates. Here, OD refers to a pair of a traveler's origin (boarding stop) and destination (drop-off stop) (=origin-destination pair).

[0014] A call failure occurs when a call is canceled by the system itself or by the user due to a long waiting time assigned to the user after dispatch; therefore, the success or failure of a call is distinguished based on the starting point (call point).

[0015] Here, the swamp phenomenon is described as a phenomenon in which a vehicle gets stuck in a frequently called area (swamp) and prioritizes handling calls from the swamp while rejecting calls coming from the outer areas (hereinafter referred to as blind spots).

[0016] Area with high call success OD = Vehicle's main operating area = Swamp

[0017] An area that is not the vehicle's main operating zone and where call failure points (call points) are concentrated = blind spot

[0018] As a result of the analysis, swamps and blind spots were repeatedly identified in similar areas. In fact, the swamp area was the main commercial district of City A with a high volume of pedestrian traffic, whereas the blind spot area was an outlying area located far from the city center with low demand for calls. Through the case of Area A, the actual swamp phenomenon and the existence of blind spots resulting from it were confirmed.

[0019] The simplest method to prevent this swamp phenomenon is to designate downtown areas as demand control zones and intentionally block traffic within the city center. However, in such cases, revenue decreases due to reduced demand in the downtown area, and the public interest and equity—the fundamental raison d'être of public transportation—may be undermined by restricting the mobility rights of downtown residents.

[0020] To resolve this issue, one method involves deploying an additional fixed-route vehicle to utilize both the free-route Demand-Responsive Transit (DRT) service and the fixed-route bus; however, if both the existing vehicles and the additional vehicle are used for DRT, there is a risk that both vehicles may fall into the swamp phenomenon.

[0021] Therefore, there is a need for research and system development on ways to make it easier for citizens in vulnerable areas to use demand-responsive public transportation (DRT) and to minimize service restrictions for citizens in non-vulnerable areas.

[0022] Accordingly, the present invention, which aims to solve the aforementioned conventional problems, provides a method for generating a fixed route for a demand-responsive transport (DRT) considering the limitations of non-vulnerable areas and a medium storing a computer-readable program for executing the fixed route generation method, which resolves the swamp phenomenon—a phenomenon in which vehicles cannot easily leave non-vulnerable areas due to a large number of calls occurring within non-vulnerable areas during the operation of a demand-responsive transport system (DRT)—and guarantees the right to mobility in transportation-vulnerable areas within the demand-responsive transport system through operational efficiency, thereby securing the public interest and equity of the demand-responsive transport system.

[0023] In addition, the present invention has another objective of providing a method for generating fixed routes and a fixed route generation system for a demand-responsive transportation system (DRT) that considers the limited conditions of non-vulnerable areas, which can guarantee mobility rights in transportation-vulnerable areas by creating routes that allow fixed routes to share the demand that significantly increases the impact of high-demand-density areas, i.e., swamp phenomena, when existing fixed bus routes have much higher transport capacity and greater efficiency than a demand-responsive transportation system (DRT) as users' movement patterns are closer in high-demand-density areas.

[0024] The problems solved by the present invention are not limited to those mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the description below.

[0025] According to one aspect of the present invention for achieving the objectives and other features of the present invention, a method for generating fixed routes for a demand-responsive transportation (DRT) system comprises: a step of constructing inter-stop travel demand data for constructing inter-stop travel demand data of users in an analysis area; a spatial cluster partitioning step of partitioning the analysis area into one or more spatial clusters; a population list initialization step of generating p.POPULATION_SIZE number of routes using a Genetic Algorithm (GA) for the analysis area of ​​the partitioned spatial clusters (wherein p.POPULATION_SIZE is the entire set of solutions constituting one generation of the genetic algorithm); and a population evolution step of evolving a population containing routes by applying a genetic modification operation process and a population maintenance process based on a population list (population) containing p.POPULATION_SIZE routes, and passing it on to the next generation. A method for generating fixed routes for a demand-responsive transportation system (DRT) is provided, characterized by including: a final route acquisition step for obtaining multiple final routes by determining whether succession proceeds or stops for a population obtained through the population evolution step.

[0026] In one aspect of the present invention, the population list initialization step may be performed to generate a route defined by the following formula.

[0027]

[0028] : Weight

[0029] : Initial standard deviation of the number of regional visits

[0030] : Standard deviation of the number of regional visits after excluding passengers assigned to fixed route r

[0031] : Number of passengers on fixed route r

[0032] (Here, B1 is the weight for variance, B2 is the weight for passenger capacity).

