A vehicle scheduling and travel path optimization method for coordinating dry and branch line bus services

Abstract

The application discloses a kind of vehicle scheduling and travel path optimization methods of coordinating trunk line and branch line public transport service, comprising: obtaining the demand information of passenger in preset time period;Based on the contact between each passenger and the selected boarding point and the selected transfer point, a virtual space-time information is constructed;Objective function is constructed;Randomly select virtual transfer point and virtual boarding point, generate current sequence;The current sequence is input into the first algorithm, and the first function value is obtained;By changing the current sequence, the optimal objective function value is found using the annealing algorithm, and the branch line vehicle scheduling and passenger travel path are planned based on the optimal objective function value.The application improves the passenger-carrying rate and operating efficiency of branch line vehicles, enhances the overall travel experience of passengers, and reduces operating costs and passenger travel costs.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent transportation. In particular, it relates to a vehicle scheduling and travel path optimization method for coordinating trunk and branch line public transport services. BACKGROUND

[0002] Overall optimization of trunk and branch line services is attracting more and more attention. A common mode is "subway / express bus / passenger bus / smart rail" + "demand-responsive feeder bus". Such trunk and branch line integrated travel mode has been widely recognized for its advantages of low travel cost, strong environmental protection and high travel efficiency.

[0003] However, the existing technology lacks globality in dealing with trunk and branch line integration. On the one hand, passengers are often limited by information when traveling and tend to choose only one boarding point and transfer point (collectively referred to as service point) according to individual preferences. When there are many discrete service points in the system, the route of the branch line transportation provided is high in road toll and low in efficiency, resulting in prolonged travel time for passengers, high operating costs for operators and reduced potential for passenger carpooling. On the other hand, the optimization of branch line transportation usually lacks effective coordination with trunk line transportation. The existing technology usually assumes that the main trunk transportation has a robust schedule and adjusts the route and schedule of the branch line transportation accordingly. However, for many trunk line transportation systems other than subways, this assumption is usually not true, and even if it is true, it is difficult to coordinate successfully in a complex network. Another strategy is to coordinate trunk and branch line services simultaneously. However, this method is suitable for specific scenarios and generally does not work well because not all users of the trunk line system will accept branch line services, and considering the coordination problem from the perspective of only some passengers may bring negative benefits to other passengers.

[0004] Therefore, there is an urgent need for a method to solve the problems of poor operation connection between trunk and branch line transportation modes in public transportation systems, individual cognitive dependence and limitations in passenger travel, and low carpooling potential of branch line transportation modes. SUMMARY

[0005] The present application is proposed based on the above-mentioned needs of the prior art. The technical problem to be solved by the present application is to provide a vehicle scheduling and travel path optimization method for coordinating trunk and branch line public transport services to improve the passenger carrying rate and operating efficiency of branch line vehicles and enhance the overall travel experience of passengers.

[0006] To solve the above problems, the present application is implemented by adopting the following technical solutions:

[0007] A vehicle scheduling and travel path optimization method for coordinating trunk and branch bus services is provided, which comprises: obtaining demand information of passengers in a preset time period, the demand information comprising at least one branch line candidate boarding point, at least one trunk and branch line candidate transfer point, a trunk line terminal and a service time window, the service time window comprising an earliest time for the passengers to board and a latest time for the passengers to arrive at the trunk line terminal; constructing virtual space-time information based on the connection between each passenger and the candidate boarding point and the candidate transfer point, the virtual space-time information comprising virtual boarding points, virtual transfer points, virtual depots and virtual road segments; constructing an objective function; randomly selecting virtual transfer points and virtual boarding points to generate a current sequence; inputting the current sequence into a first algorithm to obtain a first function value, the first algorithm comprising: pre-allocating a branch line vehicle for each input sequence; taking the first boarding point in the input sequence as the first station of the branch line vehicle; calculating the departure time of the branch line vehicle based on the time for the passengers to arrive at the first station at the earliest time; sequentially traversing the virtual boarding points in the input sequence, and when a boarding point in the input sequence is traversed, judging whether a first condition can be met if the branch line vehicle arrives at the boarding point; the first condition comprising that the number of passengers in the branch line vehicle does not exceed a preset passenger capacity, the time for the passengers to arrive at the boarding point is greater than the earliest time, and the time for the passengers to arrive at the trunk line terminal is less than the latest time; when the first condition is met, adding the boarding point to the route information of the branch line vehicle, adding the time for arriving at the boarding point to the arrival time information, and continuing to traverse the next boarding point in the input sequence; when the first condition is not met, dispatching another branch line vehicle from the virtual depot, taking the boarding point as the first station of the new branch line vehicle, and continuing to traverse the remaining boarding points in the input sequence based on the new branch line vehicle; after the traversal is completed, obtaining the route information and the arrival time information of each branch line vehicle, and determining the corresponding objective function value; changing the current sequence, finding an optimal objective function value by using a simulated annealing algorithm, and planning the branch line vehicle scheduling and passenger travel path based on the optimal objective function value.

[0008] Optionally, the current sequence is generated by randomly selecting virtual transfer points and virtual boarding points, comprising: S400 randomly selecting one of the virtual transfer points corresponding to each passenger and determining the corresponding candidate transfer point as an actual transfer point; S401 determining potential carpooling passengers based on the same actual transfer point; randomly selecting one of the virtual boarding points corresponding to the potential carpooling passengers to generate a boarding point sequence corresponding to each actual transfer point as the current sequence.

