A dynamic evolution analysis method for studying transportation mode choice in ride-sharing systems

Abstract

The present invention discloses a dynamic evolution analysis method for studying transportation mode selection in a ride-sharing system, comprising: obtaining a set of travel modes for travelers, wherein the travel mode set includes travel modes selectable by both car-based and non-car-based travelers; calculating travel costs for different travel modes based on the travel mode set; calculating generalized travel costs based on the travel costs, wherein the generalized travel costs take into account the constraint relationship between the number of passengers in a ride-sharing trip and the number of drivers in a ride-sharing trip; obtaining the probability of travelers selecting different transportation modes based on the generalized travel costs; and constructing a daily dynamic evolution model for ride-sharing travel. The method, based on the daily dynamic evolution model for ride-sharing travel, obtains the dynamic evolution of travelers' daily transportation mode selection behavior based on the selection probabilities. The present invention can simulate daily fluctuations in traffic flow for different travel modes, providing technical support for transportation planning and management departments in formulating management decisions related to ride-sharing travel.

Description

Technical Field

[0001] The present invention belongs to the technical field of dynamic analysis of ride-sharing traffic, and in particular relates to a dynamic evolution analysis method for studying traffic mode selection in a ride-sharing system. Background Art

[0002] As an emerging urban transportation mode, ridesharing is gaining popularity due to its flexibility, low cost, ease of traffic congestion, and minimal increase in vehicle ownership. It offers urban residents an alternative to driving their own car or taking public transportation, changing traveler behavior and the spatial distribution of traffic flow, posing new challenges and opportunities for urban transportation development.

[0003] While existing studies have established equilibrium models for ridesharing users, these models primarily focus on static equilibrium states of travel mode choice and fail to capture the daily dynamics of traveler mode choice. In particular, existing models often assume that a ridesharing driver can only carry one passenger at a time. This ignores the actual capacity of ridesharing vehicles and fails to accurately reflect the daily variations in traffic flow across different modes. This simplifying assumption is inconsistent with reality and limits the models' applicability in describing and predicting real-world traffic flow dynamics. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention proposes a dynamic evolution analysis method for studying the choice of transportation mode in a shared travel system to solve the problems existing in the above-mentioned prior art.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a dynamic evolution analysis method for studying transportation mode selection in a ride-sharing system, comprising:

[0006] Obtaining a travel mode set for a traveler, wherein the travel mode set includes travel modes selectable by travelers with cars and travelers without cars;

[0007] Calculating travel costs of different travel modes based on the travel mode set;

[0008] Calculating a generalized travel cost based on the travel cost, wherein the generalized travel cost takes into account the constraint relationship between the number of passengers in the ride-sharing trip and the number of drivers in the ride-sharing trip;

[0009] According to the generalized travel cost, the probability of the traveler choosing different travel modes is obtained;

[0010] A daily dynamic evolution model of carpooling is constructed, and the dynamic evolution process of travelers' daily travel mode selection behavior is obtained through the daily dynamic evolution model of carpooling according to the selection probability.

[0011] Preferably, calculating the travel costs of different travel modes includes:

[0012] Calculate the time and fuel costs of private car travel;

[0013] Calculate the inconvenience cost of ride-sharing and the driver's compensation income;

[0014] Calculate the travel time cost, average wait time cost, and bus fare for taking the bus.

[0015] Preferably, obtaining the probability of a traveler selecting different travel modes includes:

[0016] Based on the travel mode of the Logit model, the probability of travelers choosing different travel modes is obtained; wherein, the selection probability takes into account the random error term between the travel cost perceived by the traveler and the actual travel cost.

[0017] Preferably, the daily dynamic evolution model of carpooling is a model constructed based on discrete time, and the model allows a carpooling driver to carry multiple carpooling passengers in one trip.

[0018] Preferably, the dynamic evolution analysis method further includes:

[0019] The perceived travel cost in the travel cost is adjusted through a weighted average learning model; wherein, in the weighted average learning model, the perceived travel cost of each travel mode is a linear combination of the actual travel cost of the previous day and the perceived travel cost of the current day.