[0033] In one aspect of the present invention, the population list initialization step may include: a basic route generation step for generating individual basic routes; a travel time calculation step for calculating travel time for the generated basic routes; a travel frequency calculation step for calculating the number of trips for the basic routes after the travel time calculation step; a route final demand calculation step for allocating demand to the basic routes after the travel frequency calculation step; and a route suitability calculation step for calculating the suitability of the routes after the route final demand calculation step.

[0034] In one aspect of the present invention, the basic route generation step specifies the starting node and the last node of the route, and generates the intermediate nodes of the route by selecting them from a list of all stop nodes using a non-replacement sampling method, with the number of intermediate nodes of the route randomly ranging from (2 to p.S_max) (where p.S_max is the maximum number of intermediate stop nodes); the travel time calculation step calculates the travel time of the route, wherein if the travel time value cannot be obtained, the travel distance value is multiplied by a time conversion coefficient value to convert it into travel time and calculate the result; the operation frequency calculation step calculates the operation frequency (operation frequency count) by dividing the calculated travel time by the operation time; and the route final demand calculation step calculates the final demand of the route through the following equation It is structured to produce,

[0035]

[0036] (Here, is allocated demand, and is a demand allocation ratio function based on frequency)

[0037] The above-mentioned route suitability calculation step is performed using the following formula, and

[0038]

[0039] : Weight

[0040] : Initial standard deviation of the number of regional visits

[0041] : Standard deviation of the number of regional visits after excluding passengers assigned to fixed route r

[0042] : Number of passengers on fixed route r

[0043] (Here, B1 is the weight for variance, B2 is the weight for passenger capacity)

[0044] The above population evolution stage is repeated as many times as a given p.GENERATION (number of genetic algorithm generations) until the succession cessation condition is met, and in each iteration, the process of genetic modification operation and population maintenance is performed based on the inherited population of the previous generation, and the above final route acquisition stage is performed by determining and deciding that the current generation is the optimal route if at least one of the following is satisfied: the iteration has progressed through generation succession equal to p.GENERATION, or the iteration has progressed for more than 1 / 3 of p.GENERATION and there is no change in the route with the highest optimal fitness between the current generation and the previous generation, and the above population list initialization stage is performed by determining the route fitness ( If the value is greater than or equal to a preset value, the optimal route visit order search step may be further included to search for the optimal route visit order and add route information and the fitness of the route to a population list.

[0045] According to another aspect of the present invention, a computer-readable medium is provided for executing a method for generating a fixed route of a demand-responsive transportation system (DRT) according to the above-mentioned aspect.

[0046] According to the method for generating a fixed route of a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions and a medium storing a computer-readable program for executing the method for generating a fixed route according to the present invention, the following effects are provided.

[0047] First, the present invention has the effect of improving operational efficiency and guaranteeing mobility rights in traffic-vulnerable areas when operating a demand-responsive transportation (DRT), and securing the public interest and equity of the DRT.

[0048] Second, the present invention has the effect of promoting the expansion of demand-responsive transportation systems (DRT) by ensuring the utility of the DRT, by operating a free-route type demand-responsive transportation system (DRT) in outer areas with priority given to calls, and operating a fixed-route type in downtown areas where users' travel patterns are relatively monotonous.

[0049] Third, the present invention minimizes the swamp phenomenon of the demand-responsive transportation system (DRT) and allows for the visit of high-demand stops, thereby resolving the swamp phenomenon and realizing operational efficiency. This enables the placement of different route types suitable for two regions with different travel patterns and regional characteristics, thereby allowing for the expectation of synergistic effects.

[0050] The effects of the present invention are not limited to those mentioned above, and other unmentioned problems will be clearly understood by those skilled in the art from the description below.

[0051] Figure 1 is a diagram illustrating an example of a phenomenon in which swamp occurs in demand-responsive public transportation (DRT).

[0052] FIG. 2 is a flowchart schematically illustrating a method for generating a fixed route of a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention.

[0053] FIG. 3 is a flowchart schematically illustrating the process of the population list initialization step included in the method for creating a fixed route of a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention.

[0054] FIG. 4 is a diagram showing an example of a spatial cluster division step included in a fixed route generation method of a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention.

[0055] Figure 5 is a diagram showing a Toy Network analysis area for simulating the fixed route generation method of the demand-responsive transportation system of the present invention.

[0056] Figure 6 is a diagram showing the route with the highest degree of fit obtained through simulation for the fixed route generation method of the demand-responsive transportation system of the present invention.

[0057] Figure 7 is a diagram showing routes having the top 2 to 5 suitability obtained through simulation for the fixed route generation method of the demand-responsive transportation system of the present invention.

[0058] Further objects, features, and advantages of the present invention can be more clearly understood from the following detailed description and the accompanying drawings.

