An effective path set acquisition method for multi-modal transportation networks

Abstract

The present invention discloses a method for obtaining an effective path set for a multi-modal transportation network, comprising the following steps: Step 1: Establish subnets for each transportation mode: including obtaining the longitude and latitude coordinates of each node in the road network and the adjacency relationship between nodes; Step 2: Construct connections between different transportation mode networks: determine the connection relationship between nodes according to distance and determine connection attributes; Step 3: Determine the generalized cost function for each transportation mode; Step 4: Obtain n path schemes, obtain the generalized cost of all directed arcs on the path through an improved A* optimal path algorithm and record the path direction. Step 5: Return the path set. The present invention realizes path search for a multi-modal transportation road network, improves the A* algorithm, and can return multiple optimal paths for multi-modal transportation within a specified number at one time.

Description

Technical Field

[0001] The present invention belongs to the technical field of travel analysis in urban transportation systems, and particularly relates to a method for obtaining an effective path set for a multi-modal transportation system. Background Art

[0002] After years of large-scale construction of transportation infrastructure in China, each city has formed a multi-modal transportation travel system with a certain scale and distinct characteristics that integrates multiple modes. Due to the differences in service characteristics between different transportation modes, a complete trip usually requires the joint use of multiple transportation modes. Therefore, multi-modal combined trips that utilize the advantages of each mode through mode conversion at transportation hubs to achieve more efficient travel have become a prominent feature of urban transportation systems. Currently, traditional single-mode transportation network models and path optimization algorithms can no longer meet the diverse transportation travel analysis requirements.

[0003] Path search methods can generally be classified into three categories. The first category is the shortest path planning algorithm based on graph theory methods, the second category is the heuristic algorithm based on traditional artificial intelligence theory, and the third category is modern intelligent optimization algorithms that reflect the ideas and content of natural phenomena such as simulating biological evolution and interpreting neuroscience. The algorithms based on graph theory have a high time complexity, large memory consumption, and low search efficiency, and cannot meet the real-time path planning requirements. The algorithms based on heuristic thinking may not necessarily guarantee the optimal solution. Moreover, the ideas of simulating natural phenomena such as biological evolution or neuroscience may require the introduction of more parameters and complex computational models, making the implementation and debugging of the algorithms more difficult.

[0004] Although current path planning methods have been widely applied in the engineering field, there are still the following deficiencies: (1) The connection relationship between subnets of each transportation mode is not precise enough when constructing the super network; (2) Traditional basic path planning algorithms only return one result, cannot reflect the random factors when users select paths, and are even less able to reflect path selection under multiple optimization objectives. Modern intelligent algorithms have high algorithm complexity and may not necessarily return the optimal solution or even multiple solutions.

[0005] In view of this situation, in order to meet the analysis requirements of urban multi-modal travel and ensure the accuracy of subsequent distribution results, there is an urgent need for a convenient and fast method for obtaining an effective path set for a multi-modal transportation network. Summary of the Invention

[0006] In order to solve the technical problems mentioned in the above background art, the present invention proposes a method for obtaining an effective path set for a multi-modal transportation network.

[0007] In order to achieve the above technical objectives, the technical solution of the present invention is: A method for obtaining an effective path set for a multi-modal transportation network, the method comprising the following steps:

[0008] Step 1: Establish subnetworks for each transportation mode: Abstract the transportation network as a network graph, record the longitude and latitude coordinates of each node in the road network, and the adjacency relationship between nodes.

[0009] Step 2: Construct connections between different transportation mode networks: Supplement the connection relationship between nodes according to the distance between nodes, and determine the connection attributes, that is, the transportation mode attributes of the connection edges.

[0010] Step 3: Determine the generalized cost function for each transportation mode.

[0011] Step 4: Obtain the path plan: Obtain the costs of all directed arcs on the path through an improved A* path algorithm and record the path direction.

[0012] Step 5: Return the path set under specified quantity conditions.

[0013] Preferably, the steps for obtaining information in Step 1 include:

[0014] Step 11: The subnetworks for each transportation mode include the bus line network, the rail transit line network, and the road network.

[0015] Step 12: The recorded information includes the longitude and latitude coordinates of bus stops, rail transit stations, and road intersections, as well as the connection relationships of bus lines, rail transit lines, and road intersections.