[0020] In a second aspect, the present invention further provides a dynamic evolution analysis system for studying transportation mode selection in a ride-sharing system, comprising:

[0021] A travel mode acquisition module is used to obtain a travel mode set of travelers, wherein the travel mode set includes travel modes selectable by travelers with cars and travelers without cars;

[0022] A first calculation module, configured to calculate travel costs of different travel modes according to the travel mode set;

[0023] A second calculation module is configured to calculate a generalized travel cost based on the travel cost, wherein the generalized travel cost takes into account the constraint relationship between the number of passengers in the shared trip and the number of drivers in the shared trip;

[0024] A selection probability calculation module is used to obtain the probability of a traveler choosing different travel modes based on the generalized travel cost;

[0025] The model prediction module is used to construct a daily dynamic evolution model of shared travel, and obtain the dynamic evolution process of the traveler's daily travel mode selection behavior through the daily dynamic evolution model of shared travel according to the selection probability.

[0026] In a third aspect, the present invention further discloses a computer device comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the first aspect.

[0027] In a fourth aspect, the present invention further discloses a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described in the first aspect when executed by a processor.

[0028] In a fifth aspect, the present invention further discloses a computer program product, comprising a computer program, which implements the steps of the method described in the first aspect when executed by a processor.

[0029] Compared with the prior art, the present invention has the following advantages and technical effects:

[0030] The present invention provides a dynamic evolution analysis method for studying the choice of transportation mode in a shared travel system, comprising: first, obtaining a set of travel modes of travelers, wherein the travel mode set includes travel modes selectable by travelers with cars and travelers without cars; second, calculating the travel costs of different travel modes based on the travel mode set; then, calculating the generalized travel cost based on the travel cost, wherein the generalized travel cost takes into account the constraint relationship between the number of shared travel passengers and the number of shared travel drivers; further, obtaining the selection probability of different travel transportation modes by travelers based on the generalized travel cost; finally, constructing a daily dynamic evolution model of shared travel, and obtaining the dynamic evolution process of travelers' daily travel transportation mode selection behavior through the daily dynamic evolution model of shared travel based on the selection probability.

[0031] The present invention can capture the dynamic changes in travelers' transportation modes over a period of time through the daily dynamic evolution model of shared travel, simulate the fluctuations in traffic flow caused by different daily travel modes, and provide technical support and theoretical basis for traffic planning and management departments to formulate management decisions related to shared travel. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] 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:

[0033] Figure 1 This is a schematic diagram of a linear single-center city according to an embodiment of the present invention;

[0034] Figure 2 A daily evolution diagram of the traffic flow of non-car travelers in an embodiment of the present invention;

[0035] Figure 3 A daily evolution diagram of traffic flow of different travel modes for car travelers according to an embodiment of the present invention;

[0036] Figure 4 Flowchart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0037] 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.

[0038] 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.

[0039] First, the technical terms involved in the following embodiments are explained.

[0040] Ride-sharing is a form of shared travel. A ride-sharing service provider typically publishes trip information in advance, allowing people traveling the same route to choose to ride in the ride-sharing provider's vehicle, sharing some of the travel costs or providing free assistance. Ride-sharing can take many forms, including public welfare ride-sharing and mutual aid ride-sharing. Public welfare ride-sharing is free and based entirely on the principle of mutual assistance; while mutual aid ride-sharing allows for the sharing of some costs, such as fuel and tolls, based on actual circumstances.

[0041] Ride-sharing, also known as carpooling, refers to a shared travel experience where people traveling the same route choose to ride in a ride-sharing provider's car, sharing some of the travel costs or providing free transportation. Depending on whether a fee is charged, ride-sharing can be categorized as either public welfare or mutual assistance. Public welfare ride-sharing is free and based entirely on the principle of mutual assistance; mutual assistance ride-sharing allows for appropriate cost-sharing based on the actual situation.