[0009] Optionally, the optimal objective function value is found by changing the current sequence using an annealing algorithm, comprising: S600 changing the current sequence based on a first rule, inputting the changed current sequence into the first algorithm to obtain a second function value; determining whether the second function value is better than the current first optimal solution, if yes, the second function value is output as the current first optimal solution; otherwise, calculating a first probability based on the current temperature, the current first optimal solution and the second function value, and determining whether the first probability is greater than a first random value generated between 0 and 1, if yes, the second function value is taken as the current first optimal solution; the initial value of the current first optimal solution is the first function value; S601 repeating S600 until the number of sequence changes is greater than a preset first threshold, and outputting the current first optimal solution as the current second optimal solution; S602 replacing other options of the virtual transfer point based on a second rule, repeating S401 to generate a changed current sequence; inputting the changed current sequence into the first algorithm to obtain a changed first function value; repeating S600-S601 based on the changed first function value to calculate a changed current second optimal solution, and determining whether the changed current second optimal solution is better than the current second optimal solution, if yes, the changed current second optimal solution is taken as the current third optimal solution; if not, calculating a second probability based on the current temperature, the current third optimal solution and the changed current second optimal solution, and determining whether the second probability is greater than a second random value generated between 0 and 1, if yes, the changed current second optimal solution is taken as the current third optimal solution; S603 repeating S602 until the number of replacements is greater than a preset second threshold, and outputting the current third optimal solution as the current fourth optimal solution; S604 determining whether the current temperature is higher than a preset minimum temperature, if yes, changing the current temperature based on a preset cooling rate, and repeating sequentially performing S400, S401, S600-S603; otherwise, outputting the current fourth optimal solution as the optimal objective function value; the initial temperature of the current temperature is the highest temperature.

[0010] Optionally, the virtual space-time information is constructed, comprising: each candidate boarding point of each passenger corresponds to a virtual boarding point, and the virtual boarding point is represented by an identification number; each candidate transfer point of each passenger corresponds to a virtual transfer point, and the virtual transfer point is represented by an identification number; the virtual depot includes a plurality of branch vehicles located at the same starting point; and corresponding virtual road segments are determined based on the space mileage and travel time between each two of the virtual boarding point, the virtual transfer point and the virtual depot.

[0011] Optionally, the objective function is to minimize the total operating cost, and the expression is: min c1Σ j∈V Σ k∈K X ojk +c2Σ i∈V Σ j∈V Σk∈K m ij X ijk , where c1 represents the vehicle startup cost, X ojk is a decision variable, indicating that if branch vehicle k travels from virtual parking lot o to virtual node j, then X ojk is 1, otherwise, X ojk is 0; c2 represents the mileage cost, m ij is the virtual road segment E ij Space mileage, X ijk represents the decision variable, indicating that if branch vehicle k travels from virtual node i to virtual node j, then X ijk is 1, otherwise, X ijk is 0, V represents a virtual node set, which includes a virtual parking lot set, a virtual boarding point set and a virtual transfer point set, i represents an element in V, j represents another element in V, K represents a branch vehicle set, and k represents an element in K.

[0012] Optionally, the departure time of the branch line vehicle is calculated using the expression: Among them, o represents the identification number of the virtual parking lot, S ok represents the departure time of branch vehicle k from virtual parking lot o, X omk represents the decision variable, indicating that if the feeder vehicle k travels from the virtual parking lot o to the virtual node m, then X omk is 1, otherwise, X omk is 0; S mk represents the time when branch vehicle k arrives at the virtual boarding point m, h om V represents the travel time of the virtual road section from the virtual parking lot o to the virtual boarding point m, (B) Represents the set of virtual boarding points, m represents V (B) The elements in , K represents the set of branch vehicles, and k represents the elements in K.

[0013] Optionally, the number of passengers in the branch line vehicle is calculated using the expression: Among them, L mk represents the number of passengers of branch vehicle k leaving the virtual boarding point m, X imk represents the decision variable, indicating that if branch vehicle k travels from virtual node i to virtual boarding point m, then X imk is 1, otherwise, X imk is 0; L ik represents the passenger capacity of branch vehicle k leaving virtual node i, V represents the set of virtual nodes, i represents an element in V, V (B) Represents the set of virtual boarding points, m represents V (B)K is a set of branch vehicles, and k is an element in K.

[0014] Optionally, the time of branch vehicles arriving at virtual boarding points is calculated, and the expression is: wherein S mk represents the time of branch vehicle k arriving at virtual boarding point m, X mnk represents a decision variable, and X mnk is 1 if branch vehicle k travels from virtual boarding point m to another virtual boarding point n, otherwise, X mnk is 0; S nk represents the time of branch vehicle k arriving at virtual boarding point n, h mn represents the travel time of a virtual road section between virtual boarding point m and virtual boarding point n, V (B) represents a set of virtual boarding points, m represents an element in V (B) , and n represents another element in V. (B) K is a set of branch vehicles, and k is an element in K.

[0015] Optionally, the time of branch vehicles arriving at virtual transfer points is calculated, and the expression is: wherein S pk represents the time of branch vehicle k arriving at virtual transfer point p, X mpk represents a decision variable, and X mpk is 1 if branch vehicle k travels from virtual boarding point m to virtual transfer point p, otherwise, X mpk is 0; S mk represents the time of branch vehicle k arriving at virtual boarding point m, h mp represents the travel time of a virtual road section between virtual boarding point m and virtual transfer point p, V (B) represents a set of virtual boarding points, m represents an element in V (B) , and V (T) represents a set of virtual transfer points, p represents an element in V. (T) K is a set of branch vehicles, and k is an element in K.

