Customized bus route design method for regional management and control

Abstract

The invention provides a customized bus route design method for regional management and control, and relates to the technical field of traffic, and the method comprises the following steps: S1, calculating whether all related routes violate a regional management and control policy or not, and building an improved A * algorithm to correct illegal routes; s2, establishing a customized bus route optimization model based on the corrected route, and calculating a customized bus driving route network; according to the method, the influence of a regional management and control policy on customized bus route design and getting on and off at a riding site selection station is researched, a regional management and control customized bus route optimization model and a passenger site selection model are constructed, and a reasonable customized bus route is calculated by using the regional management and control customized bus route optimization model; and the getting-on and getting-off probability of each station of the passenger site selection is calculated through the passenger site selection model, so that a more accurate calculation model is formulated for a customized bus operator affected by a regional management and control policy in an emergency event, the benefit of the customized bus operator is improved, and the safety of passengers and drivers is protected.

Description

Technical Field

[0001] The present invention relates to the field of transportation technology, and in particular to a method for designing customized bus routes for regional control. Background Art

[0002] In modern cities, public transportation systems play an important role in providing residents with convenient and affordable transportation services. However, with the acceleration of urbanization and population growth, public transportation systems are facing new challenges and demands. The public transportation systems in many cities are mainly composed of conventional buses and rail transportation. These transportation tools have fixed routes and cannot achieve full coverage of the city, and it is difficult to provide complete "point-to-point services". As the scale of cities expands, the fixed route characteristics of traditional buses limit their ability to meet the diverse and specialized needs of passengers, and may not be able to meet the needs of all regions, thus restricting the further improvement of public transportation service levels. For example, sparsely populated marginal areas, industrial parks and other places often face problems such as a small number of passengers, few bus stops, empty seats and long distances between passenger destinations and bus stops. In order to alleviate the above problems and improve the level of public transportation services, some regions have begun to try to provide customized public transportation services for the public in the operating mode of taxis, flexibly adjust routes and vehicle operations according to actual needs and passengers' travel needs, and provide more personalized services;

[0003] Customized buses meet the growing and diversified transportation needs to a large extent. Therefore, customized buses have a wide range of application scenarios, such as customized tourist buses, customized school buses, customized medical lines, customized emergency buses, etc. Emergency events are sudden, highly harmful and unpredictable, and may be accompanied by sudden changes in demand and traffic paralysis. Customized buses, as a flexible and efficient transportation mode, can quickly respond to and meet the transportation needs of emergency materials and rescue personnel. Through reasonable path optimization and scheduling planning, customized buses can provide rapid and safe transportation services in dangerous areas to ensure the timely transportation of key materials and the safe evacuation of personnel. However, poor road conditions may prevent customized buses from arriving at their destinations on time, prolonging the waiting time for passengers, or even failing to provide services in certain areas. Secondly, safety risks may increase. Emergency events pose a threat to the personal safety of passengers. Passengers may be in a state of tension and panic, leading to safety issues such as congestion, disputes or even criminal behavior, thereby affecting the operation of customized buses. However, due to the lack of relevant research and technical guidance on customized bus operation strategies for regional control, customized bus operators cannot scientifically and effectively maintain regional control while maximizing their own profits, which is specifically reflected in the following aspects:

[0004] (1) The types of traffic restrictions studied in existing studies on path optimization under traffic restrictions include traffic jams, road bans, alternating traffic restrictions, and regional bans. Most studies on regional bans use the method of modifying the traffic network diagram to achieve regional bans, and do not consider the situation where the traffic network diagram cannot be used for planning when the road is severely damaged. At the same time, there are few studies using the VRP model to study path optimization models under emergency events. The VRP model can effectively solve the path optimization problem without a traffic network diagram. Therefore, the VRP model under emergency events and the corresponding solution algorithm need to be further explored;

[0005] (2) At present, the research on emergency bus site selection is mainly divided into two categories. One is to study the joint optimization problem of station selection and route optimization, which mainly considers the impact of emergency events on route optimization problems. The other is to directly study the station selection problem, which considers the impact of emergency events on station selection. The impact of emergency events on passengers is usually described by passenger waiting time constraints, and most of the constraints are soft constraints. The situation of regional control under emergency events is not taken into account. Even the research that considers regional control does not directly link the passenger feasible range with passenger decision-making. In addition, most scholars study the location of bus stops, but there is little research on passenger selection of bus stops. Although customized buses are a highly personalized bus vehicle, they still cannot achieve true "point-to-point service", which requires a lot of resources. Therefore, passengers need to go to the bus stop to receive services according to their own conditions. In this process, some passengers will be affected by regional control and need to take a detour or cannot go to the nearest bus stop. However, there is currently no direct calculation model for calculating the station selection of customized bus passengers under regional control constraints.