[0016] Preferably, Step 2 is to establish the connection relationship between each transportation subnetwork on the basis of Step 1 to achieve the interconnection and interoperability between single transportation mode networks. Specifically, it includes the following steps:

[0017] Step 21: Traverse the nodes of the single-mode network pairwise and calculate the distance between nodes:

[0018]

[0019] In the formula, Dist a,b is the distance between points a and b; r is the radius of the earth; lon a , lat b are the longitude and latitude of node a, lon b , lat b are the longitude and latitude of node b.

[0020] Step 22: According to the distance dist a,b between nodes and the maximum walking tolerance distance D walking and the maximum cycling tolerance distance D ridingGenerate connection lines with different attributes according to the size relationship. The maximum tolerable walking distance and the maximum tolerable cycling distance can be calculated based on travel data. Usually, the 85th percentile walking distance and the 85th percentile cycling distance can be taken as the maximum tolerable walking distance and the maximum tolerable cycling distance. If dist ≤ D walking Then a walking attribute section is generated between the two nodes; if D walking <dist ≤ D riding Then a bicycle attribute section can be generated between the two nodes.

[0021] Preferably, step 3 is to determine the road rights of sections on the multi-modal road network, which specifically includes the following steps: Step 31: Calculate the travel time of sections with different attributes

[0022]

[0023]

[0024]

[0025]

[0026]

[0027] In the formula, T1 is the travel time of the car section, is the zero-flow time on the car section; v 1 is the car traffic volume on the section; C1 is the traffic capacity of the car section; α, β are blocking coefficients; T2 is the travel time of the bicycle section; is the zero-flow time on the bicycle section; v 2 is the bicycle traffic volume on the section; C2 is the traffic capacity of the bicycle section; T3 is the travel time of the walking section, s3 is the average walking speed; l is the section length; T4 is the travel time of the bus section, T5 is the travel time of the subway section;

[0028] is the running time between two bus stops, is the running time between two subway stops. The departure frequencies of buses and subways are fixed and can be obtained according to the operation schedule

[0029] Step 32: Calculate the comfort loss of sections with different attributes

[0030] U i =λ·s i ·T i

[0031] where \(i\in\{1,2,3,4,5\}\), \(s_1\), \(s_2\), \(s_3\), \(s_4\), \(s_5\) respectively represent the comfort loss per unit time of car travel mode, bicycle travel mode, walking travel mode, bus travel mode, and subway travel mode; \(\lambda\) is the conversion coefficient of comfort loss time,

[0032] Step 33: Calculate the monetary costs of different attribute sections

[0033] \(P_1 = k\cdot f\cdot l\)

[0034] \(P_2 = P_3 = 0\)

[0035] where \(P_1\) is the monetary cost of the car section, \(f\) is the price of fuel consumption per kilometer of the car (\(yuan / km\)), \(k\) is the conversion coefficient of monetary cost - time; \(P_2\), \(P_3\) are the costs of the bicycle and walking sections; the value of the monetary cost \(P_4\) of the bus section is equal to the single - fare of the bus line; the monetary cost \(P_5\) of the subway section is calculated according to the actual subway line mileage,

[0036] Step 34: Combine travel time, monetary cost, and comfort loss in a linear weighted manner:

[0037]

[0038] where is the generalized cost of using travel mode \(i\) on section \(r\), the subscript \(r\) represents a specified section, and the superscript \(i\) represents a specified travel mode; \(\omega\) T is the importance coefficient of travel time, \(\omega\) P is the importance coefficient of monetary cost, \(\omega\) U is the importance coefficient of comfort, satisfying \(\omega\) T +\(\omega\) P [[ID=...]]+\(\omega\) U = 1,

[0039] Step 35: Specify the cost of the access section: The travel cost of the access section includes the cost of the walking section entering the public transportation network or the cost of the bicycle section

[0040] Step 36: Specify the cost of the egress section: The travel cost of the egress section includes the cost of the walking section leaving the public transportation network or the cost of the bicycle section

[0041] Preferably, step 4 obtains the time spent on all directed arcs on the path and records the path direction by improving the A* optimal path algorithm, specifically including the following steps:

[0042] Step 41: Maintain the minimum heap openList, closeList, and camefrom list.

