Multi-objective path planning method for intelligent public transportation system

Abstract

This invention discloses a multi-objective route planning method for an intelligent public transportation system. The method includes: generating a shortest path network topology connecting various stations based on analysis of the transportation network topology and available bus routes; dynamically and evenly allocating bus routes based on passenger waiting times and road lengths, optimally assigning buses to different trips; and determining the optimal bus route configuration by comprehensively considering average passenger travel (occupancy) distance and passenger experience. This method minimizes passenger waiting times while ensuring high passenger load factors, thereby improving the overall efficiency of the bus system and the passenger travel experience.

Description

Technical Field

[0001] The present invention belongs to the technical field of path planning, and in particular relates to a multi-objective path planning method for an intelligent public transportation system. Background Art

[0002] With the rapid development of intelligent vehicles and the connected vehicle industry, passenger demand for a more comfortable and convenient travel experience has continued to rise in recent years. Providing intelligent transportation services that are more comfortable and convenient has become increasingly important. To achieve this goal, efficiently and scientifically planning routes that meet these needs is crucial. However, the various constraints of the real-world traffic environment complicate vehicle route planning. Traditional route planning methods rely on historical experience and follow a greedy approach. However, for bus operators, reducing operating costs while improving the passenger experience is crucial. Therefore, a multi-objective route planning method for intelligent bus systems is urgently needed. Summary of the Invention

[0003] To solve the above technical problems, the present invention proposes a multi-objective path planning method for an intelligent public transportation system, which can minimize passengers' waiting time while ensuring a high passenger load factor, thereby improving the overall efficiency of the bus system and the travel experience of passengers.

[0004] To achieve the above objectives, the present invention provides a multi-objective path planning method for an intelligent public transportation system, comprising:

[0005] Generate the shortest path network topology connecting each station through analysis based on traffic network topology and bus drivable paths;

[0006] Dynamically and evenly distribute bus routes based on passenger waiting time and road length, allocating buses to different trips in an optimal manner;

[0007] The optimal bus route configuration is determined by comprehensively considering the average passenger travel (occupancy) distance and passenger experience.

[0008] Optionally, given a specific trip s, the stations are gradually added in the order of traversal, and the total path length τ(s) of the specific trip s is determined by:

[0009]

[0010]

[0011] Where i and j represent adjacent bus stops, and d(i, j) represents the distance between bus stops i and j.

[0012] Optionally, the total number of passengers in a trip s, denoted as P(s), is obtained by summing the number of passengers at each stop:

[0013] Among them, C(b) i，j (s) represents the number of passengers that bus b carries when it stops at station i.

[0014] Optionally, the change in the number of passengers on a bus can be determined based on the number of passengers at each bus stop. The update process includes:

[0015]

[0016] Among them, C(b) i，j (s) represents the number of passengers when bus b first arrives at station i, B(b) i，j (s) represents the number of passengers boarding at station i, A(b) i (s) represents the passengers who get off at station i.

[0017] Optionally, get the average passenger waiting time including:

[0018] First, the waiting time is divided into two time periods, and the weighted average waiting time of each time period is calculated; the first item is the waiting time WT in the normal time period r , each WT r Multiply by the corresponding weight ω r , then sum the values ​​of i from 1 to k1; the second term represents the waiting time WT in a specific time period s , each WT s Multiply by the weight ω s , then sum j from 1 to k2; finally, add the weighted average waiting time of these two time periods and divide it by k1+k2 to get the average waiting time WT of the passenger avg .

[0019] Optionally, obtaining the congestion degree includes:

[0020]

[0021]

[0022] in,

[0023] C d (s) is the congestion degree, C t (s) is the threshold (set according to the ride experience the operator wants to provide, usually set to 0.9).

[0024] C(b) i，j is the number of passengers when bus b first arrives at station i in trip s with end point j; s is an element of the set of trips S; β(b) is the maximum capacity of bus b.