[0059] Before providing a detailed description of the present invention, it should be understood that the present invention is capable of various modifications and may have various embodiments, and that the examples described below and illustrated in the drawings are not intended to limit the present invention to specific embodiments, but rather include all modifications, equivalents, and substitutions that fall within the spirit and scope of the present invention.

[0060] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.

[0061] The terms used in this specification are used merely to describe specific embodiments and are not intended to limit the invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as "comprising" or "having" are intended to indicate the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0062] Additionally, terms such as "...part," "...unit," and "...module" described in the specification refer to a unit that processes at least one function or operation, and this may be implemented in hardware, software, or a combination of hardware and software.

[0063] Furthermore, in the description referring to the attached drawings, identical components are assigned the same reference numeral regardless of drawing symbols, and redundant descriptions thereof are omitted. In describing the present invention, if it is determined that a detailed description of related prior art could unnecessarily obscure the essence of the present invention, such detailed description is omitted.

[0064] Hereinafter, a method for generating a fixed route for a demand-responsive transportation (DRT) system considering non-vulnerable area restriction conditions according to a preferred embodiment of the present invention and a medium storing a computer-readable program for executing the method for generating a fixed route will be described.

[0065] First, a method for generating fixed routes for a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention will be explained.

[0066] The method for generating a fixed route of a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention can be implemented to generate a route defined by the following formula.

[0067]

[0068] : Weight

[0069] : Initial standard deviation of the number of regional visits

[0070] : Standard deviation of the number of regional visits after excluding passengers assigned to fixed route r

[0071] : Number of passengers on fixed route r

[0072] Here, B1 is the weight for variance, and B2 is the passenger capacity weight, which is a pre-set weight through a fitness-based weight selection method (roulette wheel selection).

[0073] To elaborate on the above formula, when the analysis region exhibiting a strong swamp phenomenon is divided into k spaces and the call occurrences for each space unit are aggregated, it means that the variance in the number of calls occurring between space units is large, whereas conversely, when the call distribution is evenly distributed across regions, it means that the variance is small.

[0074] Therefore, in the formula This indicates how much the standard deviation of calls per spatial unit in the initial analysis area was reduced as the corresponding fixed route was introduced.

[0075] And the route must not only resolve the swamp phenomenon but also demonstrate profitability. Therefore, in the formula It is a route that is better the more passengers it carries, based on the number of passengers boarding the route.

[0076] Hereinafter, a method for generating fixed routes for a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention will be described in detail with reference to the attached drawings.

[0077] FIG. 2 is a flowchart schematically illustrating a method for generating a fixed route of a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention, FIG. 3 is a flowchart schematically illustrating the process of a population list initialization step included in the method for generating a fixed route of a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention, and FIG. 4 is a diagram showing an example of a spatial cluster division step included in the method for generating a fixed route of a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention.

[0078] The definitions of the parameters used in the method for generating fixed routes of a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention are as follows.

[0079] p.POPULATION_SIZE : The entire set of solutions constituting one generation of the genetic algorithm

[0080] p.GENERATION: Number of genetic algorithm generations

[0081] p.B1: Weights for variance

[0082] p.B2 : Passenger demand weights

[0083] p.R_max : Maximum number of routes

[0084] p.F_min : Minimum frequency (number of runs per day)

[0085] p.S_max : Maximum number of intermediate stop nodes

[0086] p.DETOUR_RATIO_max : Maximum bypass rate

[0087] p.OPERATION_HOUR : Available operating hours (Total operating hours with rest time restricted)

[0088] p.ANL_TEMPERATURE : (Simulated Annealing) Initial Temperature

[0089] p.ANL_COOLING_RATE : (Simulated Annealing) Cooling Rate

[0090] The method for generating a fixed route for a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention, as shown in FIGS. 2 and 3, largely comprises a step of constructing travel demand data between stops (S100), a spatial cluster division step (S200), a population list initialization step (S300), a population evolution step (S400), and a final route acquisition step (S500).

[0091] Specifically, the method for generating fixed routes for a demand-responsive transport system (DRT) considering non-vulnerable area restriction conditions according to the present invention comprises, as shown in FIGS. 2 and 3, a step of constructing inter-stop travel demand data (S100) for constructing inter-stop travel demand data of users utilizing a demand-responsive transport system (DRT) or a bus (hereinafter collectively referred to as 'route transport system') in an analysis area; a spatial cluster division step (S200) for dividing the analysis area into one or more, preferably a plurality of, spatial clusters after the inter-stop travel demand data construction step (S100); and a population list initialization step (S300) for generating routes p.POPULATION_SIZE in the analysis area of ​​the spatial clusters divided into a plurality in the spatial cluster division step (S200) using a Genetic Algorithm (GA). The method includes: a population evolution step (S400) in which a population list (population) containing p.POPULATION_SIZE lines generated in the population list initialization step (S300) is used to apply genetic modification operations and a population maintenance process to evolve a population containing better lines and pass them on to the next generation; and a final line acquisition step (S500) in which a determination is made to proceed with or stop the succession of the population obtained through the population evolution step (S400) to obtain multiple final lines.