[0043] Step 42: Determine the starting node start and the target node goal. By default, the initial attributes of the starting node and the target node are both walking nodes. The number of paths n to be returned is calculated. Calculate the cost function f(n) of the starting node and add the starting node to the openList. The cost function consists of the actual cost g(n) and the heuristic function h(n):

[0044] f(n) = g(n) + h(n)

[0045] g(n) = ∑ j∈N c j

[0046]

[0047]

[0048] In the formula, f(n) is the total estimated cost from the initial state through state n to the target state; g(n) is the actual cost from the initial state to state n, N is the set of road segments passed from the initial state to state n, and c j is the generalized cost of path j; h(n) is the heuristic function, which is the estimated cost from state n to the target state. It does not represent the subsequent actual moving distance but reflects the time - distance trend from the target node; Dist current,goal is the distance between the current judgment node and the target node, s represents the travel speed, and a current represents the attribute of the current evaluation node, which determines the average speed a of the transportation mode adopted by s current ∈ {1, 2, 3, 4, 5}, and when the value is 1, 2, 3, 4, 5, the average speed of a car, the average speed of a bicycle, the average speed of walking, the average speed of a bus, and the average speed of the subway are adopted respectively.

[0049] Step 43: Judge whether the openList is empty and whether the number of paths already collected has satisfied n. The two judgment conditions are taken as an intersection. If either of the above conditions is satisfied, the program ends and Step 4 ends; if neither of the above conditions is satisfied, pop the first element current of the openList list and add current to the closed list closeList.

[0050] Step 44: Sequentially access the adjacent nodes next of the current node current. If the next node is the target node goal, reconstruct the path according to the camefrom list, and add the path and the corresponding generalized cost to the result set; if the next node is not the target node, then determine whether the node already exists in the closeList.

[0051] Step 45: If the node already exists in the closeList, directly skip this node; if the node does not exist in the closeList, calculate the actual standby g(n) at this time, which can be obtained from the following state transition formula:

[0052] g(n) = g(n - 1) + c r ,

[0053] In the formula, g(n - 1) is the actual cost of the parent node state of state n, and c r is the generalized cost of the section r passed from state n - 1 to state n.

[0054] Step 46: Determine whether the next node already exists in the openList. If the next node is not in the openList, add the next node and the evaluation function f(n) of the node to the openList, and update the parent node of the node in camefrom; if the next node already exists in the openList, and the calculated actual cost g(n) at this time is less than the actual cost already recorded in the openList, then update the evaluation function f(n) of this node and update the parent node of the node in camefrom; if the node already exists and the actual cost at this time is greater than the recorded actual cost, then skip this next node and retain the previously calculated smaller cost information.

[0055] Step 47: Continuously repeat steps 43 to 46.

[0056] Preferably, step 5 returns the final path set according to the path generalized cost recorded in step 4, including the following steps:

[0057] Step 51: Output the top n shortest generalized cost paths considering multi-dimensional factors such as travel time, monetary cost, and comfort, as well as the corresponding generalized costs.

[0058] Beneficial effects brought by adopting the above technical solutions:

[0059] Compared with the existing methods, the present invention has the following obvious advantages: First, the present invention fully considers the psychological and physiological factors of travelers when using different transportation modes to determine the connection line generation rules, can efficiently construct a multi-modal travel transportation network that conforms to the actual situation, and can improve the accuracy of path search and traffic assignment; Second, the present invention improves the classic and efficient A* algorithm and applies it to the path search problem of the multi-modal transportation network, significantly reducing the number of nodes in the search process, reducing the calculation amount and the search for redundant nodes, and improving the search efficiency. At the same time, the improved algorithm can return multiple effective paths with the top n in the general cost at one time, providing a solid foundation for studying the random selection factors of travelers, enabling related traffic analysis research to more accurately predict or reproduce traffic operation conditions, and thus guiding the formulation of traffic management and induction measures by traffic planning and management departments. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is an example diagram of establishing a traffic network in Embodiment 2;

[0061] Figure 2 It is an example diagram of the connection line generation situation in Embodiment 2;

[0062] Figure 3 It is a flowchart of the method for obtaining an effective path set for a multi-modal transportation system according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0063] Hereinafter, the technical solutions of the present invention will be described in detail with reference to the accompanying drawings.