[0025] Optional, evenly distributed bus routes include:

[0026] Step 1: Select the one with the minimum cost value The bus has the highest priority in selecting the stop;

[0027] Step 2: Bus b with the highest selection weight p Select the site with the smallest site evaluation index λ from the neighboring sites and add it to the set;

[0028] Step 3: Update the cost value of the bus

[0029] Step 4: When all stops have been allocated, if there are still trips that have not yet reached the destination, directly add the destination stop to these trips and go to step 5; otherwise, return to step 1;

[0030] Step 5 The initial value starts from 0 and increases by 0.02 each time until it reaches 1; at the same time, The initial value of starts from 1 and decreases by 0.02 each time until it reaches 0; finally, the optimal result is used as the final path planning result, that is, the entire itinerary allocation procedure is completed; if Not yet reached 1, If it has not reached 0, record the current optimal itinerary result and return to step 1; otherwise, select the optimal itinerary from all recorded results as the output.

[0031] Optionally, determining the optimal bus route configuration includes:

[0032] Step 1: When a station suddenly receives a passenger request Q, the system will determine whether the station has been traversed. If not, the system will exit the dynamic routing strategy and assign the route to the itinerary allocation program. If the station has been traversed, the system will initialize s = 0 and Tag = 0, where the variable s is the itinerary and the variable Tag indicates whether the request has been assigned to a certain itinerary. If the request has been assigned to a certain itinerary, Tag = 1; otherwise, it is 0 and the system proceeds to step 2.

[0033] Step 2: Consider all trips and determine whether Q is a ride request. If it is a ride request, proceed to step 3. If it is a ride cancellation, delete the request Q from the trip S.

[0034] Step 3: Calculate the average passenger travel (occupancy) distance and ride experience from the bus stop to the requested stop, and determine whether the constraints are met. If so, record the result and set the tag to 1, indicating that the request Q meets the conditions for being included in trip s; otherwise, do not record the result and proceed to step 4.

[0035] Step 4: After determining whether all trips have been considered, if not, return to step 2. If so, continue to determine whether the Tag is 1. If the Tag is 1, the request Q is finally assigned to the optimal trip s; otherwise, the request Q is rejected and additional vehicles must be dispatched.

[0036] Technical effects of the present invention:

[0037] (1) A multi-objective optimization path planning algorithm, BusNav, is designed for flexible bus scenarios. The algorithm can efficiently dispatch buses to meet different operating cost and passenger experience requirements.

[0038] (2) The BusNav algorithm’s route extraction program and trip allocation program work together to dynamically allocate bus routes to meet various constraints. This dynamic scheduling feature allows buses to more flexibly adapt to real-time passenger demand and road conditions.

[0039] (3) Through comparative experiments with traditional manual scheduling methods, the Greedy method and the Rapid Random Search Tree (RRT) method, it was found that the BusNav algorithm performed better in terms of average passenger travel distance and riding experience. This shows that the BusNav algorithm can reduce operating costs and significantly improve the passenger riding experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0041] Figure 1 This is an example of a bus route map in Xiamen according to an embodiment of the present invention.

[0042] Figure 2 Schematic diagram of the path extraction process flow in an embodiment of the present invention.

[0043] Figure 3 Schematic diagram of the itinerary allocation procedure according to an embodiment of the present invention.

[0044] Figure 4(a) is a schematic diagram of the cost value of the first round of path selection in an embodiment of the present invention. Assume that there are three bus starting stations, namely u1, u2 and u3. Each bus starting station is individually assigned a bus b1, b2 and b3. The three buses each depart from the starting station and eventually arrive at the bus terminal T. In addition to the starting station, each transfer station has at least one bus passing through. The initial state B of each bus set is i ={u i}, First, each bus b i Find the node with the smallest station evaluation index λ from the neighboring bus stations and add it to the set B i middle.

[0045] Since the cost of all buses at the beginning of the program is are all 0, so all buses have the right to choose. Currently, b1 chooses λ 1，4 Minimum v4, b2 selects λ 2，5 Minimum v5, b3 selects λ 3，6 The smallest v6. Assuming that their corresponding distances are 5, 4, and 6 respectively, the cost of each bus in the first round is The results are shown in Figure 4(a).