[0092] The step of constructing the above-mentioned travel demand data between stops (S100) can be performed such that when the total number of stops within the analysis area is n, the data is in the form of an nxn matrix, the rows of the matrix correspond to the boarding stop ID, the columns correspond to the alighting stop ID, and the element values ​​correspond to the traffic volume.

[0093] The above spatial cluster partitioning step (S200) divides the analysis area into multiple groups (k, where k is a natural number greater than or equal to 2). Here, spatial clusters are assigned to the entire area of ​​the analysis area by grouping similar living areas into one based on living areas (distance, accessibility, etc.) and assigning areas that are not similar to a new cluster.

[0094] The above population list initialization step (S300) comprises a basic route generation step (S310) for generating individual basic routes, a travel time calculation step (S320) for calculating travel time for each basic route (individual basic route) generated in the basic route generation step (S310), a travel frequency calculation step (S330) for calculating the number of trips for each basic route after the travel time calculation step (S320), a route final demand calculation step (S340) for allocating demand to the basic routes after the travel frequency calculation step (S330), and the route performance, i.e., route suitability, after the route final demand calculation step (S340). It includes a route suitability calculation step (S350) that calculates the route suitability.

[0095] The above basic route generation step (S310) specifies the starting node and the last node of the route, and the number of intermediate nodes of the route is randomly selected from (2 to p.S_max) and can be selected from the entire list of stop nodes by a non-replacement extraction method.

[0096] The starting node and the last node of the above route may be, for example, the starting depot node and the ending depot node.

[0097] Here, the route can be composed of a depot node + multiple intermediate nodes + a depot node.

[0098] Furthermore, the travel time calculation step (S320) calculates the travel time for each of the basic routes (individual basic routes) generated in the basic route generation step (S310), and calculates the travel time (e.g., in minutes) of the individual routes generated in the basic route generation step (S310).

[0099] If the travel time value is not obtained in the travel time calculation step (S320), the travel time can be obtained by converting the travel distance value into travel time by multiplying it by a time conversion coefficient value.

[0100] And the above-mentioned operation frequency calculation step (S330) consists of calculating the operation frequency for each basic route after the above-mentioned travel time calculation step (S320).

[0101] The above operation frequency calculation step (S330) is configured to calculate the operation frequency by dividing the travel time (min) obtained by calculating in the above travel time calculation step (S320) by the operation time (p.OPERATION_HOUR * 60 min / 1 h). Here, the frequency can be truncated at the first decimal place.

[0102] Furthermore, the above-mentioned route final demand calculation step (S340) comprises allocating demand to a basic route after the above-mentioned operation frequency calculation step (S330) to obtain the allocated demand, and calculating the final demand of the corresponding route from this.

[0103] Specifically, the above route final demand calculation step (S340) is carried out through the following process.

[0104] The above route final demand calculation step (S340) allocates demand corresponding to the following conditions on the route from the demand data constructed in the above inter-stop travel demand data construction step (S100). does.

[0105] Vehicles allow boarding for passengers passing through the boarding and alighting points in sequence. However, even if the vehicle passes the passengers' boarding and alighting stops in sequence, if there is a significant detour involving additional stops between the two, the passengers do not board the route; therefore, this is summarized by the following formula.

[0106] < p.DETOUR _ RATIO_MAX

[0107] Therefore, the final demand for the route is calculated by multiplying the total number of passengers of the route by the frequency of the route calculated in the operation frequency calculation step (S330) as shown in the equation below. Calculates.

[0108]

[0109] It can be customized with a demand allocation ratio function based on frequency, and this is configured to have more demand as the operation frequency increases using a logarithmic function.

[0110] Continuing, the route suitability calculation step (S350) is, after the route final demand calculation step (S340), the performance of the route, i.e., the suitability of the route It is to produce.

[0111] The above route suitability calculation step (S350) can be performed to generate a route defined by the following formula.

[0112]

[0113] : Weight

[0114] : Initial standard deviation of the number of regional visits

[0115] : Standard deviation of the number of regional visits after excluding passengers assigned to fixed route r

[0116] : Number of passengers on fixed route r

[0117] Here, B1 is the weight for variance, and B2 is the passenger capacity weight, which is a pre-set weight through a fitness-based weight selection method (roulette wheel selection).