[0042] Ride-sharing can be applied to a variety of scenarios, including commuting, holiday travel, and travel. For example, long-term ride-sharing for commuting allows colleagues who live nearby to discuss a ride together; long-distance ride-sharing for holiday travel or travel allows people to find companions and save a lot of money.

[0043] Ride-sharing is legal, but it's important to note that sharing rides in "illegal taxis" is not permitted. To ensure your safety and rights, be sure to choose a legitimate ride-sharing platform or ride with reliable people. In addition, some cities have established "ride-sharing lanes" to encourage and guide intensive travel, guaranteeing priority for multi-passenger vehicles and buses, maximizing road capacity, alleviating traffic congestion, and reducing traffic pollution.

[0044] Dynamic evolution analysis is widely used in many fields, including but not limited to:

[0045] Economic field: Study the changing patterns of economic variables over time, such as prices and demand.

[0046] Technology theme analysis: In the dynamic evolution analysis of technology themes, the DPL-BMM model is a labeled hybrid model based on the two-term Dirichlet process, which is used to reveal the context of technology evolution.

[0047] Policy implementation: In the study of the generation and dynamic evolution of the "joint efforts and shared management" strategy, the dynamic changes in the policy implementation process and their impact on the implementation effect are analyzed.

[0048] The specific steps or processes of dynamic evolution analysis include:

[0049] Data collection: Collect time series data on relevant variables.

[0050] Time series analysis: Establish a time series model to analyze the changing trajectory of variables over time.

[0051] Statistical correlation analysis: Conduct statistical correlation analysis to determine the correlation between variables.

[0052] Regression analysis: Regression analysis is used to establish the functional relationship between variables.

[0053] Result interpretation: Explain the analysis results and put forward policy recommendations or optimization plans.

[0054] The advantages of dynamic evolution analysis include:

[0055] Comprehensiveness: Considers the interaction of internal and external factors of the system and provides a comprehensive analytical perspective.

[0056] Dynamics: emphasizes the impact of time changes on the system and is suitable for dynamic research of complex systems.

[0057] Probabilistic: The impact is considered to be probabilistic, providing a more flexible analysis framework.

[0058] Disadvantages of dynamic evolution analysis include:

[0059] High data requirements: A large amount of high-quality time series data is required, and data acquisition and processing are difficult.

[0060] Computational complexity: Dynamic evolution analysis involves complex statistical and regression analysis, which has a high computational cost.

[0061] Difficulty of interpretation: Interpretation of the results may be complex due to the involvement of multiple factors and probability relationships.

[0062] In summary, the dynamic evolution analysis method reveals the law of system change over time through time series analysis and statistical correlation analysis, and is suitable for research and application in multiple fields.

[0063] For urban transportation systems involving carpooling, this embodiment analyzes travelers' travel mode decision-making processes and proposes a discrete-time daily dynamic evolution model for carpooling. This model describes the daily dynamic evolution of travelers' travel mode choices within the urban transportation system. During model construction, this embodiment relaxes the assumption that "a carpooling driver can only carry one passenger," allowing carpooling drivers to carry multiple passengers simultaneously during a trip. This compensates for the shortcomings of existing dynamic evolution models that ignore the passenger capacity of carpooling vehicles. Furthermore, the model can capture the dynamic changes in travelers' transportation modes over time and simulate the daily fluctuations in traffic flow due to different travel modes, providing technical support and a theoretical basis for transportation planning and management departments to formulate management decisions related to carpooling.