[0046] Figure 4(b) is a schematic diagram of the cost values ​​of the second round of path selection in the embodiment of the present invention. In the second round, since the cost of b2 is the smallest, bus b2 has the priority to be selected. b2 selects the node with the smallest station evaluation index λ among the neighboring nodes and adds it to the set B2, as shown in Figure 4(b). If the station evaluation index λ corresponding to the neighboring nodes v7 and v8 is λ respectively 5，7 =3.2 and λ 3，8 =3.6, then add node v7 to set B2. Assuming the distance from point v5 to v7 is 2.3, the cost of bus b2 is updated to

[0047] Figure 4(c) is a schematic diagram of the cost value of the third round of path selection in an embodiment of the present invention. In the third round, since the cost of b1 is the smallest Therefore, b1 has a high priority, as shown in Figure 4(c). b1 selects the site with the smallest evaluation index λ from the adjacent nodes (λ 4，9 ) is added to the set B1. Assuming that the site evaluation index λ of the neighbor node v9 is 3.1, node v9 is added to the set B1. The cost of updating the vehicle Until all stations are allocated, if there are trips that have not yet reached the terminal, the terminal station will be directly added to the trips that have not yet reached the terminal station; that is, the entire trip allocation process is completed.

[0048] When all the stops have been allocated, if there are still trips that have not yet reached the destination, the destination stop is directly added to these trips, completing one round of trip allocation process.

[0049] Figure 5 Schematic diagram of the dynamic path planning strategy according to an embodiment of the present invention. DETAILED DESCRIPTION

[0050] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0051] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0052] This embodiment provides a multi-objective path planning method for an intelligent public transportation system, including:

[0053] Generate the shortest path network topology connecting each station through analysis based on traffic network topology and bus drivable paths;

[0054] Dynamically and evenly distribute bus routes based on passenger waiting time and road length, allocating buses to different trips in the best way;

[0055] The optimal bus route configuration is determined by comprehensively considering the average passenger travel (occupancy) distance and passenger experience.

[0056] Definition 1: Given a specific trip s, the total path length τ(s) of the specific trip s can be determined by gradually adding the stations in the order of traversal, as shown in Equation (1).

[0057]

[0058]

[0059] Where i and j represent adjacent bus stops, and d(i, j) represents the distance between bus stops i and j. If stop j is the next stop after stop i in the itinerary, then the binary variable The value is 1, otherwise The value is 0.

[0060] Definition 2: The total number of passengers in a trip s, denoted as P(s), is obtained by adding the number of passengers at each stop, as shown in formula (3). i，j (s) represents the number of passengers carried by bus b at stop i;

[0061] Definition 3: The change in the number of passengers on a bus can be determined based on the number of passengers at each bus stop. The update process is shown in formula (4):

[0062]

[0063] Among them, C(b)i，j (s) represents the number of passengers when bus b first arrives at station i, B(b) i，j (s) represents the number of passengers boarding at station i, A(b) i (s) represents the passengers who get off at station i, as shown in formula (4). Formula (4) is restricted by formulas (5), (6) and (7).

[0064]

[0065] In formula (5), β(b) is the total number of passengers that bus b can carry. The total number of passengers on bus b minus the total number of passengers transported between stations i and j must be greater than or equal to the total number of passengers boarding at stations i and j. Formula (6) states that the total number of passengers carried between stations i and j must be less than the number of people waiting to travel from station i to station j, where p i，j (s) is the number of people waiting at station i. Equation (7) shows that the number of passengers who board the bus at station j must be greater than or equal to the total number of passengers who get off the bus at station i.

[0066] Definition 4: Formula (8) defines the average waiting time of passengers as WT avg The weight value ω r Multiply by the regular waiting time, denoted as WT r Similarly, the waiting time in a specific time period is expressed as WT s , multiplied by the weight value ω s The specific time period can be determined based on business needs or through big data analysis. In the present invention, the peak hours are 8:00 a.m. to 9:00 a.m. and 5:00 p.m. to 5:00 p.m. The variables k1 and k2 represent the total number of passengers in the regular time period and the specific time period, respectively.