[0006] In summary, regional control policies will have a serious impact on customized bus operators and passengers. Bus routes will be modified or even cancelled due to the control area. Some passengers will not be able to accept customized bus services or change service content due to regional control. No research has added regional control constraints while designing bus routes and analyzing passenger site selection. Under the background of regional control, on the one hand, customized bus operators hope to increase passenger acceptance rate and design the shortest route to increase operating income. On the other hand, they hope to follow the control policy and cancel some passenger orders and high-risk routes, resulting in strategic planning guidance conflicts. At the same time, passengers hope to reduce risks while accepting customized public security services. Therefore, the present invention proposes a customized bus route design method for regional control to solve the problems existing in the prior art. Summary of the invention

[0007] In view of the above problems, the present invention proposes a method for designing a customized bus route for regional control, which improves the interests of customized bus operators and protects the safety of passengers and drivers.

[0008] To achieve the object of the present invention, the present invention is implemented through the following technical solutions: A customized bus route design method for area control, comprising the following steps:

[0009] S1: Calculate whether all relevant routes violate the area control policy, and establish an improved A* algorithm to correct the violated routes;

[0010] S2: Based on the corrected routes, establish an optimized model for customized bus routes, and calculate the customized bus driving route network;

[0011] S3: Based on the customized bus driving route, establish a passenger travel selection model, and calculate the probability of passengers choosing each travel plan;

[0012] S4: Based on the passenger selection probability, calculate the number of customized buses dispatched, and obtain the customized bus travel routes, quantities, and the probability of each passenger's travel choice.

[0013] Further improvement lies in: The S1 includes the following steps:

[0014] S11: Calculate whether the connection line of bus stops will cross the controlled area;

[0015] S12: Correct or cancel the route for the line that crosses the controlled area.

[0016] Further improvement lies in: The S11 is specifically: Set the controlled area as a circle or a convex polygon. According to geometric relationships, verify whether the connection line between stop A(x1,y1) and stop B(x2,y2) intersects the controlled area. The calculation method is as follows:

[0017] When the controlled area is a circle (the center is (x0,y0) and the radius is r):

[0018]

[0019] When d < r, it indicates that the AB connection line intersects the controlled area;

[0020] When the controlled area is a convex polygon (each side is (p i ,p i+1 )):

[0021] Calculate the cross product:

[0022] cross(A,B,C) = (B x -A x )(C y -A y )-(B y -A y )(C x -A x )

[0023] For each edge of the polygon (p i ,p i+1 ) The condition for determining the intersection of line segments (x1, y1) and (x2, y2) is:

[0024] cross(p i ,p i+1 ,(x1,y1))*cross(p i ,p i+1 ,(x2,y2))<0

[0025] cross((x1,y1),(x2,y2),p i )*cross((x1,y1),(x2,y2),p i+1 )<0.

[0026] Further improvement is: S12 is specifically: the intersection of the line between the sites and the control area is divided into two cases. When both sites are outside the control area, the route is corrected and the improved A* algorithm is used to avoid obstacles. The improvement is as follows:

[0027] The A* algorithm heuristic function contains both the starting point and the end point information, and its heuristic function is:

[0028] f(n)=g(n)+h(n)

[0029] In the formula, n is the current grid number, g(n) is the distance from the current grid n to the starting point, and h(n) is the distance from the current grid n to the target grid. The above formula is introduced into plane geometric obstacle avoidance, and the tangent of the circle and the corner of the convex polygon is used as a feasible point to replace the grid point. The Euclidean distance is used to calculate the distance between the g value and the h value. In two-dimensional space, given the coordinates of two points (x1, y1) and (x2, y2), the straight-line distance between the two points is calculated based on the Pythagorean theorem. The Euclidean distance is calculated using the following formula:

[0030] d=((x2-x1) 2 +(y2-y1) 2 ) 1 / 2

[0031] When two sites are located inside or outside the control area, the route is saved and marked.

[0032] A further improvement is that: S2 comprises the following steps:

[0033] S21: replacing the routes that need to be corrected and cancelled in the distance matrix;

[0034] S22: Establish a VRPPD model to calculate a customized bus route.

[0035] The further improvement lies in that: in S21, before establishing the VRPPD model, the distances between each pair of stations are calculated and the results are stored in the distance matrix. Through S12, obstacle avoidance and marking are performed on some routes, and the data in the distance matrix is updated, where the route lengths and paths of the routes with obstacle avoidance are updated; in S22, a customized bus route optimization model is established, geographical location restrictions are added to the customized bus delivery problem, the driving routes that do not meet the control requirements are cancelled, the bus routes in the controlled area are re-planned using the vehicles designated for the controlled area, and this problem is transformed into the Pickup and Delivery Vehicle Routing Problem with Area Restriction, where the locations and scopes of the controlled areas are defined independently to obtain the customized bus driving route network.

[0036] The further improvement lies in that: S3 includes the following steps:

[0037] S31: Add the locations and destinations of passengers to the customized bus driving route network;

[0038] S32: Establish a K-shortest-path model to generate k travel plans;

[0039] S33: Establish a logit model to calculate the probabilities of passengers choosing each plan.