[0016] Optionally, the time of passengers arriving at the terminal is calculated, and the expression is: wherein D mp represents the time of passenger corresponding to virtual boarding point m arriving at virtual transfer point p after transfer and then arriving at the terminal by trunk vehicle, X imk represents a decision variable, and X imkis 1, otherwise, X imk is 0; X jpk represents a decision variable, represents if branch vehicle k travels from virtual node j to virtual transfer point p, X jpk is 1, otherwise, X jpk is 0; S pk represents the time when branch vehicle k arrives at virtual transfer point p, d m represents the trunk end point of the passenger corresponding to virtual boarding point m, d represents the travel time from virtual transfer point p to trunk end point d m , V (B) represents a set of virtual boarding points, m represents an element in V (B) , V (T) represents a set of virtual transfer points, p represents an element in V (T) , K represents a set of branch vehicles, k represents an element in K, V represents a set of virtual nodes, i represents an element in V, and j represents another element in V.

[0017] Compared with the prior art, the present application provides a vehicle scheduling and travel path optimization method for coordinating trunk and branch line bus services, which supports passengers to simultaneously candidate multiple boarding points and transfer points, and considers the influence of trunk travel time on branch traffic routing and scheduling. In consideration of transportation demand and system limitations, the travel path of the passenger, the routing information and the time information of the branch vehicle are determined to minimize the total cost. The present application solves the problems of poor connection between trunk and branch traffic mode operation, individual cognitive dependence and limitation in passenger travel, low shared ride potential of branch traffic mode, etc. in the public transportation system, improves the passenger carrying rate and operation efficiency of the branch vehicle, enhances the overall travel experience of the passengers, and reduces the operation cost and passenger travel cost. In addition, the present application helps to promote public transportation travel selection, reduce carbon emissions and energy consumption. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments or prior art of the present application, the drawings needed in the embodiment or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art according to these drawings.

[0019] Figure 1 is a flowchart of a vehicle scheduling and travel path optimization method for coordinating trunk and branch line bus services provided by the present embodiment;

[0020] Figure 2 is a schematic diagram of the relationship between the selected virtual transfer point and the corresponding actual transfer point in the present embodiment;

[0021] Figure 3 is a schematic diagram of the relationship between the virtual boarding point selected in this embodiment and the corresponding actual boarding point;

[0022] Figure 4 Schematic diagram of a sequence of virtual boarding points corresponding to actual transfer points in this embodiment. DETAILED DESCRIPTION

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0024] To facilitate understanding of the embodiments of the present invention, the following will be further explained with reference to specific embodiments in conjunction with the accompanying drawings. The embodiments do not constitute a limitation on the scope of protection of the present invention.

[0025] This embodiment provides a method for coordinating vehicle scheduling and travel route optimization for trunk and branch bus services, the process of which is as follows: Figure 1 Shown, including:

[0026] S1 obtains passenger demand information within a preset time period.

[0027] The scenario of this embodiment is as follows: a passenger walks from the starting point to the boarding point, takes a branch line bus to the transfer point, and then takes a main line bus to the main line terminal. The transfer point is a stop for the main line bus. Each branch line bus corresponds to a transfer point.

[0028] The demand information includes at least one branch line waiting point, at least one branch line waiting point, the main line terminal and the service time window, as shown in Table 1.

[0029] Table 1

[0030]

[0031] Passenger identification numbers can be distributed based on the order in which they submit their requests. Branch line boarding points and branch-to-main line transfer points can be identified based on location information or other rules. The service time window includes the earliest time a passenger can board the bus and the latest time they can arrive at the main line destination. The earliest and latest times can be different for each passenger. In this embodiment, the earliest time for each passenger is uniformly set to 8:00 and the latest time is set to 9:00. The preset time period is set to 8:00-9:00.

[0032] S2 constructs virtual space-time information based on the relationship between each passenger and the candidate boarding point and the candidate transfer point.

[0033] The virtual space-time information includes virtual boarding points, virtual transfer points, virtual depots and virtual road segments.

[0034] Each candidate boarding point of each passenger corresponds to a virtual boarding point, and the virtual boarding point is represented by an identification number; each candidate transfer point of each passenger corresponds to a virtual transfer point, and the virtual transfer point is represented by an identification number; the virtual depot includes multiple branch vehicles located at the same starting point; based on the space mileage and travel time between each two of the virtual boarding points, the virtual transfer points and the virtual depots, the corresponding virtual road segments are determined.

[0035] For example, the identification number of the virtual depot is 0; based on the identification number of the passenger, each candidate boarding point of the passenger is identified in sequence to obtain the virtual boarding point, for example, the candidate boarding point of passenger 1 is 1 and 4, and the virtual boarding point of the candidate boarding point 1 of passenger 1 is 1, and the virtual boarding point of the candidate boarding point 4 of passenger 1 is 2. Assuming that all passengers generate b virtual boarding points, a virtual boarding point set V (B) ={1,2,…,b} is formed, as shown in Table 2.

[0036] Table 2

[0037]

[0038] Each candidate transfer point of the passenger is identified, and it is assumed that all passengers generate t virtual transfer points to form a virtual transfer point set V (T) ={b+1,b+2,…,b+t}, as shown in Table 3.