[0067]

[0068] Average passenger waiting time WT avg The calculation of can be done as follows: First, divide the waiting time into two time periods and calculate the weighted average waiting time of each time period. The first term is the waiting time WT of the normal time period r , each WT r Multiply by the corresponding weight ω r , and then sum it for i from 1 to k1. The second term represents the waiting time WT in a specific time period s , each WT s Multiply by the weight ω s , then sum the j from 1 to k2. Finally, add the weighted average waiting time of the two time periods and divide it by k1+k2 to get the average waiting time WT of the passenger avg .

[0069] Definition 5: Crowding is defined based on the number of passengers on a bus and the vehicle's carrying capacity. Crowding affects the passenger's riding experience. A less crowded riding environment can provide a better riding experience. In addition, the present invention uses the crowding degree C d (s) is defined as the number of passengers C(b) i，j (s) divided by the maximum passenger capacity of the bus β(b), where C t (s) is the threshold value, which is set to avoid seat congestion, as shown in equations (9) and (10).

[0070]

[0071]

[0072] 2. Define the objective function

[0073] The first objective of the present invention is to determine the bus route with the shortest path length among the trips in the set S, where D(s) represents the trip with the shortest distance in the set S. Equations (11) and (12) ensure the optimization of passenger travel time, because the path length directly affects the time it takes for passengers to travel.

[0074]

[0075]

[0076] In formula (12), Q + Refers to a correct rational number. According to equations (11) and (12), while minimizing the travel distance, the two objectives are combined into one objective, called the average passenger travel (occupancy) distance function, which is the average distance traveled by each passenger in trip s, denoted as τp(s), as shown in equation (13). Obviously, when τ(s) reaches its minimum value and P(s) reaches its maximum value, τp(s) also reaches its minimum value, and this optimization criterion achieves the goal of minimizing operating costs.

[0077]

[0078] Next, we define the second objective of the present invention, namely the passenger's riding experience function. Since waiting time and congestion level will affect the passenger's riding experience, in the present invention, it is expected that the passenger's waiting time WT avg and the congestion degree of the riding section C d (s). In order to comprehensively consider the impact of passenger waiting time and congestion level on passenger riding experience, as shown in formula (14).

[0079]

[0080] The calculation method of passenger waiting time is defined by equations (15) and (16). v (s) is the time it takes for the bus to arrive at station v, pt v (s) is the time it takes for a passenger to arrive at station v, and wt(s) represents the passenger's tolerance for waiting time. The waiting time for each passenger does not exceed the specified threshold wt(s), and the operator can dynamically adjust and provide customized services.

[0081] Adjust the defined threshold for different user experiences. For example, a smaller threshold can lead to a better riding experience. Consider the waiting time of all passengers. This constraint brings higher complexity to the problem because pt v (s) and the passenger arrival time distribution are not restricted; that is, the passenger arrival time has an uncertain (random) characteristic.

[0082] bt v (s)-pt v (s)≤wt(s) (15)

[0083]

[0084] In the third objective of the present invention, in order to reduce operating costs, the number of passengers per trip should be maximized to reduce the required trips and number of departures, that is, to increase the passenger capacity of buses on each route. Operating costs can be reduced by reducing the required trips and number of departures, as shown in Equations (17) and (18).

[0085]

[0086]

[0087] At the same time, in order to reduce operating costs, we need to minimize the number of buses required while meeting the travel needs of all passengers. Where nb represents the number of buses, and the minimum value of nb is 1, which means that at least one bus needs to be dispatched, as shown in Equations (19) and (20):

[0088] min(nb), (19)

[0089]

[0090] From the perspective of the trip, each trip has its own optimization objectives, namely τp(s) and be(s). Since all trips have the same optimization objectives, each trip is optimized independently; that is, in each round, a route is selected from the candidate routes based on our optimization objectives τp(s) and be(s). Therefore, the optimization of each trip is equivalent to the optimization of all trips. The multi-objective problem in this invention can be expressed as shown in Equation (21):

[0091]

[0092] 3. Data Structure

[0093] The bus route map can be defined as a directed graph G(V, E), where V is the set of bus stops and transfer points, V = {v1, .., v n}. E is the set of edges, expressed as E = {e(u, v), where u, v∈V∧u≠v}. The set of m bus trips can be expressed as S = {s1, s2, ..., s m}. A set of itineraries k ∈S can be expressed as τ(s)={v i ,...,v j}, where v i is the starting station, v j For the terminal.