[0040] The further improvement lies in that: in S32, according to the new road network formed in S31, with the passenger's location as the starting point and the passenger's destination as the ending point, k shortest travel plans are recommended for each passenger through the K-shortest-path algorithm. The specific steps are as follows:

[0041] Initialization:

[0042] Use the Dijkstra algorithm or A* algorithm, etc. to find the shortest path from the source point s to the target point t, denoted as P1;

[0043] Initialize a set R to store the paths that have been found;

[0044] Initialize a set S to store candidate paths (initially empty);

[0045] Initialize a set U that contains all vertices (except the source point and the target point) to store the deviation points that may generate new paths;

[0046] Let k = 1 (the current number of paths found);

[0047] Search for new paths:

[0048] When k < K, perform the following steps:

[0049] Select a deviation point u from the set U so that the candidate path generated by u is the best among the found paths;

[0050] Use Dijkstra algorithm to find the shortest path from source point s to deviation point u, denoted as P su ;

[0051] Starting from the deviation point u, along the current shortest path P k (the part from u to t, but not including the part before u) in the reverse direction, find a nearest point v, that is, v is P k A point after u, such that the distance from u to v plus the distance from v (excluding v) to t is minimized;

[0052] Construct a new path from P su to u, then from u to v (but not including v), and finally from v along P k The remaining part of the path to t, this new path is recorded as P′;

[0053] When P′ is different from any path in set R (i.e., not a duplicate path), add it to R and remove all points on P′ from set U;

[0054] Update k = k + 1 and repeat the above steps until K paths are found or no new paths can be found;

[0055] Output:

[0056] Output the K shortest paths stored in the set R (when K are found).

[0057] Further improvement is: S33 is specifically: through S32, each passenger has a corresponding recommended travel plan, each passenger is analyzed, and a logit model is established to calculate the probability of the passenger choosing each travel plan. The calculation steps are as follows:

[0058] Calculate the negative travel effects in each plan, and calculate the probability of plan selection based on the negative effects. The formula is as follows:

[0059]

[0060] where p i is the probability of choosing this option, is the cost of taking the customized bus in this plan, is the distance to be walked in this scheme, α, β, γ, θ, C α ,C β are all constants, (x o ,y o ) is the passenger's location, (x d ,yd ) is the passenger’s destination, is the boarding point of plan i, is the alighting point of plan i.

[0061] Further improvement is: in S4, the number of customized bus departures is calculated based on the passenger selection probability, based on the probability of passengers choosing each plan calculated in S3, the probability of passengers choosing the plan is used as the updated demand, the boarding and alighting points of the plan are used as new OD pairs, and the maximum load capacity of the bus during driving is recalculated according to the new demand to obtain the minimum number of departures for the route. The calculation formula is as follows:

[0062]

[0063] Where N k is the number of departures on route k, Q is the maximum vehicle load, is the passenger capacity of route k at station i, K is the set of customized bus routes, and C is the set of bus stations.

[0064] The beneficial effects of the present invention are:

[0065] 1. The present invention studies the impact of regional control policies on customized bus route design and the location selection of boarding and alighting stations, constructs a customized bus route optimization model and a passenger station location selection model for regional control, uses the customized bus route optimization model for regional control to calculate reasonable customized bus routes, and uses the passenger station location selection model to calculate the probability of passengers boarding and alighting at each station, and develops a more accurate calculation model for customized bus operators affected by regional control policies in emergency events, thereby improving the interests of customized bus operators and protecting the safety of passengers and drivers. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 is a flow chart of the present invention;

[0067] Figure 2 It is a schematic diagram of the traditional A* algorithm of the present invention;

[0068] Figure 3 This is a rendering of the improved A* algorithm of the present invention;

[0069] Figure 4 A route diagram of a customized bus route network with respect to stations of the present invention;

[0070] Figure 5 A schematic diagram of adding the location and destination of passengers to the road network of the present invention;

[0071] Figure 6 This is a schematic diagram of the final road network of the present invention. DETAILED DESCRIPTION

[0072] To deepen the understanding of the present invention, the present invention will be further described in detail below in conjunction with embodiments. These embodiments are only used to explain the present invention and do not constitute a limitation on the protection scope of the present invention.

[0073] Embodiment 1

[0074] According to Figure 1 、 2 As shown in 3, 4, 5, and 6, this embodiment proposes a customized bus route design method for area control, including the following steps:

[0075] S1: Calculate whether all relevant routes violate the area control policy, and establish an improved A* algorithm to correct the violated routes; the S1 includes the following steps: S11: Calculate whether the connection line of bus stops passes through the controlled area; S12: Correct or cancel the route of the line passing through the controlled area.