[0039] Table 3

[0040]

[0041] Each two of the virtual nodes i,j forms a directed virtual road segment E ij , and a virtual road segment set E={E ij} is formed. The virtual space-time information G=(V,E) is recorded, that is, a weighted directed graph, wherein V represents a virtual node set, and the virtual node set includes a virtual depot set, a virtual boarding point set and a virtual transfer point set.

[0042] The mapping of the virtual boarding point set V (B) , the virtual transfer point set V (T) , the passenger set P, the candidate boarding point set B and the candidate transfer point set T is established:

[0043] f1:V (B)→P represents the surjective non-injective relationship between the set of virtual boarding points and the set of passengers;

[0044] f2:V (T) →P represents the surjective non-injective mapping from the set of virtual transfer points to the set of passengers;

[0045] f3:V (B) →B indicates that the set of virtual boarding points is neither surjective nor injective to the set of actual boarding points;

[0046] f4:V (T) →T represents the non-surjective and non-injective transition from the set of virtual transfer points to the set of actual transfer points.

[0047] By establishing a mapping relationship, you can query the other party's information through the identification number.

[0048] Each virtual road segment includes spatial mileage m ij and travel time h ij Table 4 details the spatial mileage and travel time for various situations. For example, when the starting point of the road segment is a virtual parking lot and the end point is also a virtual parking lot, the spatial mileage and travel time are both ∞; when the starting point of the road segment is a virtual boarding point i2 and the end point is also a virtual boarding point j2, the spatial mileage is determined based on the mapping relationship f3 between the virtual boarding point and the actual boarding point. for Travel time ij for Among them, f3(i2) represents the real boarding point corresponding to the virtual boarding point i2, and f3(j2) represents the real boarding point corresponding to the virtual boarding point j2. Represents the spatial mileage between two real ride points, Represents the travel time between two real boarding points; similarly, the spatial mileage and travel time of any virtual road segment in the table can be obtained.

[0049] Table 4

[0050]

[0051] S3 builds the target function.

[0052] In this embodiment, the objective function is to minimize the total operating cost, and its expression is:

[0053]

[0054] Among them, c1 represents the vehicle startup cost, X ojk is a decision variable, indicating that if branch vehicle k travels from virtual parking lot o to virtual node j, then x ojk is 1, otherwise, X ojk is 0; c2 represents the mileage cost, m ij is the virtual road segment Eij Space mileage, X ijk represents the decision variable, indicating that if branch vehicle k travels from virtual node i to virtual node j, then X ijk is 1, otherwise, X ijk is 0, V represents the set of virtual nodes, i represents an element in V, j represents another element in V, K represents the set of branch vehicles, and k represents an element in K.

[0055] S4 randomly selects a virtual transfer point and a virtual boarding point to generate the current sequence.

[0056] This step includes:

[0057] S400 randomly selects one of the virtual transfer points corresponding to each passenger, and determines the corresponding transfer point to be selected as the actual transfer point.

[0058] like Figure 2 As shown, each passenger randomly selects a virtual transfer point based on mapping f2, and the actual transfer point corresponding to the selected virtual transfer point is determined based on mapping f4. For example, passenger number 1 selects virtual transfer point 20, and the candidate transfer point 22 is determined as the actual transfer point.

[0059] S401 determines potential carpooling passengers based on the same actual transfer point; randomly selects one of the virtual boarding points corresponding to the potential carpooling passengers, and generates a boarding point sequence corresponding to each actual transfer point as the current sequence.

[0060] Based on the same actual transfer point, potential carpooling passengers are determined. As shown in Table 5, for the actual transfer point identified as 22, passenger 1 and passenger 2 may share a feeder vehicle.

[0061] Table 5

[0062] Actual transfer point Passenger 22 1,2 23 4,5,6,7 24 3,8,9 25 10

[0063] Randomly select one of the virtual boarding points corresponding to the potential carpooling passengers, and generate a boarding point sequence corresponding to each actual transfer point as the current sequence.

[0064] like Figure 3 As shown, each passenger randomly selects a virtual boarding point based on the mapping f1, and the actual boarding point corresponding to the selected virtual boarding point is determined based on the mapping f3.

[0065] like Figure 4As shown, from the virtual boarding points 1 and 2 corresponding to the passenger 1, the virtual boarding point 2 is randomly selected, from the virtual boarding points 3, 4 and 5 corresponding to the passenger 2, the virtual boarding point 3 is randomly selected, thereby generating the boarding point sequence corresponding to the actual transfer point 22 as {virtual boarding point 2, virtual boarding point 3}. Similarly, the boarding point sequences corresponding to other actual transfer points can be obtained. Figure 3 The boarding point sequence is taken as the current sequence.

[0066] S5 inputs the current sequence into the first algorithm to obtain a first function value.

[0067] The first algorithm includes:

[0068] A feeder vehicle is pre-assigned to each input sequence.

[0069] For example, the input sequence is {virtual boarding point 1, virtual boarding point 2, virtual boarding point 3, virtual boarding point 4}, and the current sequence obtained is {virtual boarding point 2, virtual boarding point 3}. Figure 4 The current sequence has four actual transfer points, and each actual transfer point corresponds to a current sequence, so four feeder vehicles are initially assigned.

[0070] The first station of the feeder vehicle is the first boarding point in the input sequence.

[0071] For example, the boarding point sequence corresponding to the actual transfer point 22 is {virtual boarding point 2, virtual boarding point 3}, and the first station of the corresponding feeder vehicle is the virtual boarding point 2. A feeder vehicle is dispatched from the virtual parking lot to the corresponding first station to pick up the corresponding passengers.