[0064] This embodiment provides a complete technical solution, such as Figure 4 As shown, the following steps are included:

[0065] Step 1: Definition of available travel modes and related symbols;

[0066] This example considers a linear single-center city with a ride-sharing service (e.g. Figure 1 As shown in the figure, there is a road and a bus lane, and travelers travel from their place of residence to their place of work every day by various modes of transportation. These travelers can be divided into two categories according to their vehicle ownership: car-owning travelers and car-free travelers. Taking into account the car usage habits of car-owning travelers, it is believed that there are three travel modes for car-owning travelers to choose from: driving alone, traveling as a carpooling driver (that is, participating in carpooling services as a driver), and traveling as a carpooling passenger (that is, participating in carpooling services as a passenger). For car-free travelers, there are two travel modes to choose from: traveling as a carpooling passenger and traveling as a bus passenger. Therefore, there are four travel roles in the urban transportation system with the participation of carpooling: single-driving drivers, carpooling drivers, carpooling passengers, and bus passengers.

[0067] In order to distinguish between car-owning and car-free travelers, the variables and parameters related to car-owning travelers are marked with a superscript 1, while the variables and parameters related to car-free travelers are marked with a superscript 0. To simplify the representation, the variables or parameters related to the four travel roles (single driver, carpool driver, carpool passenger, and bus passenger) are represented by subscripts v, d, p, and u. In order to better represent the set of travel roles that can be selected by car-owning and car-free travelers, this embodiment introduces H for representation. Therefore, the set of travel roles that can be selected by car-owning and car-free travelers is represented as H 1 = {v, d, p} and H 0 = {p,u}. The symbol i is introduced to represent the travel role chosen by the traveler. In the city, there are q 1 Car owners and q 0 The number of people without cars who travel from their place of residence to their place of work is The number of car-free travelers who choose travel role i is For example represents the number of car-owning travelers who choose to share a ride, Represents the number of car-free travelers who choose to share rides.

[0068] Step 2: Calculate travel costs for different travel modes;

[0069] Step 2-1: Cost composition analysis of different travel modes;

[0070] The first mode of travel, basic cost calculation of private car:

[0071] 1) Private car travel time

[0072] Vehicles on road a can be divided into two categories: ride-sharing vehicles and single-driver vehicles. The number of ride-sharing vehicles is equal to the number of ride-sharing drivers on road a. The number of single-driver vehicles is equal to the number of single-driver drivers on road a Therefore, the traffic flow on road a can be calculated as the sum of the traffic flow of carpooling drivers and the traffic flow of single drivers. This embodiment uses the travel time function of the Bureau of Public Roads (BPR) of the United States, which is widely used in academia, to calculate the travel time of a private car on road a. The calculation formula is as follows:

[0073]

[0074] in represents the free flow travel time on road a; c arepresents the traffic capacity of road a; α, β are non-negative parameters in the BPR function.

[0075] By introducing a time value parameter, i.e., the monetary amount corresponding to a unit of time, this embodiment can convert travel time into travel time costs. However, considering that the time value of travelers with and without cars is different, to ensure that the travel costs of the two types of travelers are more accurately reflected, this embodiment introduces two time value parameters, η1 and η0, where η1 represents the time value of travelers with cars, and η0 represents the time value of travelers without cars. Therefore, this embodiment can obtain the travel time costs of travelers with and without cars when using private cars:

[0076] <1> Travel time costs for car users

[0077]

[0078] <2> Travel time costs for people without cars

[0079]

[0080] 2) Fuel costs for private cars

[0081] In addition to the travel time cost, private cars also incur fuel costs. The formula for calculating private car fuel costs is: τ f l, where τ f is the fuel cost per unit distance during driving, and l is the length of road a.

[0082] The second mode of travel, calculation of additional costs for shared travel:

[0083] 1) Inconvenience cost

[0084] Unlike travelers who choose to drive alone or take the bus, travelers who choose to share a car, whether they are passengers or drivers, need to endure the inconvenience of sharing a car with others, such as the discomfort caused by sharing the same space with strangers and the extra detour time caused by picking up and dropping off passengers. This inconvenience is closely related to the average number of passengers per carpool, and as the number of passengers in a car increases, the inconvenience cost felt by the driver and each passenger will also increase. Therefore, the inconvenience cost is calculated as: where τ u Indicates the inconvenience fee when sharing a ride with a stranger. represents the average number of passengers carried by any vehicle on road a, that is, the ratio of the number of carpooling passengers to the number of carpooling drivers. They represent the flow of people with cars serving as carpooling passengers and the flow of people without cars serving as carpooling passengers respectively.