[0076] Specifically, S11 is as follows: Set the controlled area as a circle or a convex polygon. According to the geometric relationship, verify whether the connection line between station A(x1, y1) and station B(x2, y2) intersects the controlled area. The calculation method is as follows:

[0077] When the controlled area is a circle (the center is (x0, y0) and the radius is r):

[0078]

[0079] When d < r, it means that the AB connection line intersects the controlled area;

[0080] When the controlled area is a convex polygon (each side is (p i , p i+1 )):

[0081] Calculate the cross product:

[0082] cross(A, B, C) = (B x - A x )(C y - A y ) - (B y - A y )(C x - A x )

[0083] For each side (p i , p i+1 ) of the polygon, the condition for judging that the line segment (x1, y1) and (x2, y2) intersect is:

[0084] cross(p i , p i+1 , (x1, y1)) * cross(pi ,p i+1 ,(x2,y2))<0

[0085] cross((x1,y1),(x2,y2),p i )*cross((x1,y1),(x2,y2),p i+1 )<0.

[0086] Specifically, S12 is as follows: the intersection of the line between the sites and the control area is divided into two cases. When both sites are outside the control area, the route is corrected and the improved A* algorithm is used to avoid obstacles. The improvement is as follows:

[0087] The A* algorithm heuristic function contains both the starting point and the end point information, and its heuristic function is:

[0088] f(n)=g(n)+h(n)

[0089] In the formula, n is the current grid number, g(n) is the distance from the current grid n to the starting point, and h(n) is the distance from the current grid n to the target grid. The above formula is introduced into plane geometric obstacle avoidance, and the tangent of the circle and the corner of the convex polygon is used as a feasible point to replace the grid point. The Euclidean distance is used to calculate the distance between the g value and the h value. In two-dimensional space, given the coordinates of two points (x1, y1) and (x2, y2), the straight-line distance between the two points is calculated based on the Pythagorean theorem. The Euclidean distance is calculated using the following formula:

[0090] d=((x2-x1) 2 +(y2-y1) 2 ) 1 / 2

[0091] When two sites are located inside or outside the control area, the route is saved and marked.

[0092] S2: Establish a customized bus path optimization model based on the corrected route and calculate the customized bus route network; S2 includes the following steps: S21: Replace the routes that need to be corrected and cancelled in the distance matrix; S22: Establish a VRPPD model to calculate the customized bus route.

[0093] In S21, before establishing the VRPPD model, calculate the distances between each pair of stations and store the results in a distance matrix. Through S12, avoid obstacles and mark some routes, and update the data in the distance matrix, where the lengths and paths of the routes with obstacle avoidance are updated; in S22, establish an optimized model for customized bus routes, add geographical location restrictions to the customized bus delivery problem, cancel the driving routes that do not meet the control requirements, re-plan the bus routes in the controlled areas with the specified vehicles in the controlled areas, and transform this problem into a Pickup and Delivery Vehicle Routing Problem with Area Restriction, where the locations and scopes of the controlled areas are defined independently to obtain the customized bus driving route network.

[0094] S3: Based on the customized bus driving route, establish a passenger travel choice model to calculate the probabilities of passengers choosing each travel plan; S3 includes the following steps: S31: Add the locations and destinations of passengers to the customized bus driving route network; S32: Establish a K-shortest-path model to generate k travel plans; S33: Establish a logit model to calculate the probabilities of passengers choosing each plan.

[0095] In S32, according to the new road network formed in S31, with the passenger's location as the starting point and the passenger's destination as the ending point, recommend k shortest travel plans for each passenger through the K-shortest-path algorithm. The specific steps are as follows:

[0096] Initialization: Use algorithms such as Dijkstra's algorithm or A* algorithm to find the shortest path from the source point s to the target point t, denoted as P1;

[0097] Initialize a set R to store the paths that have been found;

[0098] Initialize a set S to store candidate paths (initially empty);

[0099] Initialize a set U that contains all vertices (except the source point and the target point) to store the deviation points that may generate new paths;

[0100] Let k = 1 (the current number of paths found);

[0101] Search for new paths: When k < K, perform the following steps:

[0102] Select a deviation point u from the set U such that the candidate path generated through u is the optimal among the paths that have been found;

[0103] Use Dijkstra's algorithm to find the shortest path from the source point s to the deviation point u, denoted as Psu ;

[0104] Starting from the deviation point u, along the current shortest path P k (the part from u to t, but not including the part before u) in the reverse direction, find a nearest point v, that is, v is P k A point after u, such that the distance from u to v plus the distance from v (excluding v) to t is minimized;

[0105] Construct a new path from P su to u, then from u to v (but not including v), and finally from v along P k The remaining part of the path to t, this new path is recorded as P′;

[0106] When P′ is different from any path in set R (i.e., not a duplicate path), add it to R and remove all points on P′ from set U;

[0107] Update k = k + 1 and repeat the above steps until K paths are found or no new paths can be found;

[0108] Output result: Output the K shortest paths stored in the set R (when K are found).