[0072] The departure time of the feeder vehicle is calculated based on the time at which the passenger arrives at the first station at the earliest time.

[0073] The earliest time for the passenger 1 to board at the virtual boarding point 2 is 8:00, that is, the time of arriving at the first station is 8:00, and based on this time, the departure time of the feeder vehicle is calculated, and the expression is:

[0074]

[0075] Wherein, o represents the identification number of the virtual parking lot, S ok represents the departure time of the feeder vehicle k from the virtual parking lot o, X omk represents a decision variable, and represents that if the feeder vehicle k travels from the virtual parking lot o to the virtual node m, X omk is 1, otherwise, X omk is 0; S mk represents the time at which the feeder vehicle k arrives at the virtual boarding point m, h om represents the travel time of the virtual road section between the virtual parking lot o and the virtual boarding point m, V (B) represents a set of virtual boarding points, and m represents V(B) K represents a set of branch vehicles, and k represents an element in K.

[0076] The virtual boarding points in the input sequence are traversed in sequence. When a boarding point in the input sequence is reached, it is determined whether the first condition can be met if the branch vehicle reaches the boarding point.

[0077] The first condition includes that the number of passengers in the branch vehicle does not exceed the preset passenger capacity, the time at which the passengers arrive at the boarding point is greater than the earliest time, and the time at which the passengers arrive at the trunk terminal is less than the latest time.

[0078] It should be noted that the embodiment needs to determine, for each passenger of the branch vehicle, whether the time at which the passenger arrives at the boarding point is greater than the earliest time, and whether the time at which each passenger arrives at the trunk terminal is less than the latest time. If one of the passengers in the branch vehicle does not meet the requirements, the branch vehicle does not go to the boarding point; and in the determination process, the operation is based on the boarding point as the last boarding point of the branch vehicle.

[0079] The expression for calculating the number of passengers in the branch vehicle is:

[0080]

[0081] wherein L mk represents the number of passengers of the branch vehicle k leaving the virtual boarding point m, X imk represents a decision variable, X imk is 1 if the branch vehicle k travels from the virtual node i to the virtual boarding point m, otherwise, X imk is 0; L ik represents the passenger capacity of the branch vehicle k leaving the virtual node i, V represents a set of virtual nodes, i represents an element in V, V (B) represents a set of virtual boarding points, m represents an element in V (B) represents a set of branch vehicles, and k represents an element in K.

[0082] When the first condition is met, the boarding point is added to the route information of the branch vehicle, the time at which the boarding point is reached is added to the arrival time information, and it is determined whether the next boarding point in the input sequence meets the first condition.

[0083] When the first condition is not met, the last boarding point is taken as the terminal station of the branch vehicle, a new branch vehicle is dispatched from the virtual depot, the boarding point is taken as the first station of the new branch vehicle, and the remaining boarding points in the input sequence are traversed.

[0084] As Figure 4As shown, for the virtual transfer point 24, the traversal sequence is {virtual boarding point 6, virtual boarding point 15, virtual boarding point 16}. When the traversal reaches the virtual boarding point 6, the first condition is satisfied because the branch vehicle arrives at the first station at the earliest time. When the traversal reaches the virtual boarding point 15, it is determined whether the first condition can be satisfied if the branch vehicle arrives at the virtual boarding point 15. If the first condition can be satisfied, the virtual boarding point 15 is added to the route information of the branch vehicle, the time of arrival at the virtual boarding point 15 is added to the arrival time information, and the traversal continues to the virtual boarding point 16. If the first condition cannot be satisfied, the virtual boarding point 6 is taken as the terminal station of the branch vehicle, and another branch vehicle is dispatched from the virtual depot. The virtual boarding point 15 is taken as the first station of the new branch vehicle, the traversal continues to the virtual boarding point 16, and it is determined whether the first condition can be satisfied.

[0085] When the traversal of the input sequence is completed, the route information and the arrival time information of each vehicle are obtained, and the corresponding objective function value is determined based on the constructed objective function.

[0086] The route information is composed of a plurality of virtual nodes. The branch vehicle departs from the virtual depot, visits a plurality of virtual boarding points, and arrives at a plurality of virtual transfer points. The virtual nodes include the virtual depot, the virtual boarding points arrived at by the branch vehicle, and the virtual transfer points. Exemplarily, as shown in Table 6, the route information of the branch vehicle corresponding to the actual transfer point 22 is that the branch vehicle departs from the virtual depot 0, carries passengers at the virtual boarding point 2 and the virtual boarding point 3, and travels to the actual transfer point 22. Since the actual transfer point 22 corresponds to the virtual transfer point 20 and the virtual transfer point 22, the route information of the branch vehicle traveling to the actual transfer point 22 is [0, 2, 3, 20, 22].

[0087] Table 6

[0088]

[0089] The arrival time information records the time of arrival of the branch vehicle at each virtual node in the route information, including the departure time of the branch vehicle, the arrival time at the virtual boarding point, and the arrival time at the transfer point. Except for the virtual depot, each virtual node uniquely corresponds to an actual boarding point or an actual transfer point of a passenger. Exemplarily, as shown in Table 7, the arrival time information is that the branch vehicle departs from the virtual depot 0 at 472 minutes, arrives at the virtual boarding point 2 at 480 minutes, arrives at the virtual boarding point 3 at 484 minutes, and arrives at the virtual transfer point 20 and the virtual transfer point 22 at 499 minutes.