[0064] Embodiment 1: Refer to Figure 3 , a method for obtaining an effective path set for a multi-modal transportation network, the method comprising the following steps:

[0065] Step 1: Establish subnetworks for each transportation mode: Abstract the transportation network into a network diagram, record the longitude and latitude coordinates of each node in the road network, and the adjacency relationship between nodes.

[0066] Step 2: Construct connections between different transportation mode networks: Supplement the connection relationship between nodes according to the distance between nodes and determine the connection attributes, that is, the transportation mode attributes of the connection edges.

[0067] Step 3: Determine the general cost function for each transportation mode.

[0068] Step 4: Obtain a path plan: Obtain the costs of all directed arcs on the path and record the path direction through the improved A* path algorithm.

[0069] Step 5: Return a path set under specified quantity conditions.

[0070] Preferably, the steps of obtaining information in Step 1 include:

[0071] Step 11: Each of the transportation mode subnets includes a bus line network, a rail transit line network, and a road network.

[0072] Step 12: The recorded information includes the longitude and latitude coordinates of bus stops, rail transit stations, and road intersections, as well as the connection relationships among bus lines, rail transit lines, and road intersections.

[0073] Preferably, Step 2 is to establish the connection relationships among the transportation sub-networks based on Step 1, so as to achieve the interconnection and interoperability among single transportation mode networks. It specifically includes the following steps:

[0074] Step 21: Traverse the nodes of the single-mode network pairwise and calculate the distance between the nodes:

[0075]

[0076] In the formula, Dist a,b is the distance between points a and b; r is the radius of the earth; lon a , lat b are the longitude and latitude of node a, lon b , lat b are the longitude and latitude of node b,

[0077] Step 22: Generate connection lines with different attributes according to the magnitude relationship between the distance dist a,b between the nodes and the maximum walking tolerance distance D walking , and the maximum cycling tolerance distance D riding . The maximum tolerance walking distance and the maximum tolerance cycling distance can be calculated based on travel data. Usually, the 85th percentile walking distance and the 85th percentile cycling distance can be taken as the maximum tolerance walking distance and the maximum tolerance cycling distance. If dist ≤ D walking , then a walking attribute section is generated between the two nodes; if D walking <dist ≤ D riding , then a bicycle attribute section can be generated between the two nodes.

[0078] Preferably, Step 3 is to determine the road rights of the sections on the multi-mode road network, which specifically includes the following steps: Step 31: Calculate the travel time of sections with different attributes

[0079]

[0080]

[0081]

[0082]

[0083]

[0084] In the formula, T1 is the travel time of the car section, is the zero-flow time on the car section; v 1 is the car traffic volume on the section; C1 is the traffic capacity of the car section; α and β are blocking coefficients; T2 is the travel time of the bicycle section; is the zero-flow time on the bicycle section; v 2 is the bicycle traffic volume on the section; C2 is the traffic capacity of the bicycle section; T3 is the travel time of the walking section, s3 is the average walking speed; l is the section length; T4 is the travel time of the bus section, T5 is the travel time of the subway section;

[0085] is the running time between two bus stops, is the running time between two subway stops. The departure frequencies of buses and subways are fixed, and can be obtained according to the operation schedule

[0086] Step 32: Calculate the comfort loss of sections with different attributes

[0087] U i = λ·s i ·T i

[0088] In the formula, i ∈ {1, 2, 3, 4, 5}, s1, s2, s3, s4, and s5 respectively represent the comfort loss per unit time of the car travel mode, bicycle travel mode, walking travel mode, bus travel mode, and subway travel mode; λ is the comfort loss time conversion coefficient,

[0089] Step 33: Calculate the monetary costs of sections with different attributes

[0090] P1 = k·f·l

[0091] P2 = P3 = 0

[0092] In the formula, P1 is the monetary cost of the car section, f is the fuel consumption price per kilometer of the car (yuan / km), k is the monetary cost-time conversion coefficient; P2 and P3 are the costs of the bicycle and walking sections; the value of the monetary cost P4 of the bus section is equal to the single fare of the bus line; the monetary cost P5 of the subway section is calculated according to the actual subway line mileage,