[0109] S33 is specifically as follows: through S32, each passenger has a corresponding recommended travel plan, each passenger is analyzed, and a logit model is established to calculate the probability of the passenger choosing each travel plan. The calculation steps are as follows:

[0110] Calculate the negative travel effects in each plan, and calculate the probability of plan selection based on the negative effects. The formula is as follows:

[0111]

[0112]

[0113] where p i is the probability of choosing this option, is the cost of taking the customized bus in this plan, is the distance to be walked in this scheme, α, β, γ, θ, C α ,C β are all constants, (x o ,y o ) is the passenger's location, (x d ,y d ) is the passenger’s destination, is the boarding point of plan i, is the alighting point of plan i.

[0114] S4: Based on the passenger selection probability, calculate the number of customized bus departures, and obtain the customized bus travel routes, number, and travel selection probability of each passenger. In S4, the number of customized bus departures is calculated based on the passenger selection probability, based on the probability of passengers choosing each plan calculated in S3, the probability of passengers choosing the plan is used as the updated demand, and the boarding and alighting points of the plan are used as the new OD pair. The maximum load during the bus travel is recalculated according to the new demand, and the minimum number of departures for the route is obtained. The calculation formula is as follows:

[0115]

[0116] Where N k is the number of departures on route k, Q is the maximum vehicle load, is the passenger capacity of route k at station i, K is the set of customized bus routes, and C is the set of bus stations.

[0117] Embodiment 2

[0118] according to Figure 1 , 2 As shown in , 3, 4, 5, and 6, this embodiment proposes a customized bus route design method for regional control, including the following steps:

[0119] S1: Calculate whether all relevant routes will violate the regional control policy, and establish an improved A* algorithm to correct the illegal routes;

[0120] S2: Establish a customized bus route optimization model based on the revised route and calculate the customized bus route;

[0121] S3: Based on the customized bus routes in S2, a passenger travel choice model is established to calculate the probability of passengers choosing each travel plan;

[0122] S4: Based on the passenger selection probability in S3, the number of customized buses is calculated, and finally the customized bus travel routes and quantity as well as the travel selection probability of each passenger are obtained.

[0123] The S1 comprises the following steps:

[0124] S11: Calculate whether the bus stop connection line will cross the control area;

[0125] S12: Modify or cancel the route that crosses the controlled area;

[0126] S11 specifically includes: setting the control area to be a circle or a convex polygon, and verifying whether the line connecting site A (x1, y1) and site B (x2, y2) intersects with the control area based on the geometric relationship. The calculation method is as follows:

[0127] If the controlled area is circular (with the center at (x0, y0) and the radius r):

[0128]

[0129] If d < r, it indicates that the line segment AB intersects the controlled area

[0130] If the controlled area is a convex polygon (each side is (p i , p i+1 )):

[0131] Calculate the cross product:

[0132] cross(A, B, C) = (B x - A x )(C y - A y ) - (B y - A y )(C x - A x )

[0133] For each side (p i , p i+1 ) of the polygon, the condition for determining whether the line segment (x1, y1) and (x2, y2) intersect is:

[0134] cross(p i , p i+1 , (x1, y1)) * cross(p i , p i+1 , (x2, y2)) < 0

[0135] cross((x1, y1), (x2, y2), p i ) * cross((x1, y1), (x2, y2), p i+1 ) < 0

[0136] The specific content of S12 includes:

[0137] The intersection of the connection line between stations and the controlled area is divided into two cases. If both stations are outside the controlled area, the route is corrected and the improved A* algorithm is used for obstacle avoidance processing. The improvement principle is as follows:

[0138] The heuristic function of the traditional A* algorithm contains information about both the starting point and the ending point, and its heuristic function is:

[0139] f(n) = g(n) + h(n)

[0140] Where n is the current grid number, g(n) is the distance from the current grid n to the starting point, and h(n) is the distance from the current grid n to the target grid. The idea of ​​the above formula is introduced into the plane geometry obstacle avoidance problem. Unlike the "gridding" processing of the traditional A* algorithm, this paper uses the corners of convex polygons and the tangents of circles and corners of convex polygons as feasible points to replace grid points. Since the A* algorithm of this model does not use gridding processing, points can travel in any direction, so the distances calculated for the g and h values ​​above use Euclidean distance. Euclidean distance is a formula for calculating the straight-line distance between two points. In two-dimensional space, given the coordinates of two points (x1, y1) and (x2, y2), this formula is based on the Pythagorean theorem and can be used to calculate the straight-line distance between two points. The Euclidean distance can be calculated using the following formula (14):

[0141] d=((x2-x1) 2 +(y2-y1) 2 ) 1 / 2

[0142] Traditional A* algorithm such as Figure 2 As shown; the effect diagram of the improved A* algorithm is as follows Figure 3 As shown; if the two sites are located inside or outside the control area, the route is saved and marked.