[0090] Table 7

[0091]

[0092] The expression for calculating the time of arrival of the branch vehicle at the virtual boarding point is:

[0093]

[0094] where S mk denotes the time that branch vehicle k arrives at virtual boarding point m, X mnk denotes a decision variable, and X mnk is 1 if branch vehicle k travels from virtual boarding point m to another virtual boarding point n, otherwise, X mnk is 0; S nk denotes the time that branch vehicle k arrives at virtual boarding point n, h mn denotes the travel time of the virtual link between virtual boarding point m and virtual boarding point n, V (B) denotes a set of virtual boarding points, m denotes an element in V (B) , and n denotes another element in V (B) ; K denotes a set of branch vehicles, and k denotes an element in K.

[0095] The expression for calculating the time that a branch vehicle arrives at a virtual transfer point is:

[0096]

[0097] where S pk denotes the time that branch vehicle k arrives at virtual transfer point p, X mpk denotes a decision variable, and X mpk is 1 if branch vehicle k travels from virtual boarding point m to virtual transfer point p, otherwise, X mpk is 0; S mk denotes the time that branch vehicle k arrives at virtual boarding point m, h mp denotes the travel time of the virtual link between virtual boarding point m and virtual transfer point p, V (B) denotes a set of virtual boarding points, m denotes an element in V (B) ; V (T) denotes a set of virtual transfer points, p denotes an element in V (T) ; K denotes a set of branch vehicles, and k denotes an element in K.

[0098] The expression for calculating the time that a passenger arrives at the terminal is:

[0099]

[0100] where D mp denotes the time that passenger corresponding to virtual boarding point m arrives at virtual transfer point p after transferring and then takes the trunk vehicle to arrive at the terminal, X imk denotes a decision variable, and X imkis 1, otherwise, X imk is 0; X jpk represents a decision variable, represents if branch vehicle k travels from virtual node j to virtual transfer point p, X jpk is 1, otherwise, X jpk is 0; S pk represents the time when branch vehicle k arrives at virtual transfer point p, d m represents the trunk end point of the passenger corresponding to virtual boarding point m, d represents the travel time from virtual transfer point p to trunk end point d m , V (B) represents a set of virtual boarding points, m represents an element in V (B) , V (T) represents a set of virtual transfer points, p represents an element in V (T) , K represents a set of branch vehicles, k represents an element in K, V represents a set of virtual nodes, i represents an element in V, and j represents another element in V.

[0101] S6 finds the optimal objective function value by changing the current sequence using the annealing algorithm, and plans the branch vehicle scheduling and passenger travel path based on the optimal objective function value.

[0102] Finding the optimal objective function value by changing the current sequence using the annealing algorithm, comprising:

[0103] S600 changes the current sequence based on the first rule, inputs the changed current sequence into the first algorithm to obtain a second function value; judges whether the second function value is better than the current first optimal solution, if it is better than the current first optimal solution, the second function value is output as the current first optimal solution; otherwise, based on the current temperature, the current first optimal solution and the second function value, a first probability is calculated, and it is judged whether the first probability is greater than a first random value generated between 0 and 1, if it is greater, the second function value is taken as the current first optimal solution; the initial value of the current first optimal solution is the first function value.

[0104] The first rule includes reset, exchange, move and reverse, and a new current sequence is randomly selected with the same probability.

[0105] Reset: randomly select any point in the sequence, and reset the boarding point at the point to any other virtual boarding point corresponding to the same passenger. If the same passenger has no other virtual boarding point, the point is reselected.

[0106] Exchange: randomly select any two points in the sequence, and exchange the positions of the virtual boarding points at the points.

[0107] Move: randomly select any two points in the sequence, and move the virtual boarding point at the front point to the position immediately after the rear point.

[0108] Reverse: randomly select any two points in the sequence, reverse the sub-sequence composed of the two points and the virtual transfer points between the two points.

[0109] According to the formula Calculate the first probability, where a represents the first probability, exp() represents the natural exponential function, represents the second function value, represents the current first optimal solution, and T represents the current temperature.

[0110] S601 repeats S600 until the number of sequence changes is greater than a preset first threshold, and the output current first optimal solution is taken as the current second optimal solution.

[0111] The first threshold represents the maximum number of times the current sequence is changed.

[0112] S602 replaces other choices of virtual transfer points based on a second rule, repeats S401 to generate a replaced current sequence; inputs the replaced current sequence into the first algorithm to obtain a replaced first function value; repeats S600-S601 based on the replaced first function value to calculate a replaced current second optimal solution, judges whether the replaced current second optimal solution is better than the current second optimal solution, if yes, takes the replaced current second optimal solution as the current third optimal solution; if not, calculates a second probability based on the current temperature, the current third optimal solution and the replaced current second optimal solution, and judges whether the second probability is greater than a second random value generated between 0 and 1, if yes, takes the replaced current second optimal solution as the current third optimal solution.

[0113] The second rule includes randomly selecting a virtual transfer point selected by a passenger, and resetting the virtual transfer point to any other virtual transfer point corresponding to the same passenger. If the same passenger has no other virtual transfer point, it is reselected.

[0114] According to the formula Calculate the second probability, where b represents the second probability, exp() represents the natural exponential function, represents the replaced current second optimal solution, represents the current third optimal solution, and T represents the current temperature.

[0115] S603 repeats S602 until the number of replacements is greater than a preset second threshold, and the output current third optimal solution is taken as the current fourth optimal solution.

[0116] The second threshold represents the maximum number of times the virtual transfer point is replaced.