[0118] As mentioned earlier, to elaborate on the above formula, when dividing the analysis region where the swamp phenomenon is strong into multiple spaces and aggregating the call occurrences for each space unit, it means that the variance in the number of calls occurring between space units is large, whereas conversely, when the call distribution is evenly distributed across regions, it means that the variance is small.

[0119] Therefore, in the formula This indicates how much the standard deviation of calls per spatial unit in the initial analysis area was reduced as the corresponding fixed route was introduced.

[0120] And since the route must consider profitability as well as resolving the swamp phenomenon, in the formula It is a route that is better the more passengers it carries, based on the number of passengers boarding the route.

[0121] Meanwhile, in the present invention, the population list initialization step (S300) is the route suitability in the route suitability calculation step (S350). If the value is greater than or equal to a preset value, for example, 5 or greater, the route optimal visit order search step (S360) may be further included to search for the route optimal visit order and add route information and the suitability of the route to a population list.

[0122] The above-mentioned optimal route visit order search step (S360) is configured to search for and obtain the route stop visit order that has the optimal suitability by changing the visit order of intermediate stops, as there is a high possibility of inefficiency in the visit order because intermediate stops of the route are randomly extracted.

[0123] The above-mentioned optimal route visit order search step (S360) can be performed by searching using a simulated annealing technique.

[0124] Specifically, the initial route is , goodness of fit is Set it to , and the initial temperature T is set by multiplying the number of intermediate nodes by p.ANL_TEMPERATURE.

[0125] And randomly swap the order of the intermediate nodes of the route to create a new path Create and fit Check .

[0126] and If, , otherwise probability Depending on Update to and decrease temperature T by the cooling rate p.ANL_COOLING_RATE.

[0127] Here, the number of iterations is performed multiplied by 10 of the number of intermediate nodes to obtain the final optimal path overfit Returns.

[0128] Next, the aforementioned evolutionary stage (S400) is configured to evolve a population containing better lines by applying a genetic modification operation process and a population maintenance process based on a population list (population) containing p.POPULATION_SIZE lines generated in the population list initialization stage (S300), and to pass this evolutionary stage on to the next generation.

[0129] The above population evolution stage (S400) is repeated a given number of iterations (p.GENERATION) until the succession cessation condition is met, and in each iteration, the process of genetic modification operations and population maintenance is carried out based on the inherited population of the previous generation.

[0130] In the above population evolution stage (S400), the genetic modification operation is performed through the following process by sequentially executing the following genetic operators, namely the selection operator, crossover operator, mutation operator, and mathematical operator.

[0131] - Selection Operator

[0132] The application process of the selection operator selects two routes by making the route with high fitness highly probable (roulette wheel selection), and the selected route becomes the parent route.

[0133] total_futness =

[0134] Here, the fitness of all routes is summed for the population. N is the number of populations (p.POPULATION_SIZE).

[0135]

[0136] Calculate the selection probability of an individual route by dividing the fitness of the individual route by the total fitness.

[0137] - Cross operator

[0138] The two parent lines selected through the execution of the selection operator are shuffled, and an intersection point is randomly assigned. The line created by this shuffling becomes the child line.

[0139] The first child is created by combining the front part of parent 1 and the back part of parent 2, and the second child is created by combining the front part of parent 2 and the back part of parent 1.

[0140] - Variant operator

[0141] Various variations are created by applying mutation operations to each of the two child lines obtained through the execution of the above crossover operator.

[0142] Specifically, first, the probability of mutation occurrence is examined. When there is a 15% probability of mutation occurring and an 85% probability of remaining unchanged, or when there is a 15% probability of mutation being determined, one of the following four mutation types is selected with equal probability.

[0143] Insertion: Adds a random stop to the route.

[0144] Deletion: Removes stops excluding the start and end.

[0145] Swap: Swap the locations of two stops on a single route.

[0146] Transfer: Move one of the stops on the current route to another route.

[0147] Routes that have been determined to be mutated return a new route that meets the total travel time conditions of the route (e.g., 30 minutes or more, 3 hours or less), and if the conditions are not met, attempt the mutation again.

[0148] - Mathematical operator

[0149] In the mathematical operator, the process of identifying and correcting two child routes that violated constraints due to deformation in the above process is performed, and the following correction work is carried out.

[0150] - Check if the route starts and ends at the depot, and modify it.

[0151] - Modify if checking for duplicate nodes at intermediate stops

[0152] - Adjust the total travel time of the route to, for example, between 30 minutes and 3 hours; modify it by adding stops if the time is short and removing stops if it is long.