[0094] Figure 1 This is a route map from Xiamen Bus Station to the subway station (Xiamen North Station). There are three starting points, one subway station (terminal point), and six general bus stops. The small dots on the right side of the figure are road intersections. The present invention will find three routes that pass through all bus stops in parallel and ultimately reach the subway station (terminal point). Passengers can only board the bus at the six stops or the three starting points. This satisfies the requirements of dispatching the least number of buses, the shortest path, and satisfying the passenger riding experience. To better describe the bus scheduling model of the present invention, the parameters and decision variables used in the optimization model are summarized in Table 1.

[0095] When a map and bus station information are given, it is transformed into an adjacency matrix M∈R n+1×n+1 (R is a real number), the adjacency matrix M records the information of each bus stop and the intersection of each road. The intersection of the road is regarded as a transfer point, where M is defined as the Euclidean distance between any two points. The diagonal distance adjacency matrix M is 0, as shown in Equations (22) and (23). In addition, the adjacency matrix allows roads that are closed to traffic (for example, for regulation, maintenance, etc.) to be set to infinity in advance.

[0096] M=M T (twenty two)

[0097] Diag(M)=0 (23)

[0098] The BusNav algorithm consists of two clearly defined phases. In the first phase, a path extraction procedure is used to calculate the shortest route to each bus stop, effectively building the shortest path connecting all stops together. In the subsequent phase, a trip allocation procedure is used to determine the driving sequence between bus stops, taking into account a series of constraints. The goal is to find a driving route that meets the specified operating cost constraints while ensuring a satisfactory boarding experience for passengers. Table 1 shows the parameters:

[0099] Table 1

[0100]

[0101]

[0102] The first stage of path planning in this invention is to extract the path. In this process, some unnecessary paths can be filtered or removed to reduce the amount of data calculation. To achieve this, the path extraction process uses the Floyd algorithm, which is good at calculating the shortest distance between any two stations. The Floyd algorithm uses dynamic programming techniques and the principle of greedy method to convert dist_var w，v，k is considered as the length of the shortest path from node w to v, containing only nodes from the set (1, ..., k) as transit nodes. In Floyd's algorithm, the decision to update the path length can be divided into the following two cases:

[0103] In formula (24), if a shortest path passes through a transit node k and generates a shorter path length, the path length and path information are updated:

[0104] dist_va w,v，k =dist_var w，k，k-1 +dist_var k，v，k-1 (twenty four)

[0105] Otherwise, the path is not updated and the original path is maintained, as shown in formula (25):

[0106] dist_var w，v，k =dist_var w，v，k-1 (25)

[0107] Therefore, with the above relationship, equations (24) and (25) can be rewritten as equation (26).

[0108]

[0109] In the process of finding the shortest path between any two stations using the Floyd algorithm, some unnecessary paths can be removed or filtered out. There are multiple transfer points between stations. If these unnecessary transfer points can be removed in advance, it will help the calculation speed. The time complexity of the Floyd algorithm is O(n 3 ), so as the number of transfer points increases, the time complexity increases. Therefore, in the first stage, the inapplicable transfer points should be removed to reduce the amount of calculation. Figure 2 Flowchart of the path acquisition process.

[0110] After determining the shortest path between any two stations, it is necessary to plan the driving path trajectory from the starting station to the terminal station. In order to reduce operating costs, minimizing the length of the driving path is our goal. The length of the driving distance and the number of departures have an impact on operating costs. The passenger's riding experience will be affected by waiting time and congestion. In order to assign the optimal driving station and sequence to each bus, we define a station evaluation index (λ), which is the average passenger travel (occupancy) distance multiplied by the weight Then multiply the passenger's riding experience value by the weight Added together, the site evaluation index is the key indicator of whether a site is included in the itinerary s, as shown in formula (27).