[0117] S604 judges whether the current temperature is higher than the preset minimum temperature. If the current temperature is higher than the preset minimum temperature, the current temperature is changed based on the preset cooling rate, and S400, S401, S600-S603 are repeatedly executed in turn. Otherwise, the output current fourth optimal solution is taken as the optimal objective function value. The initial temperature of the current temperature is the highest temperature.

[0118] Based on the optimal objective function value, corresponding routing information and arrival time information are obtained, thereby planning branch vehicle scheduling and passenger travel paths.

[0119] In addition, the method for finding the optimal objective function value can also use other heuristic algorithms such as genetic algorithm, particle swarm algorithm or variable field search algorithm.

[0120] Compared with the prior art, the embodiment provides a vehicle scheduling and travel path optimization method for coordinating trunk and branch line bus services, supports passengers to simultaneously candidate multiple boarding points and transfer points, and considers the influence of trunk travel time on branch traffic routing and scheduling. In consideration of transportation demand and system limitations, the travel path of the passenger, the routing information and the time information of the branch vehicle are determined to minimize the total cost. The embodiment solves the problems of poor operation connection of trunk-branch traffic mode in the public transportation system, individual cognitive dependence and limitation in passenger travel, low shared potential of branch traffic mode, etc., improves the passenger carrying rate and operation efficiency of the branch vehicle, enhances the overall travel experience of the passengers, and reduces the operation cost and passenger travel cost. In addition, the embodiment helps to promote public transportation travel selection, reduces carbon emissions and energy consumption.

[0121] The specific embodiments described above further illustrate the purposes, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for coordinating vehicle dispatch and trip path optimization for trunk-and-feeder bus services, the method comprising: The method comprises the following steps: obtaining demand information of passengers in a preset time period, the demand information comprising at least one branch line candidate boarding point, at least one branch line candidate transfer point, a trunk line terminal and a service time window, the service time window comprising an earliest time for the passengers to board and a latest time for the passengers to arrive at the trunk line terminal; constructing virtual space-time information based on the relationship between each passenger and the candidate boarding point and the candidate transfer point, the virtual space-time information comprising a virtual boarding point, a virtual transfer point, a virtual depot and a virtual road section; constructing an objective function; randomly selecting the virtual transfer point and the virtual boarding point to generate a current sequence; inputting the current sequence into a first algorithm to obtain a first function value, the first algorithm comprising: pre-allocating a branch line vehicle for each input sequence; taking the first boarding point in the input sequence as the first station of the branch line vehicle; calculating the departure time of the branch line vehicle based on the time at which the passengers arrive at the first station at the earliest time; sequentially traversing the virtual boarding points in the input sequence, and when a boarding point in the input sequence is traversed, judging whether a first condition can be met if the branch line vehicle arrives at the boarding point; the first condition comprising that the number of passengers in the branch line vehicle does not exceed a preset passenger capacity, the time at which the passengers arrive at the boarding point is greater than the earliest time, and the time at which the passengers arrive at the trunk line terminal is less than the latest time; when the first condition is met, adding the boarding point to the route information of the branch line vehicle, adding the time at which the boarding point is arrived to the arrival time information, and continuing to traverse the next boarding point in the input sequence; when the first condition is not met, dispatching another branch line vehicle from the virtual depot, taking the boarding point as the first station of the new branch line vehicle, and continuing to traverse the remaining boarding points in the input sequence based on the new branch line vehicle; after the traversal is completed, obtaining the route information and the arrival time information of each branch line vehicle, and determining the corresponding objective function value; finding an optimal objective function value by changing the current sequence using a simulated annealing algorithm, and planning branch line vehicle scheduling and passenger travel paths based on the optimal objective function value.

2. The method of claim 1, wherein, randomly selecting the virtual transfer point and the virtual boarding point to generate a current sequence, comprising: S400 randomly selecting one of the virtual transfer points corresponding to each passenger and determining the corresponding candidate transfer point as an actual transfer point; S401 determining passengers who can potentially share a vehicle based on the same actual transfer point, and randomly selecting one of the virtual boarding points corresponding to the passengers who can potentially share a vehicle to generate a boarding point sequence corresponding to each actual transfer point as the current sequence.

3. The method of claim 2, wherein, finding an optimal objective function value by changing the current sequence using a simulated annealing algorithm, comprising: S600 changing the current sequence based on a first rule, inputting the changed current sequence into the first algorithm to obtain a second function value, judging whether the second function value is better than the current first optimal solution, if the second function value is better than the current first optimal solution, taking the second function value as the current first optimal solution; otherwise, calculating a first probability based on the current temperature, the current first optimal solution and the second function value, and judging whether the first probability is greater than a first random value generated between 0 and 1, if the first probability is greater than the first random value, taking the second function value as the current first optimal solution; the initial value of the current first optimal solution being the first function value. S601 repeats S600 until the number of sequence changes is greater than a preset first threshold, and outputs the current first optimal solution as a current second optimal solution; S602 replaces other options of the virtual transfer point based on a second rule, repeats S401 to generate a replaced current sequence, inputs the replaced current sequence into the first algorithm to obtain a replaced first function value, repeats S600-S601 based on the replaced first function value to calculate a replaced current second optimal solution, judges whether the replaced current second optimal solution is better than the current second optimal solution, if yes, takes the replaced current second optimal solution as a current third optimal solution, if not, calculates a second probability based on the current temperature, the current third optimal solution and the replaced current second optimal solution, and judges whether the second probability is greater than a second random value generated between 0 and 1, if yes, takes the replaced current second optimal solution as the current third optimal solution; S603 repeats S602 until the number of replacements is greater than a preset second threshold, and outputs the current third optimal solution as a current fourth optimal solution; S604 judges whether the current temperature is higher than a preset minimum temperature, if yes, changes the current temperature based on a preset cooling rate, and repeats sequentially S400, S401, S600-S603, otherwise, outputs the current fourth optimal solution as an optimal target function value, and an initial temperature of the current temperature is a highest temperature.