[0153] And when the process from the population evolution step (S400) to the mathematical operator is completed, for two child routes, the process proceeds from the travel time calculation step (S320) to the route fitness calculation step (S350). If the fitness is a preset value (for example, if the fitness is 5 or higher), the route optimal visit order search step (S360) is performed, and the route in the optimal order and the fitness obtained through the route optimal visit order search step (S360) are added to the new generation's population list.

[0154] In the above population evolution stage (S400), the genetic modification operation process is performed to repeat until the list of populations of the new generation has p.POPULATION_SIZE lines.

[0155] Furthermore, in the above population evolution stage (S400), the population maintenance process is a process of determining whether offspring generated during the genetic modification operation process will be included as the next successor individuals.

[0156] In other words, in the population evolution stage (S400) above, the population maintenance process is configured to combine the population list (parent line) inherited from the previous generation and the new generation population list newly created during the population maintenance process to create a population list with 2 × p.POPULATION_SIZE lines, and then to create a population list with p.POPULATION_SIZE lines to be passed on to the next generation through a diversity control process and an elite preservation process.

[0157] Specifically, the diversity control of the population maintenance process in the population evolution stage (S400) is performed as follows.

[0158] - Diversity control process in the population maintenance process during the population evolution stage (S400)

[0159] Routes within the population list maintain diversity by excluding similar individuals. This is to prevent the problem of getting stuck in local optimization.

[0160] First, the best route with the highest fitness among the objects is selected, and its similarity with the remaining routes is evaluated. The similarity evaluation calculates the Hamming distance (h) and the route length (L).

[0161] Hamming distance refers to how different consecutive pairs of nodes on two lines are.

[0162] The Hamming distance (h) is calculated using the following formula.

[0163]

[0164] δ is a function that returns 1 when two pairs of nodes are different and 0 when they are the same.

[0165] is the additional distance due to the difference in route length.

[0166] And the route length (L) is calculated using the following formula.

[0167]

[0168] Then, the survival probability Ps of the corresponding route is calculated based on the Hamming distance (h) and route length (L). The lower the similarity to the optimal route, the higher the survival probability to the next generation (using c = 0.5 and α = 2.0).

[0169]

[0170] Here, a random number is generated, and if the random number probability is smaller than the survival probability Ps, the route is maintained in the population list, and if it is larger, one route is created through the population list initialization step (S300) and included in the population list.

[0171] In the present invention, the population maintenance process in the population evolution stage (S400) is configured such that even after the process is terminated, the diversity control process maintains a population list of 2 × p.POPULATION_SIZE routes.

[0172] - Elite preservation process during population maintenance in the population evolution stage (S400)

[0173] The elite preservation process can be accomplished by using a fitness-based weighted selection method (roulette wheel selection) to select the top half of individuals and determine the final population list of p.POPULATION_SIZE to send to the next generation.

[0174] Next, the final route acquisition step (S500) is configured to obtain multiple final routes by determining whether to proceed with or stop succession for the population obtained through the population evolution step (S400).

[0175] The above final route acquisition step (S500) may be performed by determining and deciding that the current generation is the optimal route if at least one of the following two conditions is satisfied.

[0176] - When Iteration has progressed generational succession by p.GENERATION

[0177] - If iterations have progressed more than 1 / 3 of p.GENERATION and there are no changes to the top (p.POPULATION_SIZE / 4) routes with the highest optimal fitness between the current and previous generations

[0178] In the present invention, the final route acquisition step (S500) may be performed to produce n route results with the highest suitability.

[0179] In other words, the above final route acquisition step (S500) can be configured to provide n routes with high suitability and non-overlapping suitability along with route information from the results of the finally calculated generations.

[0180] Meanwhile, the inventor of the present invention confirmed through simulation that a fixed route generation method for a demand-responsive transportation system (DRT) considering non-vulnerable area restriction conditions according to the present invention can generate a route with optimal suitability, and this is explained with reference to FIGS. 5 to 7.

[0181] FIG. 5 is a diagram showing a Toy Network analysis area for simulation regarding the fixed route generation method of the demand-responsive transportation system of the present invention, FIG. 6 is a diagram showing the route with the highest suitability obtained through simulation regarding the fixed route generation method of the demand-responsive transportation system of the present invention, and FIG. 7 is a diagram showing the routes with the top 2 to 5 suitability obtained through simulation regarding the fixed route generation method of the demand-responsive transportation system of the present invention.

[0182] Demand was generated in a Toy Network environment to exhibit regional demand imbalance. The total demand is 1,000 cases, and the total number of stops in the analysis area is 64. The route depot nodes were selected randomly.