[0093] Step 34: Combine travel time, monetary cost, and comfort loss in a linear weighted manner:

[0094]

[0095] In the formula, is the generalized cost of using transportation mode i on road section r. The subscript r represents a specified road section, and the superscript i represents a specified transportation mode; ω T is the importance coefficient of travel time, ω P is the importance coefficient of monetary cost, ω U is the importance coefficient of comfort, and it satisfies ω T +ω P +ω U = 1,

[0096] Step 35: Specify the cost of the access road section: The travel cost of the access road section includes the cost of the walking road section entering the public transportation road network or the cost of the bicycle road section

[0097] Step 36: Specify the cost of the egress road section: The travel cost of the egress road section includes the cost of the walking road section leaving the public transportation road network or the cost of the bicycle road section

[0098] Preferably, step 4 obtains the time cost of all directed arcs on the path and records the path direction by improving the A* optimal path algorithm, which specifically includes the following steps:

[0099] Step 41: Maintain the minimum heap openList, closeList, and camefrom list,

[0100] Step 42: Determine the starting node start and the target node goal. By default, the initial attributes of the starting node and the target node are both walking nodes. The number of paths n to be returned is calculated, and the cost function f(n) of the starting node is calculated. The starting node is added to the openList. The cost function is composed of the actual cost g(n) and the heuristic function h(n):

[0101] f(n) = g(n) + h(n)

[0102] g(n) = ∑ j∈N c j

[0103]

[0104]

[0105] In the formula, f(n) is the total estimated cost from the initial state through state n to the target state; g(n) is the actual cost from the initial state to state n, N is the set of road sections passed from the initial state to state n, and c jis the generalized cost of path j; h(n) is the heuristic function, which is the estimated cost from state n to the target state. It does not represent the subsequent actual moving distance, but reflects the time - distance trend from the current node to the target node; Dist current,goal is the distance between the current judgment node and the target node, s represents the travel speed, a current represents the attribute of the current evaluation node, which determines the average speed a of the transportation mode adopted by s current ∈{1, 2, 3, 4, 5}. When the value is 1, 2, 3, 4, 5, the average speed of car, bicycle, walking, bus, and subway is adopted respectively,

[0106] Step 43: Judge whether openList is empty and whether the number of collected paths has satisfied n. Take the intersection of the two judgment conditions. If either of the above conditions is satisfied, the program ends and Step 4 ends; if neither of the above conditions is satisfied, pop the first element current of the openList list, and add current to the closed list closeList.

[0107] Step 44: Access the adjacent node next of the current node current in turn. If the next node is the target node goal, reconstruct the path according to the camefrom list, and add the path and the corresponding generalized cost to the result set; if the next node is not the target node, judge whether the node already exists in the closeList.

[0108] Step 45: If the node already exists in the closeList, directly skip this node; if the node does not exist in the closeList, calculate the actual standby g(n) at this time, which can be obtained by the following state - transfer formula:

[0109] g(n) = g(n - 1)+c r ,

[0110] In the formula, g(n - 1) is the actual cost of the parent - node state of state n, c r is the generalized cost of the road section r passed from state n - 1 to state n,

[0111] Step 46: Determine whether the next node already exists in the openList. If the next node is not in the openList, add the next node and the node's evaluation function f(n) to the openList, and update the parent node of the node in camefrom; if the next node already exists in the openList, and the actual cost g(n) calculated at this time is less than the actual cost recorded in the openList, then update the evaluation function f(n) of the node and update the parent node of the node in camefrom; if the node already exists and the actual cost at this time is greater than the actual cost recorded, then skip the next node and retain the previously calculated smaller cost information.

[0112] Step 47: Repeat steps 43 to 46.

[0113] Preferably, step 5 returns the final path set according to the generalized path costs recorded in step 4, including the following steps:

[0114] Step 51: Output the top n shortest generalized cost paths and corresponding generalized costs taking into account the multi-dimensional factors of travel time, monetary cost, and comfort.