[0143] The S2 comprises the following steps:

[0144] S21: Replace the routes that need to be corrected and cancelled in the distance matrix:

[0145] S22: Establish VRPPD model to calculate customized bus routes

[0146] The S21 specifically includes: before establishing the VRPPD model, the distance between each site is calculated and the result is kept in the distance matrix. Through S1, obstacle avoidance and marking are performed on some routes, and the data of the distance matrix is ​​updated, wherein the route for obstacle avoidance will update the route length and path, and the route for marking will not update the length and path, but the cost will be greatly increased in subsequent calculations.

[0147] S22 includes building a customized bus route optimization model

[0148] Research the customized bus model, add geographical location restrictions to the customized bus distribution problem, cancel the driving routes that do not comply with the lockdown policy, re-plan the bus routes in the lockdown area with designated vehicles in the lockdown area, and transform the problem into a Pickup and Delivery Vehicle Routing Problem with Area Restriction, in which the location and scope of the lockdown area will be defined independently.

[0149] In this paper, the model formulation is developed under the following assumptions:

[0150] (1) All customized buses are dispatched from the same bus terminal;

[0151] (2) The time it takes for passengers to get on and off the bus at the station is negligible;

[0152] (3) There are enough customized buses at the bus terminal;

[0153] (4) the location of bus stops, where passenger demand is collected in advance and does not change;

[0154] (5) Passenger demand cannot be divided into batches;

[0155] (6) Each customized bus travels at a uniform and identical speed;

[0156] (7) The bus terminal cannot be located in the lockdown area;

[0157] (8) Customized public transportation to enter the closed area needs to bear costs;

[0158] Mathematical model

[0159] Sets and indexes

[0160]

[0161] Decision variables

[0162]

[0163]

[0164] Parameters

[0165] symbol Symbols Q Maximum load capacity of custom buses <![CDATA[e ij ]]> Passenger demand from station i to station j P Unit distance operating cost of a single customized bus <![CDATA[d ij ]]> The distance from site i to site j <![CDATA[P g ]]> Fixed costs of a single custom bus <![CDATA[P s ]]> Fine for a single customized bus entering a closed area L Maximum path length for each line F fare

[0166] Functions

[0167]

[0168] Through the above analysis, a distribution route optimization model is constructed based on customized bus revenue as the goal:

[0169]

[0170] st

[0171]

[0172] N k ≥0,Nk =int, (9)

[0173] Formula (1) represents the objective function of the model. The model objective is to maximize revenue. The total revenue is equal to the passenger revenue minus the distance operation cost, the penalty for entering and leaving the closed area, and the fixed cost of customized bus. k The calculation formula is shown in formula (10); formula (2) indicates that each bus stop must ensure flow balance, that is, the bus must depart from a bus stop to arrive at the bus stop; formula (3) indicates that the vehicle departs from the bus terminal and eventually returns to the bus terminal; formula (4) indicates that the length of each route cannot exceed the maximum path length L; formula (5) is the passenger capacity calculation formula for line k at station i; formula (6) indicates that when there are too many passengers, it can be handled by adjusting the number of vehicles.

[0174]

[0175] The S3 comprises the following steps:

[0176] S31: Add the passenger’s location and destination to the customized bus network

[0177] S32: Establish a K-shortest-path model to generate k travel plans

[0178] S33: Establish a logit model to calculate the probability of passengers choosing each option

[0179] The S31 specifically includes: the customized bus network produced by S2 is about the route of the station, such as Figure 4 As shown in the figure, the location and destination of the passengers are not included, so the location and destination of the passengers need to be added to the road network. Usually, passengers are free to choose any station to get on and off the bus, such as Figure 5 However, in order to improve the calculation efficiency, this design removes the routes with too long walking distance. The final road network is as follows Figure 6 shown.

[0180] S32 specifically includes: according to the new road network formed in S32, taking the passenger's location as the starting point and the passenger's destination as the end point, recommending k shortest travel plans for each passenger through the K-shortest-path algorithm. The specific steps are:

[0181] initialization:

[0182] Use Dijkstra algorithm or A* algorithm to find the shortest path from source point s to target point t, denoted as P1.

[0183] Initialize a set R to store the found paths.

[0184] Initialize a set S to store candidate paths (initially empty).

[0185] Initialize a set U that contains all vertices in the graph (except the source and target vertices), which is used to store deviation points that may generate new paths.

[0186] Let k = 1 (the number of paths found so far).

[0187] Find new paths:

[0188] When k < K, perform the following steps:

[0189] a. Select a deviation point u from the set U such that the candidate path generated through u is optimal among the paths that have been found (this usually involves calculating the distance to reach the target vertex after deviating from the current shortest path through u and choosing the u with the minimum distance).

[0190] b. Use Dijkstra's algorithm to find the shortest path from the source vertex s to the deviation point u, denoted as P su .