[0183] The analysis area was divided into five spaces (Area 1 to 5), and Area 4 and Area 0, corresponding to the left side, were assumed to be areas with high demand generation rates of 29% and 28%, respectively, Area 2, which has moderate demand, was assumed to be an area with 19%, while Area 1 and Area 3, the outer right side, were assumed to be vulnerable areas with rates of 14% and 10%, respectively.

[0184] The intensity of the bus stop color indicates the demand density, and a darker color signifies higher demand.

[0185] In this simulation, the goodness-of-fit function was applied as follows.

[0186]

[0187] : Weight

[0188] : Initial standard deviation of the number of regional visits

[0189] : Standard deviation of the number of regional visits after excluding passengers assigned to fixed route r

[0190] : Number of passengers on fixed route r

[0191] The route must be operated to resolve demand imbalances across the entire service area, and it can be selected as a good route only if it also has a high volume of passenger demand.

[0192] The inventor of the present invention was able to determine the direction of the route to allocate demand in demand-overcrowded areas when the route is generated through the above-described simulation, and to visit stops with high demand density within the direction, and confirmed from the results of the optimal Best route and the routes with the 2nd to 5th highest suitability that the routes were generated to visit stops with high demand in the center of demand-overcrowded areas.

[0193] As described above, the method for generating a fixed route for a demand-responsive transport (DRT) system considering non-vulnerable area constraints according to the present invention derives an optimal route by utilizing a genetic algorithm, which is a metaheuristic technique, as the main framework; customizes the genetic algorithm by applying engineering concepts and judgments at each individual step during this process; and generates a route by enhancing the precision of the metaheuristic approach by additionally combining a simulated annealing technique to search for a route with a higher degree of fit.

[0194] Meanwhile, although the above has described a method for generating fixed routes for a demand-responsive transport (DRT) system considering the restrictive conditions of non-vulnerable areas, it goes without saying that a computer-readable recording medium storing a program for implementing the method for generating fixed routes for a demand-responsive transport (DRT) system considering the restrictive conditions of non-vulnerable areas, and a program stored on such a computer-readable recording medium for implementing the method for generating fixed routes for a demand-responsive transport (DRT) system considering the restrictive conditions of non-vulnerable areas, can also be implemented.

[0195] In other words, those skilled in the art will readily understand that the method for generating fixed routes for a demand-responsive transport (DRT) system, considering the aforementioned limitations of non-vulnerable areas, may be provided by being tangibly implemented as a program of instructions for implementing it, and included in a computer-readable recording medium. That is to say, it may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, etc., either individually or in combination. The program instructions recorded on the computer-readable recording medium may be those specifically designed and configured for the present invention, or they may be those known and available to those skilled in computer software. Examples of the above computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, and USB memory. The above computer-readable recording media may also be transmission media such as optical or metallic wires, waveguides, etc., containing a carrier wave that transmits signals specifying program instructions, data structures, etc. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The above hardware devices may be configured to operate as one or more software modules to perform the operation of the present invention, and vice versa.

[0196] According to the method for generating a fixed route for a demand-responsive transportation system (DRT) considering the restricted conditions of non-vulnerable areas according to the present invention as described above, and a medium storing a computer-readable program for executing the method for generating a fixed route, when operating a demand-responsive transportation system (DRT), it is possible to ensure operational efficiency and mobility rights in traffic-vulnerable areas and secure the public interest and equity of the demand-responsive transportation system. Furthermore, by operating a free-route type demand-responsive transportation system (DRT) in outer areas with priority given to calls, and operating a fixed route in urban areas where users' travel patterns are relatively monotonous, there is an advantage of being able to promote the expansion of the demand-responsive transportation system (DRT) by securing the utility of the demand-responsive transportation system (DRT).

[0197] In addition, the present invention has the advantage of allowing for the expectation of synergy effects by arranging different route types suitable for two regions with different travel patterns and regional characteristics, thereby minimizing the swamp phenomenon of the demand-responsive transportation system (DRT) and ensuring that high-demand stops are visited, thereby resolving the swamp phenomenon and realizing operational efficiency.

[0198] The embodiments described in this specification and the accompanying drawings are merely illustrative of a part of the technical concept included in the present invention. Accordingly, since the embodiments disclosed in this specification are intended to explain, not limit, the technical concept of the present invention, it is obvious that the scope of the technical concept of the present invention is not limited by these embodiments. All variations and specific embodiments that can be easily deduced by a person skilled in the art within the scope of the technical concept included in the specification and drawings of the present invention should be interpreted as being included within the scope of the rights of the present invention.