[0115] Example 2: See Figure 1 、 Figure 2 The present invention introduces a method for obtaining an effective path set for a multi-modal transportation system. The technical solution of the present invention is further described in detail below with reference to examples and drawings. Figure 1 , including the following steps:

[0116] (1) Establish each transportation mode subnetwork. This includes recording and obtaining the latitude and longitude coordinates of each node in the road network, as well as the adjacency relationship between nodes. The network construction result can be expressed as follows: Figure 1 As shown:

[0117] (2) Construct connections between networks with different transportation modes. Supplement the connection relationship between nodes based on the distance between nodes and determine the connection attributes, that is, the transportation mode attributes of the connecting edge; assume that D walking =1km,

[0118] D riding = 2km. Based on the distance between nodes dist and the maximum walking distance D walking , Maximum riding distance D riding The size relationship generates connecting lines with different attributes. If dist≤D walking Then a walking attribute section can be generated between the two nodes; if D walking <dist≤D riding Then a bicycle attribute section can be generated between the two nodes; if dist>D ridingOnly car - attribute road segments can be generated between the two nodes. The details of generating the connecting road segments and the network after connection are as Figure 2 shown;

[0119] (3) Determine the generalized cost functions of each transportation mode. Here, it is assumed that P = U = 0, the average walking speed is 1.1 (m / s), and the average cycling speed is 3.3 (m / s).

[0120] (4) Obtain the path plan. The time cost of all directed arcs on the path is obtained through the improved A* path algorithm, and the path direction is recorded. Due to the large number of road intersections, the iterative data update display is cumbersome. Therefore, the rail transit and bus line networks are used as examples for path search here.

[0121] When specifically implementing the algorithm program, the longitude and latitude information of the nodes and the affiliated transportation sub - networks can be stored using a dictionary. The structure is: Nodes = {'node number': [(lon, lat), node attribute]}. The transportation modes corresponding to the numbers 1, 2, 3, 4, 5 are car, bicycle, walking, bus, and subway respectively. The adjacency relationship of the nodes can be stored by storing the connected nodes and the corresponding connection road segment attributes. The structure is: Graph = {'node number': [('neighbor node number', connection road segment attribute)]};

[0122] (41) Maintain the minimum heap openList, closeList, and camefrom list

[0123]

[0124] (42) Determine the starting node start and the target node goal. By default, the initial attributes of the starting node and the target node are both walking nodes, and the number of paths to be returned is 2.

[0125] Calculate the cost function f(start) of the starting node, and add the starting node to the openList. The cost function is composed of the actual cost g(n) and the heuristic function h(n):

[0126] g(start) = 0

[0127]

[0128] f(n) = 0 + 2.02 = 2.02 (h)

[0129]

[0130] (43) Determine whether the openList is empty and whether the number of collected paths has satisfied n. If either of the above conditions is met, the program ends and step (4) ends; if neither of the above conditions is met, pop the first element current of the openList and add it to the closed list closeList.

[0131]

[0132] (44) Visit the adjacent nodes next of the current node current in sequence. At this time, the adjacent nodes of the start node are C, G, and A.

[0133] (45) If the nodes C, G, and A do not exist in the closeList, further determine whether they already exist in the openList.

[0134] (46) Calculate the actual cost g(n) and f(n):

[0135] g(C) = 0 + 1.41 / (3.3 * 3.6) = 0.118 (h)

[0136] g(G) = 0 + 1 / (1.1 * 3.6) = 0.253 (h)

[0137] g(A) = 0 + 1.41 / (3.3 * 3.6) = 0.168 (h)

[0138]

[0139]

[0140]

[0141] The nodes C, G, and A are not in the openList and are directly added to the openList. The update situation is as follows:

[0142]

[0143] After traversing the adjacent nodes of the start node, pop the node with the smallest estimated value from the openList again. At this time, the C node is popped and added to the closeList. Next, traverse the adjacent nodes of the C node. The adjacent nodes of the C node are start, B, and D; the start node already exists in the closeList, so skip it; the B and D nodes are not in the closeList and are also not in the closeList, so add them to the openList; update the openList for the second time, and the update situation is as follows:

[0144]

[0145] At this time, the adjacent nodes of node C have been traversed. Pop an element from the openList again, and node D is popped. Update the OpenList again. The update situation is as follows:

[0146]

[0147] Until the target node goal is traversed, reconstruct the path according to the camefrom dictionary, and record the total travel time at this time, and add it to the result list. The code for reconstructing the path is as follows:

[0148]

[0149] (47) Continuously repeat steps (43) to (46) until the openList list is empty or the number of path solutions in the result reaches 2, and step four ends.