[0191] c. Starting from the deviation point u, along the inverse direction of the current shortest path P k (the part from u to t, but not including the part before u), find a nearest point v (i.e., v is a point after u on P k and such that the distance from u to v plus the distance from v (excluding v) to t is minimized).

[0192] d. Construct a new path: from P su to u, then from u to v (but not including v), and finally from v along the remaining part of P k to t. This new path is denoted as P'.

[0193] e. If P' is different from any path in the set R (i.e., it is not a duplicate path), then add it to R and remove all points on P' from the set U (because these points have been used to construct a new path and are less likely to generate a better path).

[0194] f. Update k = k + 1 and repeat steps a - e until K paths are found or no new paths can be found.

[0195] Output the result:

[0196] Output the K shortest paths stored in the set R (if K paths are found).

[0197] The S33 specifically includes: in S32, each passenger has a corresponding recommended travel plan. This step analyzes each passenger and establishes a logit model to calculate the probability of the passenger choosing each travel plan. The calculation steps are as follows:

[0198] Calculate the negative travel effects in each plan, and calculate the probability of plan selection based on the negative effects. The formula is as follows:

[0199]

[0200] in is the cost of taking the customized bus in this plan, is the distance to be walked in this scheme, α, β, γ, θ, C α ,C β are all constants, (x o ,y o ) is the passenger's location, (x d ,y d ) is the passenger’s destination, is the boarding point of plan i, is the alighting point of plan i.

[0201] The S4 comprises the following steps:

[0202] The number of customized bus departures is calculated based on the passenger selection probability. The probability of the passenger choosing each plan is calculated in S3. In this step, the probability of the passenger choosing the plan is used as the updated demand, and the boarding and alighting points of the plan are used as the new OD pair. For example, the probability of passenger A choosing to go from his location to station 2, get off at station 8 and go to the destination is 0.6, that is, the new demand is that 0.6 people want to go from station 2 to station 8 during this time period, and the OD pair is changed from the passenger's location to the destination to station 2 to station 8. Recalculate the maximum load during the bus driving process according to the new demand, and get the minimum number of departures for this line. The calculation formula is as follows:

[0203]

[0204] Where N k is the number of departures on route k, Q is the maximum vehicle load, is the passenger capacity of route k at station i, K is the set of customized bus routes, and C is the set of bus stations.

[0205] The customized bus route design method for regional control studies the impact of regional control policies on customized bus route design and passenger station selection and boarding and alighting, constructs a regionally controlled customized bus route optimization model and a passenger station selection model, uses the regionally controlled customized bus route optimization model to calculate reasonable customized bus routes, and uses the passenger station selection model to calculate the probability of passengers boarding and alighting at each station, develops a more accurate calculation model for customized bus operators affected by regional control policies in emergency events, improves the interests of customized bus operators and protects the safety of passengers and drivers.

[0206] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for designing customized bus routes for regional control, characterized in that: It includes the following steps: S1: Calculate whether all relevant routes will violate the regional control policy, and establish an improved A* algorithm to correct the违规 routes; S2: Based on the corrected routes, establish a customized bus path optimization model to calculate the customized bus driving route network; S3: Based on the customized bus driving route, establish a passenger travel choice model to calculate the probability of passengers choosing each travel plan; S4: Based on the passenger selection probability, calculate the number of customized buses dispatched, and obtain the customized bus travel route, quantity, and the probability of each passenger's travel choice.

2. The method for designing a customized bus route for regional control according to claim 1, characterized in that: The S1 includes the following steps: S11: Calculate whether the connection line between bus stops will cross the control area; S12: Correct or cancel the route that crosses the control area.

3. The method for designing a customized bus route for regional control according to claim 2, characterized in that: The S11 is specifically as follows: Set the control area as a circle or a convex polygon. According to the geometric relationship, verify whether the connection line between station A(x1, y1) and station B(x2, y2) intersects with the control area. The calculation method is as follows: When the control area is a circle (the center is (x0, y0) and the radius is r): When d < r, it means that the AB connection line intersects with the control area; When the control area is a convex polygon (each edge is (p i ,p i+1 )): Calculate the cross product: cross(A,B,C)=(B x -A x )(C y -A y )-(B y -A y )(C x -A x ) For each edge of the polygon (p i , p i+1 ) The condition for determining the intersection of line segments (x1, y1) and (x2, y2) is: cross(p i ,p i+1 (x1,y1))*cross(p i ,p i+1 (x2,y2))<0 Cross((x1,y1),(x2,y2),p i )*cross((x1,y1),(x2,y2),p i+1 )<0。 4. The method for designing a customized bus route for regional control according to claim 3, characterized in that: The S12 is specifically as follows: Divide the intersection of the connection line between stations and the control area into two cases. When both stations are outside the control area, correct the route and use the improved A* algorithm for obstacle avoidance. The improvement is as follows: The A* algorithm heuristic function contains the information of both the starting point and the ending point at the same time. Its heuristic function is: f(n) = g(n) + h(n) In the formula, n is the current grid number, g(n) is the distance from the current grid n to the starting point, h(n) is the distance from the current grid n to the target grid. Introduce the above formula into the plane geometry obstacle avoidance. Use the tangent of the circle and the convex polygon angle as the feasible points to replace the grid points, and calculate the g value and the h value distance using the Euclidean distance. In the two-dimensional space, given the coordinates (x1, y1) and (x2, y2) of two points, based on the Pythagorean theorem, calculate the straight-line distance between the two points. The Euclidean distance is calculated by the following formula: d=((x2-x1) 2 +(y2-y1) 2 ) 1 / 2 When the two stations are distributed inside and outside the control area, save the route and mark it.