[0199] The modes for carrying out the invention are described together in the best mode for carrying out the invention.

[0200] The present invention can improve operational efficiency and guarantee mobility rights in traffic-vulnerable areas when operating a demand-responsive transportation (DRT) system, and can secure the public interest and equity of the demand-responsive transportation system.

[0201] In addition, the present invention enables the operation of a free-route demand-responsive transportation system (DRT) in outer areas with priority given to calls, and a fixed-route system in downtown areas where users' travel patterns are relatively monotonous, thereby ensuring the utility of the demand-responsive transportation system (DRT) and promoting the expansion of the demand-responsive transportation system (DRT).

Claims

1. As a method for generating fixed routes for a demand-responsive transportation system (DRT), Inter-stop travel demand data construction step for building inter-stop travel demand data for users in the analysis area; A spatial cluster partitioning step for dividing the above analysis region into one or more spatial clusters; A population list initialization step for generating p.POPULATION_SIZE number of routes using a Genetic Algorithm (GA) for the analysis regions of the above-described spatial clusters (wherein p.POPULATION_SIZE is the entire set of solutions constituting one generation of the genetic algorithm); A population evolution stage in which a population containing a line to which a genetic modification operation process and a population maintenance process have been applied is evolved based on a population list (population) containing p.POPULATION_SIZE lines generated above, and is passed down to the next generation; and Characterized by including a final route acquisition step for obtaining multiple final routes by determining whether succession proceeds or is suspended for the population obtained through the above population evolution step. Method for creating fixed routes in a demand-responsive transportation system (DRT).

2. In Paragraph 1, The above population list initialization step is characterized by being configured to generate a route defined by the following formula, which serves as a basis for determining the optimal route by judging the suitability of the route. : Weight : Initial standard deviation of the number of regional visits : Standard deviation of the number of regional visits after excluding passengers assigned to fixed route r : Number of passengers on fixed route r (Here, B1 is the weight for variance, B2 is the weight for passenger capacity) Method for creating fixed routes in a demand-responsive transportation system (DRT).

3. In Paragraph 1, The above population list initialization step Basic route generation step for generating individual basic routes; A travel time calculation step for calculating travel time for the above-mentioned generated basic route; A step for calculating the frequency of operation for a basic route following the above-mentioned travel time calculation step; A route final demand calculation step for allocating demand to a basic route after the above-mentioned operation frequency calculation step; and Characterized by including a route suitability calculation step for calculating the route suitability after the above-mentioned route final demand calculation step. Method for creating fixed routes in a demand-responsive transportation system (DRT).

4. In Paragraph 3, The above basic route generation step specifies the starting node and the last node of the route, and generates the route by selecting a random number of intermediate nodes from the entire list of intermediate nodes without replacement, with the number of intermediate nodes ranging from (2 to p.S_max) (where p.S_max is the maximum number of intermediate node stops). The above travel time calculation step is performed by calculating the travel time of the route, and if the travel time value cannot be obtained, the travel distance value is multiplied by a time conversion factor value to convert it into travel time and calculate it. The above-mentioned operation frequency calculation step is performed to calculate the operation frequency by dividing the above-mentioned travel time by the operation time, and The above-mentioned final demand calculation step is the final demand of the route through the following formula It is structured to produce, (Here, is allocated demand, and is a demand allocation ratio function based on frequency) The above-mentioned route suitability calculation step is performed using the following formula, and : Weight : Initial standard deviation of the number of regional visits : Standard deviation of the number of regional visits after excluding passengers assigned to fixed route r : Number of passengers on fixed route r (Here, B1 is the weight for variance, B2 is the weight for passenger capacity) The above population evolution stage is repeated (iterated) as many times as a given p.GENERATION (number of genetic algorithm generations) until the succession cessation condition is met, and in each iteration, the process of genetic modification operations and population maintenance processes is carried out based on the inherited population of the previous generation. The above final route acquisition step is determined by judging and deciding that the current generation is the optimal route when at least one of the following is satisfied: when the iteration has progressed generational succession by the above p.GENERATION, or when the iteration has progressed by more than 1 / 3 of p.GENERATION and there is no change in the route with the highest optimal fitness between the current generation and the previous generation. The above population list initialization step is the route suitability ( in the above route suitability calculation step (S350) Characterized by further including a route optimal visit order search step that, when the value is greater than or equal to a preset value, searches for the optimal visit order of a route and adds route information and the fitness of the corresponding route to a population list. Method for creating fixed routes in a demand-responsive transportation system (DRT).

5. A computer-readable medium storing a method for generating a fixed route of a demand-responsive transport (DRT) system according to any one of claims 1 through 4.

Images