4. The method of claim 1, wherein, Constructing virtual space-time information, comprising: Each candidate boarding point of each passenger corresponds to a virtual boarding point, and the virtual boarding point is represented by an identification number; each candidate transfer point of each passenger corresponds to a virtual transfer point, and the virtual transfer point is represented by an identification number; the virtual depot includes a plurality of branch vehicles located at the same starting point; based on the space mileage and travel time between each two of the virtual boarding point, the virtual transfer point and the virtual depot, a corresponding virtual road section is determined.

5. The method of claim 1, wherein, The target function is to minimize the total operating cost, and its expression is: wherein cl represents a vehicle start cost, X ojk is a decision variable, representing if feeder vehicle k travels from virtual depot o to virtual node j, X ojk is 1, otherwise, X ojk is 0; c2 represents a mileage travel cost, m ij is a spatial mileage of virtual route E ij , X ijk represents a decision variable, representing if feeder vehicle k travels from virtual node i to virtual node j, X ijk is 1, otherwise, X ijk is 0, V represents a virtual node set, the virtual node set including a virtual depot set, a virtual boarding point set and a virtual transfer point set, i represents an element in V, j represents another element in V, K represents a feeder vehicle set, and k represents an element in K.

6. The method of claim 1, wherein, The expression for calculating the departure time of the branch vehicle is: wherein o denotes an identification number of a virtual yard, S ok denotes a departure time of branch vehicle k from virtual yard o, X omk denotes a decision variable, and X omk is 1 if branch vehicle k travels from virtual yard o to virtual node m, otherwise, X omk is 0; S mk denotes a time at which branch vehicle k arrives at virtual boarding point m, h om denotes a travel time of a virtual route between virtual yard o and virtual boarding point m, V (B) denotes a set of virtual boarding points, m denotes an element in V (B) , and K denotes a set of branch vehicles, k denotes an element in K.

7. The method of claim 1, wherein, The expression for calculating the number of passengers in the branch vehicle is: where L mk represents the number of passengers that branch vehicle k leaves at virtual boarding point m, X imk represents a decision variable, X imk is 1 if branch vehicle k travels from virtual node i to virtual boarding point m, otherwise, X imk is 0; L ik represents the number of passengers that branch vehicle k leaves at virtual node i, V represents a set of virtual nodes, i represents an element in V (B) represents a set of virtual boarding points, m represents an element in V (B) represents a set of branch vehicles, k represents an element in K.

8. The method of claim 1, wherein, The expression for calculating the time when the branch vehicle arrives at the virtual boarding point is: where S mk denotes the time when branch vehicle k arrives at virtual boarding point m, X mnk denotes a decision variable, X mnk is 1 if branch vehicle k travels from virtual boarding point m to another virtual boarding point n, otherwise, X mnk is 0; S nk denotes the time when branch vehicle k arrives at virtual boarding point n, μ denotes the average service time per passenger at a station, h mn denotes the travel time of a virtual link between virtual boarding point m and virtual boarding point n, V (B) denotes a set of virtual boarding points, m denotes an element in V (B) , n denotes another element in V (B) , and K denotes a set of branch vehicles, k denotes an element in K.

9. The method of claim 1, wherein, The expression for calculating the time when the branch vehicle arrives at the virtual transfer point is: where S pk denotes the time at which branch vehicle k arrives at virtual transfer point p, X mpk denotes a decision variable, and X mpk is 1 if branch vehicle k travels from virtual boarding point m to virtual transfer point p, otherwise, X mpk is 0; S mk denotes the time at which branch vehicle k arrives at virtual boarding point m, h mp denotes the travel time of the virtual link between virtual boarding point m and virtual transfer point p, V (B) denotes a set of virtual boarding points, m denotes an element of V (B) , V (T) denotes a set of virtual transfer points, p denotes an element of V (T) , K denotes a set of branch vehicles, and k denotes an element of K.

10. The method of claim 1, wherein, The expression for calculating the time when the passenger arrives at the terminal is: where D mp represents the time when the passenger corresponding to the virtual boarding point m arrives at the virtual transfer point p and takes the trunk vehicle after transfer to arrive at the terminal, X imk represents a decision variable, and represents that if the branch vehicle k travels from the virtual node i to the virtual boarding point m, X imk is 1, otherwise, X imk is 0; X jpk represents a decision variable, and represents that if the branch vehicle k travels from the virtual node j to the virtual transfer point p, X jpk is 1, otherwise, X jpk is 0; S pk represents the time when the branch vehicle k arrives at the virtual transfer point p, d m represents the trunk terminal of the passenger corresponding to the virtual boarding point m, represents the travel time from the virtual transfer point p to the trunk terminal d m , V (B) represents a set of virtual boarding points, m represents an element in V (B) , V (T) represents a set of virtual transfer points, p represents an element in V (T) , K represents a set of branch vehicles, k represents an element in K, V represents a set of virtual nodes, i represents an element in V, and j represents another element in V.

Images