[0085] 2) Compensation income for ride-sharing drivers

[0086] Each carpooling passenger needs to pay the carpooling driver a fare τ according to the pre-determined pricing rules a , to compensate the carpool driver for providing travel services. Therefore, the compensation income received by the carpool driver is equal to the sum of the fares paid by all the passengers he carries, and the calculation formula is:

[0087] 3) Rideshare driver's passenger fees

[0088] Each carpool passenger needs to pay the carpool driver a fare τ a , so the fare for the shared passenger is τ a .

[0089] The third mode of travel, cost calculation of taking the bus:

[0090] 1) Bus operating hours

[0091] Since there is a bus lane between the residence and the workplace, the running time of the bus can be considered to be fixed at t b .

[0092] 2) Average waiting time

[0093] Due to the limited capacity of buses, passengers traveling by bus will have to wait for the bus at the departure station for an extra time. The bus runs on the dedicated route b according to the scheduled schedule, and the headway is fixed at h b Assuming that the passenger's departure station is the starting node of dedicated line b, and that passengers arrive at the station evenly and take the bus in a first-come, first-served order, the average waiting time is a piecewise function of the passenger flow. Using linear approximation techniques, the average waiting time can be approximately calculated as: in, represents the number of passengers taking the bus, c b Indicates the capacity of the bus.

[0094] 3) Bus fares

[0095] Passengers riding the bus need to pay a fixed fare τ b .

[0096] Step 2-2: Calculate travel costs for different travel modes;

[0097] Based on the above analysis of the cost structure of different travel modes, this embodiment can calculate the travel costs corresponding to different travel modes for travelers with cars and travelers without cars as follows:

[0098] (1) For a car owner who chooses to drive alone, his travel cost is composed of travel time and fuel cost, namely:

[0099]

[0100] (2) For car owners who choose to be a ride-sharing driver, their travel costs include travel time, fuel costs, and inconvenience costs minus compensation income, that is:

[0101]

[0102] (3) For car owners who choose to travel as a shared passenger, their travel costs are composed of travel time costs, inconvenience costs and fare costs, namely:

[0103]

[0104] (4) For those who do not have a car and choose to travel as a shared passenger, their travel cost is composed of travel time cost, inconvenience cost and fare, namely:

[0105]

[0106] (5) For those who do not have a car and choose to travel as a bus passenger, their travel cost is composed of the bus running time cost, the average waiting time cost and the bus fare, that is:

[0107]

[0108] Step 3: Calculation of generalized travel costs;

[0109] In a transportation system with ride-sharing services, there is a constraint between the number of ride-sharing passengers and the number of ride-sharing drivers:

[0110]

[0111] Where Cap represents the upper limit of the number of carpool passengers that a carpool vehicle can carry. Formula (9) indicates that the number of carpool passengers should be greater than or equal to the number of carpool drivers, and formula (10) indicates that the number of carpool passengers should be less than or equal to the upper limit of the passenger capacity that all carpool drivers can provide.

[0112] Assume that travelers choose their travel mode based on the cost calculated in step 2. As rational individuals, they usually tend to choose the travel mode with the lowest cost. However, in this case, the number of carpooling drivers and carpooling passengers may not meet the above constraints. This means that in the carpooling system, the supply side (the number of seats provided by carpooling drivers) is higher than the demand side (the number of carpooling passengers), or the demand side is lower than the supply side. Therefore, in order to ensure that the constraint relationship between the number of carpooling passengers and the number of carpooling drivers is satisfied, it is necessary to consider the impact of these constraints on the cost of different travel modes. This impact can be reflected by introducing Lagrange multipliers associated with the constraints. Let ν + For The associated Lagrange multiplier represents the reduction in fare due to the reduction in the number of carpooling passengers. - For The related Lagrange multipliers represent the increase in fares due to the increase in the number of carpooling passengers. The conditions satisfied by the two multipliers are as follows:

[0113]

[0114] Therefore, this embodiment can obtain the generalized travel costs corresponding to different travel modes chosen by car travelers and non-car travelers, which are calculated as follows:

[0115] (1) For a car owner who chooses to drive alone, the generalized travel cost is:

[0116]

[0117] (2) For car owners who choose to be a ride-sharing driver, their generalized travel cost is:

[0118]

[0119] (3) For car owners who choose to travel as carpool passengers, their generalized travel cost is:

[0120]

[0121] (4) For those who do not have a car and choose to travel as a carpool passenger, their generalized travel cost is:

[0122]

[0123] (5) For those who do not have a car and choose to travel as a bus passenger, their generalized travel cost is:

[0124]

[0125] Step 4: Probability formula for travel mode selection based on the Logit model;

[0126] Due to the limited travel information available, travelers can only estimate but cannot accurately predict the actual travel costs of various travel modes. Therefore, there is always a random error term between the travel costs perceived by travelers and the actual travel costs. The Logit model is used to describe the travel mode selection behavior of travelers under the principle of random user equilibrium. Therefore, the probabilities of car travelers and non-car travelers choosing travel mode i are:

[0127]

[0128] Among them, P i 1 P represents the probability that a car traveler chooses transportation mode i, i 0 represents the probability that a traveler without a car chooses travel mode i, and θ is the perception error parameter.

[0129] Step 5: Construct a daily dynamic evolution model of carpooling;

[0130] Travelers need to choose their transportation mode every day, so the dynamic system that explores their daily transportation mode choice behavior is discrete. Furthermore, rationality generally drives travelers to choose the transportation mode with the lowest perceived travel cost. They use available traffic information to estimate the costs of various transportation modes and update these estimates based on past travel experience.

[0131] To describe the adjustment process of perceived travel costs, this embodiment adopts a widely used weighted average learning model, in which the perceived travel cost of each mode of transportation is a linear combination of the actual travel cost of the previous day and the perceived travel cost of the current day. However, in a transportation network with carpooling services, there is a quantity constraint between carpooling passengers and carpooling drivers. Therefore, the flow of various types of travelers in the system on day t will be regulated by the carpooling market. This regulation is achieved through the multiplier ν corresponding to the constraint relationship. + 、ν - reflect.

[0132] As an additional implementation method, the multiplier introduced by the relationship between the number of drivers and passengers is unique, and the established daily dynamic evolution model of shared travel can be expressed by the following equation, which can be represented by the adjustment process of perceived travel costs.

[0133]

[0134] Among them, x (t) is the matrix composed of the flow of various types of travelers on day t, is the matrix of the perceived travel costs of various modes of transportation for travelers on day t, δ is the weight factor required by the learning model, which is used to indicate the degree of influence of previous travel experience on the current cost estimate. The smaller its value, the more the traveler relies on traffic information when making decisions; the larger its value, the more habitual the traveler's choice behavior is. C(x (t-1) ) is the matrix consisting of the actual travel costs corresponding to various modes of transportation on day t-1, q is the matrix of the demand of people with cars and people without cars, is a matrix consisting of the probabilities of choosing various modes of transportation. ν (t) is a matrix of multipliers corresponding to the quantity constraint relationship between carpooling passengers and carpooling drivers, ν (t) =(v + ,ν - );Φ(ν (t) ) is the matrix of functions that the multiplier needs to satisfy,

[0135] As an additional implementation, it can be shown mathematically that the multiplier ν + 、ν - Uniqueness in dynamic evolution model. The nonlinear complementarity problem in formula (22) 0≤ν (t) ⊥Φ(ν (t) )≥0 can be equivalent to the variational inequality problem Φ(ν (t)* )(ν (t) -ν (t)* )≥0, where ν (t)* represents ν in equilibrium (t) Since it can be mathematically proved that Φ(ν (t) ) is about ν (t) A monotonic function, so according to the known It can be determined by 0≤ν (t) ⊥Φ(ν (t) )≥0 and Get the only ν (t) =(v + ,ν - ) and x (t) Therefore, the established dynamic evolution model can ensure that the multiplier ν (t) =(v + ,ν - ) and x (t)This ensures that the established daily evolution model of carpooling can simulate the dynamic evolution of travelers' daily travel mode selection behavior and simulate the fluctuations of daily traffic flow under different travel modes.