[0111]

[0112] The steps of the BusNav algorithm's trip allocation procedure are as follows:

[0113] First, we and (in ) to initialize, The increment is 0.02, The reduction is 0.02, i is the ID of the bus, and each bus B i , 1≤i≤nb, record your own path trajectory, the initial state only records the starting site u, that is, B i ={u i To distribute the work of each bus evenly, a cost value is set for each bus Cost value Record the total distance of a journey, where The initial state is 0. Figure 3 Here are the steps in the itinerary allocation process:

[0114] Step 1: Select the one with the minimum cost value The bus with the highest priority in selecting a station. Since the cost value of each bus is initially 0, if the cost is the same, the bus with the smallest ID is selected with priority to select the station. For each bus, select the node v with the smallest station evaluation index λ from the adjacent stations that have not yet been selected, and assign the node v to the bus set B i Here, α(s) represents the threshold, and the value of τp(s) affects the operating cost of the bus. A larger τp(s) results in higher operating costs for the bus company, while a smaller τp(s) results in lower operating costs. Equation (28) shows how the operating cost is calculated.

[0115]

[0116] Another objective to consider in the station evaluation metric λ is the passenger experience. The selected station should have a passenger experience value be(s) that is greater than or equal to the experience threshold γ(s). In other words, lower congestion or shorter average waiting time corresponds to a better passenger experience, as shown in Equation (29).

[0117]

[0118] With minimum cost value Bus B p Has the highest selection priority. If the costs are the same, the bus with the smallest ID is given a higher selection priority, as shown in formula (30).

[0119]

[0120] Step 2: Bus b with the highest choice p Select the site with the smallest site evaluation index λ from the neighboring sites and add it to the set B p As shown in formula (31).

[0121] B p ={v|min(λ)} (31)

[0122] Step 3: Update the cost value of the bus Where len is the set B p The total number of nodes, expressed as len = |B p |. Set the total distance of the path to bus b p Cost value As shown in formula (32).

[0123]

[0124] Step 4: When all stops have been assigned, if there are still trips that have not yet reached their destination, the destination stop is directly added to these trips, completing one round of trip assignment and proceeding to Step 5. Otherwise, return to Step 1. The advantage of this design is that it avoids the problem of a single bus bearing the entire workload, ensuring fairer work distribution, more efficient resource utilization, and better overall balance.

[0125] Step 5: The initial value of starts from 0 and increases by 0.02 each time until it reaches 1. The initial value of starts from 1 and decreases by 0.02 each time until it reaches 0. Finally, the optimal result is used as the final path planning result, which completes the entire itinerary allocation process. Not yet reached 1, If it has not reached 0, record the current optimal trip result and return to step 1. Otherwise, select the optimal trip from all recorded results as the output.

[0126] An example of a trip allocation program is given to illustrate the process of the trip allocation program. Taking Figure 4(a) as an example, assume that there are three bus starting stations, namely u1, u2 and u3. Each bus starting station is individually assigned a bus b1, b2 and b3. The three buses each depart from the starting station and eventually arrive at the bus terminal T. In addition to the starting station, each transfer station has at least one bus passing through. The initial state of each bus set is B i ={u i}, First, each bus b i Find the node with the smallest station evaluation index λ from the neighboring bus stations and add it to the set B i middle.

[0127] Since the cost of all buses at the beginning of the program is are all 0, so all buses have the right to choose. Currently, b1 chooses λ 1，4 Minimum v4, b2 selects λ 2，5 Minimum v5, b3 selects λ 3，6 The smallest v6. Assuming that their corresponding distances are 5, 4, and 6 respectively, the cost of each bus in the first round is The results are shown in Figure 4(a).

[0128] Figure 4(a) Since the cost of the three buses is 0 at the beginning of the process, each bus has the priority to select a station and add its adjacent stations to its respective set B. i , where the corresponding costs are 5, 4, and 6.