[0150] (51) Output each path solution: the nodes passed by the path and the generalized cost. According to the assumptions of the embodiment, the following can be finally obtained:

[0151] Path 1: start->C->D->E->F->3->2->1->goal, with a time consumption of 0.539h;

[0152] Path 2: start->C->D->E->2->-1->goal, with a time consumption of 0.555h;

[0153] If the specified number of solutions n = 3 at initialization, in addition to the above solutions, the following can also be obtained after the algorithm is executed:

[0154] Path 3: start->C->D->E->F->2->1->goal, with a time consumption of 0.576h.

[0155] The embodiments are only for explaining the technical idea of the present invention, and the protection scope of the present invention cannot be limited thereby. Any changes made on the basis of the technical solution according to the technical idea proposed by the present invention fall within the protection scope of the present invention.

Claims

1. A method for obtaining an effective path set for a multimodal transportation network, characterized in that: The method comprises the following steps: Step 1: Establish each transportation mode subnet: Abstract the transportation network into a network diagram, record the longitude and latitude coordinates of each node in the road network, and the adjacency relationship between the nodes. Step 2: Build connections between networks with different transportation modes: Supplement the connection relationship between nodes based on the distance between nodes and determine the connection attributes, that is, the transportation mode attributes of the connecting edges. Step 3: Determine the generalized cost function for each mode of transportation, Step 4: Get the path plan: Get the cost of all directed arcs on the path by improving the A* path algorithm and record the path direction. Step 5: Return the path set under the specified quantity conditions; Step 1 is as follows: Step 11: The transportation mode subnetworks include a bus line network, a rail transit line network, and a road network. Step 12: The recorded information includes the longitude and latitude coordinates of bus stops, rail transit stops, and road intersections, as well as the connection relationships among bus lines, rail transit lines, and road intersections. Step 2 is as follows: Step 2 is to establish a connection relationship between each transportation sub-network based on step 1 to achieve interconnection and intercommunication between single transportation mode networks, which specifically includes the following steps: Step 21: Traverse the nodes of the single-mode network two by two and calculate the distance between the nodes: Where Dist a,b is the distance between points a and b; r is the radius of the earth; lon a ,lat b is the latitude and longitude of node a, lon b ,lat b is the latitude and longitude of node b, Step 22: According to the distance dist between nodes a,b Maximum walking distance D walking , Maximum riding distance D riding The maximum tolerable walking distance and the maximum tolerable cycling distance can be calculated based on the travel data. Usually, the 85% walking distance and the 85% cycling distance can be taken as the maximum tolerable walking distance and the maximum tolerable cycling distance. If dist≤D walking Then a walking attribute section is generated between the two nodes; if D walking <dist≤D riding Then a bicycle attribute section can be generated between the two nodes.