[0136] Step 6: Programming solution;

[0137] MATLAB (R2023a) is used to program the contents involved in the above steps, and GAMS (version win6425.1.3) is used to program the flow distribution part (Formulas (1)-(21)). After completing the programming of MATLAB and GAMS, interactive programming is performed between the two to achieve the solution purpose. Specifically, first, in MATLAB, this embodiment will set the demand number of travelers with cars and travelers without cars, the parameters required by the model of this embodiment, and the feasible initial flow (i.e., the initial number of people who choose each type of travel mode) that meets the constraints (9)-(10). Then, MATLAB will solve the perceived travel cost of each type of travel mode. Subsequently, through the interaction between GAMS and MATLAB, these data (feasible initial flow and perceived travel cost of each type of travel mode) will be automatically passed to GAMS. In GAMS, by using the built-in solver PATH, this embodiment can solve the flow distribution part and obtain the actual number of people who choose each type of travel mode (i.e., the actual flow corresponding to each type of travel mode) and the multiplier. After the solution is completed, the GAMS solution results (the actual flow and multipliers corresponding to each travel mode) will be automatically returned to MATLAB for the next calculation of the perceived cost of each travel mode (that is, the perceived travel cost of each travel mode for the next day). Through this interaction, the calculated perceived travel cost data for each travel mode will be automatically passed to GAMS to solve the actual flow corresponding to each travel mode for the next day. This process will continue to loop until the actual flow of each travel mode for all days is solved. Finally, MATLAB will plot the curve of the actual flow corresponding to each travel mode over time.

[0138] This embodiment provides an example based on a linear single-center city (e.g. Figure 1 As shown in Figure 2, there is a corridor between the residence and the workplace. This corridor consists of a bus lane road b and a parallel road a. The capacity and free flow time of the two roads are: a =1.5×10^5Veh / h,c b =3.5×10^4Veh / h, The headway of the bus is: h b= 5min. The daily dynamic evolution model proposed in the present embodiment is used to explore the travel mode selection behavior of travelers in the city within one month (30 days).

[0139] Other parameters required by the model are detailed in Table 1. The demand for car travelers and non-car travelers is: q 1 =1.5×10^5 peopleq 0 =1.5×10^5 people. Assume that the initial flow of various types of travelers is That is, the number of car-owning travelers who choose to be solo drivers, carpool drivers, and carpool passengers is 50*10^3 people; the number of car-free travelers who choose to be carpool passengers and take the bus are 1×10^5 and 5×10^4 people respectively.

[0140] Table 1

[0141] parameter Value α,β 0.15，4 <![CDATA[η1,η0]]> 40$ / h, 20$ / h <![CDATA[τ f ]]> 0.1$ / km l 0.1$ / km <![CDATA[τ u ]]> $2 τ $8 <![CDATA[τ b ]]> $2 θ 0.1 δ 0.2

[0142] According to the above steps, interactive programming between GAMS and MATLAB was performed to obtain the actual traffic flow evolution over time corresponding to the various travel modes chosen by non-car travelers in the urban transportation system. Figure 2 The actual traffic flow corresponding to the various travel modes chosen by car travelers evolves over time. Figure 3 . Figure 2 and Figure 3 It not only shows the actual daily traffic flow of various modes of transportation, but also clearly reflects the evolution of the actual daily traffic flow of various modes of transportation over time. In addition, the evolution diagram can also capture the stable state that the system eventually reaches.