[0129] In the second round, bus b2 has the priority to be selected because its cost is the smallest. b2 selects the node with the smallest station evaluation index λ among the neighboring nodes and adds it to the set B2, as shown in Figure 4(b). If the station evaluation index λ corresponding to the neighboring nodes v7 and v8 is λ 5，7 =3.2 and λ 3，8 =3.6, then add node v7 to set B2. Assuming the distance from point v5 to v7 is 2.3, the cost of bus b2 is updated to

[0130] Figure 4(b). Since bus b2 has the lowest cost, it has priority and is added to set B2, where the corresponding cost is and The costs are 3.3 and 3.5 respectively.

[0131] In the third round, since the cost of b1 is the smallest Therefore, b1 has a high priority, as shown in Figure 4(c). b1 selects the site with the smallest evaluation index λ from the adjacent nodes (λ 4，9 ) is added to the set B1. Assuming that the site evaluation index λ of the neighbor node v9 is 3.1, node v9 is added to the set B1. The cost of updating the vehicle Until all stations are allocated, if there are trips that have not yet reached the terminal, the terminal station will be directly added to the trips that have not yet reached the terminal station; that is, the entire trip allocation process is completed.

[0132] Figure 4(c). Since bus b1 has the lowest cost, it has priority and is added to set B1 with the corresponding cost, where d 1，6 The cost is 3.1.

[0133] However, the above procedure, the itinerary allocation procedure is a one-time allocation process. The present invention takes the map of the route planning as input. Increment from 0 to 1, It decreases from 1 to 0, and finds the optimal τp(s) and be(s) as the final result.

[0134] In reality, when passengers suddenly make or cancel rides, the BusNav system needs to monitor their bookings in real time and make timely routing arrangements. The system obtains information on confirmed bookings, cancellations, and new booking requests through communication with passengers and drivers and real-time updates to the system database.

[0135] To handle these unexpected situations, the BusNav system incorporates a dynamic routing strategy into its trip allocation process. This allows it to flexibly adjust itineraries to meet passenger needs. When a passenger cancels a reservation, the system simply subtracts the number of canceled passengers from the total number of passengers at the corresponding stop. This ensures that the passenger count at each stop is updated promptly during the trip allocation process.

[0136] Step 1. When a station suddenly receives a passenger request Q, the system determines whether the station has been traversed. If not, the system exits the dynamic routing strategy and assigns a route to the itinerary. If the station has been traversed, the system initializes s = 0 and Tag = 0, where the variable s represents the itinerary and the variable Tag indicates whether the request has been assigned to a specific itinerary. If the request has been assigned to a specific itinerary, Tag = 1; otherwise, it is 0. The system then proceeds to Step 2.

[0137] Step 2. Consider all trips and determine whether Q is a ride request. If it is a ride request, proceed to step 3. If it is a ride cancellation, delete the request Q from the trip S.

[0138] Step 3. Calculate the average passenger travel (occupancy) distance and ride experience from the bus stop to the requested stop, and determine whether the constraints are met. If so, record the result and set the tag to 1, indicating that the request Q meets the conditions for being included in trip s; otherwise, do not record it and proceed to step 4.

[0139] Step 4. After determining whether all trips have been considered, if not, return to step 2. If so, continue to determine whether the Tag is 1. If the Tag is 1, the request Q is finally assigned to the optimal trip s. Otherwise, the request Q is rejected. At this time, additional vehicles must be dispatched to meet the needs of passengers. Figure 5 It is a flow chart of dynamic path planning strategy.