2. According to the method for obtaining an efficient path set for a multimodal transportation network according to claim 1, step 3 is to define the generalized cost of a road segment on the road network, wherein the generalized cost function includes three components: travel time T, cost P, and comfort loss U, and specifically comprises the following steps: Step 31: Calculate travel time for different attribute road segments Where T1 is the travel time of the car section, is the zero flow time on the car road section; v 1 is the car traffic volume on the road section; C1 is the traffic capacity of the car section; α and β are the blocking coefficients; T2 is the travel time of the bicycle section; is the zero flow time on the bicycle section; v 2 is the bicycle traffic volume on the road section; C2 is the traffic capacity of the bicycle section; T3 is the travel time of the walking section, s3 is the average walking speed; l is the length of the road section; T4 is the travel time of the bus section, and T5 is the travel time of the subway section; is the running time of the bus between two stops, The running time between two subway stations. The bus and subway departure frequencies are fixed and can be obtained according to the running timetable. Step 32: Calculate the comfort loss of different attribute sections U i =λ·s i ·T i , Where, i∈{1,2,3,4,5}, s1, s2, s3, s4, and s5 represent the comfort loss per unit time of car travel, bicycle travel, walking travel, bus travel, and subway travel, respectively; λ is the comfort loss time conversion coefficient, Step 33: Calculate the monetary cost of different attribute segments P1=k·f·l, P2=P3=0, Where P1 is the monetary cost of the car section, f is the fuel consumption price per kilometer, RMB / km, and k is the monetary cost-time conversion coefficient; P2 and P3 are the costs of bicycle and walking sections; the monetary cost of the bus section P4 is equal to the single bus line fare; the monetary cost of the subway section P5 is calculated based on the actual subway line mileage. Step 34: Combine travel time, monetary cost, and comfort loss using a linear weighted approach: Where, is the generalized cost of using transportation mode i on road section r, where the subscript r represents a specific road section and the superscript i represents a specific transportation mode; ω T is the travel time importance coefficient, ω P is the monetary cost importance coefficient, ω U is the comfort importance coefficient, satisfying ω T +ω P +ω U =1, Step 35: Define the cost of the Internet access route: The travel cost of the Internet access route includes the walking route cost of entering the public transportation network or bicycle route fees Step 36: Define the cost of the off-network segment: The off-network segment travel cost includes the cost of the walking segment leaving the public transportation network or bicycle route fees 3. The method for obtaining an effective path set for a multimodal transportation network according to claim 2, characterized in that: Step 4 is as follows: Step 41: Maintain the minimum heap openList, closeList, camefrom list, Step 42: Determine the starting node start and the target node goal. By default, the initial attributes of the starting node and the target node are both walking nodes. The number of paths to return is n. Calculate the cost function f(n) of the starting node and add the starting node to the openList. The cost function consists of the actual cost g(n) and the heuristic function h(n): f(n)=g(n)+h(n) g(n)=∑ j∈N c j Where f(n) is the estimated total cost from the initial state to the target state via state n; g(n) is the actual cost from the initial state to state n; N is the set of road sections from the initial state to state n; c j is the generalized cost of path j; h(n) is the heuristic function, which is the estimated cost from state n to the target state. It does not represent the subsequent actual moving distance, but reflects the time distance trend from the target node; Dist current,goal is the distance between the current judgment node and the target node, s represents the travel speed, a current Indicates the attributes of the current evaluation node, which determines the average speed a of the transportation mode used by s current ∈{1, 2, 3, 4, 5}, when the values are 1, 2, 3, 4, and 5, the average speed of cars, bicycles, pedestrians, buses, and subways are used respectively. Step 43: Determine whether openList is empty and whether the number of collected paths meets n. The two judgment conditions are intersected. If one of the above conditions is met, the program ends and step 4 ends. If the above conditions are not met at the same time, the first element current in the openList list is popped out and current is added to the closed list closeList. Step 44: Visit the adjacent nodes next of the current node in sequence. If the next node is the target node goal, rebuild the path according to the camefrom list and add the path and the corresponding generalized cost to the result set. If the next node is not the target node, determine whether the node already exists in the CloseList. Step 45: If the node already exists in the closeList, skip the node directly; if the node does not exist in the CloseList, calculate the actual standby g(n) at this time, which can be obtained by the following state transition formula: g(n)=g(n-1)+c r , In the formula, g(n-1) is the actual cost of the parent node state of state n, c r is the generalized cost of the road segment r from state n-1 to state n, Step 46: Determine whether the next node already exists in the OpenList. If the next node is not in the openList, add the next node and the node's evaluation function f(n) to the openList, and update the parent node of the node in camefrom. If the next node already exists in the openList, and the actual cost g(n) calculated at this time is less than the actual cost recorded in the openList, update the evaluation function f(n) of the node, and update the parent node of the node in camefrom. If the node already exists and the actual cost at this time is greater than the actual cost recorded, then skip the next node and retain the previously calculated smaller cost information. Step 47: Repeat steps 43 to 46.

4. The method for obtaining an effective path set for a multimodal transportation network according to claim 3, characterized in that: The step 5 is specifically as follows: Step 51: Output the top n shortest generalized cost paths and the corresponding generalized costs taking into account the multi-dimensional factors of travel time, monetary cost, and comfort.

Images