[0143] 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 dynamic evolution analysis method for studying the choice of transportation mode in a ride-sharing system, characterized by: The following steps are involved: Obtaining a travel mode set for a traveler, wherein the travel mode set includes travel modes selectable by travelers with cars and travelers without cars; Calculating travel costs of different travel modes based on the travel mode set; Calculating a generalized travel cost based on the travel cost, wherein the generalized travel cost takes into account the constraint relationship between the number of passengers in the ride-sharing trip and the number of drivers in the ride-sharing trip; According to the generalized travel cost, the probability of the traveler choosing different travel modes is obtained; Constructing a daily dynamic evolution model of shared travel, and obtaining a dynamic evolution process of travelers' daily travel mode selection behavior through the daily dynamic evolution model of shared travel according to the selection probability; The daily dynamic evolution model of carpooling is a discrete-time model that allows a carpooling driver to carry multiple carpooling passengers on a single trip. The model simulates the dynamic evolution of travelers' daily transportation mode selection behavior and the fluctuations in traffic flow for different travel modes. The multiplier introduced by the relationship between the number of drivers and passengers is unique. The established daily dynamic evolution model of carpooling is expressed by the following equation, which is represented by the adjustment process of perceived travel costs: Among them, x (t) is the matrix composed of the flow of various types of travelers on day t, is the matrix of the perceived travel costs of various modes of transportation for travelers on day t, δ is the weight factor required by the learning model, which is used to indicate the influence of previous travel experience on the current cost estimate, C(x (t-1) ) is the matrix of the actual travel costs corresponding to various modes of transportation on day t-1, q is the matrix of the demand for car travelers and non-car travelers, is a matrix consisting of the probabilities of choosing various modes of transportation, ν (t) is a matrix composed of multipliers corresponding to the quantity constraint relationship between carpooling passengers and carpooling drivers, Φ(ν (t) ) is the matrix of functions that the multipliers need to satisfy.

2. The dynamic evolution analysis method for studying transportation mode selection in a ride-sharing system according to claim 1 is characterized in that: Calculating travel costs for different travel modes includes: Calculate the time and fuel costs of private car travel; Calculate the inconvenience cost of ride-sharing and the driver's compensation income; Calculate the travel time cost, average wait time cost, and bus fare for taking the bus.

3. The dynamic evolution analysis method for studying transportation mode selection in a ride-sharing system according to claim 1 is characterized in that: Obtaining the probability of travelers choosing different travel modes includes: Based on the travel mode of the Logit model, the probability of travelers choosing different travel modes is obtained; wherein, the selection probability takes into account the random error term between the travel cost perceived by the traveler and the actual travel cost.

4. The dynamic evolution analysis method for studying transportation mode selection in a ride-sharing system according to claim 1 is characterized in that: Also includes: The perceived travel cost in the travel cost is adjusted through a weighted average learning model; wherein, in the weighted average learning model, the perceived travel cost of each travel mode is a linear combination of the actual travel cost of the previous day and the perceived travel cost of the current day.

5. A dynamic evolution analysis system for studying the choice of transportation mode in a ride-sharing system, characterized by: include: A travel mode acquisition module is used to obtain a travel mode set of travelers, wherein the travel mode set includes travel modes selectable by travelers with cars and travelers without cars; A first calculation module, configured to calculate travel costs of different travel modes according to the travel mode set; A second calculation module is configured to calculate a generalized travel cost based on the travel cost, wherein the generalized travel cost takes into account the constraint relationship between the number of passengers in the shared trip and the number of drivers in the shared trip; A selection probability calculation module is used to obtain the probability of a traveler choosing different travel modes based on the generalized travel cost; The model prediction module is used to construct a daily dynamic evolution model of shared travel, and obtain the dynamic evolution process of the traveler's daily travel mode selection behavior through the daily dynamic evolution model of shared travel according to the selection probability.

6. A computer device comprising a memory, a processor, and a computer program stored in the memory, wherein: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

8. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

Images