[0140] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A multi-objective path planning method for an intelligent public transportation system, characterized in that: include: Generate the shortest path network topology connecting each station through analysis based on traffic network topology and bus drivable paths; Dynamically and evenly distribute bus routes based on passenger waiting time and road length, allocating buses to different trips in an optimal manner; Determine the optimal bus route configuration by comprehensively considering average passenger travel distance and passenger experience; Evenly distributed bus routes include: Step 1: Select the one with the minimum cost value The bus has the highest priority in selecting the stop; Step 2: The bus with the highest selection weight Select the site with the smallest site evaluation index λ from the neighboring sites and add it to the set; Step 3: Update the cost value of the bus ; Step 4: When all stops have been allocated, if there are still trips that have not yet reached the destination, directly add the destination stop to these trips and go to step 5; otherwise, return to step 1; Step 5 The initial value starts from 0 and increases by 0.02 each time until it reaches 1; at the same time, The initial value of starts from 1 and decreases by 0.02 each time until it reaches 0; finally, the optimal result is used as the final path planning result, that is, the entire itinerary allocation procedure is completed; if Not yet reached 1, If it has not reached 0, record the current optimal itinerary result and return to step 1; otherwise, select the optimal itinerary from all recorded results as output; in, ； ； ； i is the ID of the bus; is the total path length of trip s; is the total number of passengers picked up in trip s; represents the threshold; s is a subset of the trip set S; is the average waiting time for passengers; is the congestion degree in trip s; is the experience threshold; The total distance of the driving route is set to bus Cost value ; Determining the optimal bus route configuration involves: Step 1: When a station suddenly receives a passenger request Q, the system will determine whether the station has been traversed. If not, the system will exit the dynamic routing strategy and assign the route to the itinerary allocation program. If the station has been traversed, the system will initialize s = 0 and Tag = 0, where the variable s is the itinerary and the variable Tag indicates whether the request has been assigned to a certain itinerary. If the request has been assigned to a certain itinerary, Tag = 1; otherwise, it is 0 and the system proceeds to step 2. Step 2: Consider all trips and determine whether Q is a ride request. If it is a ride request, proceed to step 3. If it is a ride cancellation, delete the request Q from the trip S. Step 3: Calculate the average passenger travel distance and ride experience from the bus stop to the requested stop, and determine whether the constraints are met. If so, record the result and set the tag to 1, indicating that the request Q meets the conditions for being included in trip s. Otherwise, do not record the result and proceed to step 4. Step 4: After determining whether all trips have been considered, if not, return to step 2. If so, continue to determine whether the Tag is 1. If the Tag is 1, the request Q is finally assigned to the optimal trip s; otherwise, the request Q is rejected and additional vehicles must be dispatched.

2. The multi-objective path planning method for an intelligent public transportation system according to claim 1, wherein: Given a specific itinerary , the stations are gradually added in the order of traversal to determine the specific itinerary The total path length include: （1） in, and Indicates the adjacent bus stop. express and The distance between bus stops; if the stops It is a station on the itinerary The next station, the binary variable The value is 1, otherwise The value is 0; n is the total number of stations; V is the set of bus stations and transfer points, V = {v1, .., v n }.

3. The multi-objective path planning method for an intelligent public transportation system according to claim 2, wherein: In a certain itinerary The total number of passengers in , expressed as , obtained by adding up the number of passengers at each station: in, It represents the number of passengers when bus b first arrives at station i in trip s with the destination j.

4. The multi-objective path planning method for an intelligent public transportation system according to claim 3, wherein: The change in the number of passengers on a bus can be determined based on the number of passengers at each bus stop. The update process includes: in, represents the number of passengers when bus b first arrives at station i in trip s with the destination j; represents the number of passengers who board bus b at stop i in trip s with the destination j; Indicates that it is at the station Passengers getting off the bus.

5. The multi-objective path planning method for an intelligent public transportation system according to claim 4, characterized in that: Obtaining the average passenger waiting time includes: First, the waiting time is divided into two time periods, and the weighted average waiting time of each time period is calculated separately; finally, the weighted average waiting time of the two time periods is added together and divided by the total number of passengers to obtain the average waiting time of the passengers.

6. The multi-objective path planning method for an intelligent public transportation system according to claim 5, characterized in that: Obtaining congestion includes: in, For the itinerary The congestion in The threshold for congestion can be set based on the ride experience the operator wants to provide; is the journey with the end point j In the bus First arrival Number of passengers at the station; is the element of itinerary S; is the maximum capacity of bus b.

Images