5. The method for designing customized bus routes for regional control according to claim 4, characterized in that: The S2 includes the following steps: S21: Replace the routes that need to be corrected and cancelled in the distance matrix; S22: Establish a VRPPD model to calculate the customized bus driving route.

6. A method for designing customized bus routes for regional control according to claim 5, characterized in that: In the S21, before establishing the VRPPD model, calculate the distance between each station and keep the result in the distance matrix. Through S12, perform obstacle avoidance and marking on some routes, and update the data in the distance matrix. The routes that perform obstacle avoidance update the route length and path. In the S22, establish a customized bus path optimization model, add geographical location restrictions to the customized bus distribution problem, cancel the driving routes that do not meet the control requirements, re-plan the bus routes in the control area with the specified vehicles in the control area, and transform the problem into the Pickup and Delivery Vehicle Routing Problem with Area Restriction, where the locations and scopes of the control areas are defined independently to obtain the customized bus driving route network.

7. The method for designing a customized bus route for regional control according to claim 1, characterized in that: The said S3 includes the following steps: S31: Add the location and destination of the passenger to the customized bus route network; S32: Establish a K - shortest - path model to generate k travel plans; S33: Establish a logit model to calculate the probability of the passenger choosing each plan.

8. A method for designing customized bus routes for regional control according to claim 7, characterized in that: In the said S32, according to the new road network formed in S31, with the passenger's location as the starting point and the passenger's destination as the ending point, use the K - shortest - path algorithm to recommend k shortest travel plans for each passenger. The specific steps are as follows: Initialization: Use Dijkstra's algorithm or A* algorithm, etc. to find the shortest path from the source point s to the target point t, denoted as P1; Initialize a set R for storing the paths that have been found; Initialize a set S for storing candidate paths (initially empty); Initialize a set U containing all vertices (except the source point and the target point) for storing the deviation points that may generate new paths; Let k = 1 (the number of paths found currently); Search for new paths: When k < K, execute the following steps: Select a deviation point u from the set U such that the candidate path generated through u is the optimal among the paths that have been found; Use Dijkstra algorithm to find the shortest path from source point s to deviation point u, denoted as P su ; Starting from the deviation point u, along the current shortest path P k (the part from u to t, but not including the part before u) in the reverse direction, find a nearest point v, that is, v is P k A point after u, such that the distance from u to v plus the distance from v (excluding v) to t is minimized; Construct a new path from P su to u, then from u to v (but not including v), and finally from v along P k The remaining part of the path to t, this new path is recorded as P ′ ; When P ′ If it is different from any path in set R (i.e., not a duplicate path), then add it to R and remove all paths in P from set U. ′ The point on; Update k = k + 1, and repeat the above steps until K paths are found or no new paths can be found; Output results: Output the K shortest paths stored in the set R (when K paths are found).

9. A method for designing customized bus routes for regional control according to claim 8, characterized in that: The said S33 is specifically as follows: Through S32, each passenger has a corresponding recommended travel plan. Analyze each passenger and establish a logit model to calculate the probability of the passenger choosing each travel plan. The calculation steps are as follows: Calculate the negative travel effect in each plan, and calculate the plan selection probability according to the negative effect. The formula is as follows: where p i is the probability of choosing this option, is the cost of taking the customized bus in this plan, is the distance to be walked in this scheme, α, β, γ, θ, C α ,C β are all constants, (x o ,y o ) is the passenger's location, (x d ,y d ) is the passenger’s destination, is the boarding point of plan i, is the alighting point of plan i.

10. A method for designing customized bus routes for regional control according to claim 9, characterized in that: In the said S4, calculate the departure quantity of the customized bus based on the passenger selection probability. Based on the probability of each passenger choosing each plan calculated in S3, take the probability of the passenger choosing the plan as the updated demand, and the boarding and alighting points of the plan as the new OD pair. Recalculate the maximum loading capacity during the bus travel according to the new demand to obtain the minimum departure quantity of this line. The calculation formula is as follows: Where N k is the number of departures on route k, Q is the maximum vehicle load, is the passenger capacity of route k at station i, K is the set of customized bus routes, and C is the set of bus stations.

Images