Emergency Bus Transfer Scheduling Method Considering Passengers' Travel Choice Behavior under Sudden Subway Interruptions

Abstract

The present invention relates to an emergency bus transfer scheduling method considering passengers' travel choice behavior under sudden subway interruptions. This method obtains the basic attribute values of travel alternative plans, the influence data of different sudden interruption scenarios on passengers' travel choice behavior, the preset scenario variables, passenger attribute variables, and travel characteristic variables that affect travel choice behavior, calculates the value function value, decision weight function value, and the cumulative prospect value corresponding to special attributes under different plans, constructs a passenger travel choice behavior model based on subway interruptions, further establishes an emergency bus transfer scheduling model considering passengers' travel choice behavior, and uses an optimization algorithm to obtain the optimal solution of the decision variables of the scheduling model. This method is beneficial to accurately analyze passengers' travel mode choice behavior and improve the efficiency of emergency bus transfer scheduling.

Description

Technical Field

[0001] The present invention belongs to the technical field of passenger emergency evacuation, and in particular relates to an emergency bus connection scheduling method that takes into account passengers' travel selection behaviors when a subway is suddenly interrupted. Background Art

[0002] Emergency shuttle buses are often used to connect stranded passengers due to subway interruptions, but currently, the connection is often inefficient and the emergency response is not matched. For example, a large number of passengers are piled up at some stations and wait for a long time for the shuttle bus, while at some stations, the shuttle bus seats are vacant and no passengers get on. The direct cause of this situation is mainly: the calculation of the connection demand of blocked passengers is unreasonable, that is, all blocked passenger flows or blocked passenger flows are uniformly multiplied by a fixed coefficient and assumed to be the total connection demand, which obviously does not conform to the travel rules of blocked passengers.

[0003] The characterization of passengers' travel choice behavior is crucial to exploring the travel patterns of passengers. It is affected by the interruption scenario, personal characteristics, travel habits, and other factors, and considers the proportion of passengers choosing any travel mode under interruption conditions from multiple perspectives. Therefore, it is necessary to re-examine the emergency bus connection scheduling problem under subway interruption considering passengers' travel choice behavior.

[0004] At present, both at home and abroad, attention has been paid to the study of the problem of "emergency bus connection scheduling under sudden subway interruption". However, most studies focus on the operation end of the connection plan, such as departure interval optimization, driving plan optimization, and bus operation timetable optimization. Few studies focus on the calculation of blocked passenger connection demand and explore the passenger travel choice behavior. At the same time, existing studies in this field often regard passengers as completely rational individuals. This assumption is too ideal. Obviously, passengers have differences in travel experience, travel needs, and psychology, and it is impossible to predict all uncertainties. Therefore, how to accurately characterize the travel choice behavior of blocked passengers under the condition of ensuring the limited rationality of passengers has become a key problem that needs to be solved in the overall arrangement of emergency bus connection scheduling, and it has important practical significance. Summary of the invention

[0005] The purpose of the present invention is to provide an emergency bus connection scheduling method that takes into account passengers' travel selection behavior under sudden subway interruption. The method is conducive to accurately analyzing passengers' travel mode selection behavior and improving the efficiency of emergency bus connection scheduling.

[0006] To achieve the above object, the technical solution adopted by the present invention is: an emergency bus connection scheduling method considering passengers' travel choice behavior under sudden subway interruption. This method obtains the basic attribute values of travel alternative plans, the influence data of different sudden interruption scenarios on passengers' travel choice behavior, the preset scenario variables, passenger attribute variables, and travel characteristic variables that affect travel choice behavior, calculates the value function value, decision weight function value, and the cumulative prospect value corresponding to special attributes under different plans, constructs a passenger travel choice behavior model based on subway interruption, further establishes an emergency bus connection scheduling model considering passengers' travel choice behavior, and uses an optimization algorithm to obtain the optimal solution of the decision variables of the scheduling model.

[0007] Furthermore, the specific implementation steps of this method are as follows:

[0008] Step S1: Collect questionnaires to obtain the basic attribute values of travel alternative plans;

[0009] The travel alternative plans at least include: waiting in place, standard connection bus, direct connection bus, skip-stop connection bus, regular bus, taxi / online car-hailing, and shared bicycle;

[0010] The basic attribute values of the travel alternative plans at least include: the possible travel time of travel plan k between OD pairs rs and the number of transfers required for travel plan k between OD pairs rs;

[0011] Step S2: Use the questionnaires collected in Step S1 to obtain the influence data of different sudden interruption scenarios on passengers' travel choice behavior, and through data analysis, obtain the preset scenario variables, passenger attribute variables, and travel characteristic variables that affect passengers' travel choice behavior after a sudden subway interruption event;

[0012] The influence data at least includes: the number of people whose travel time for choosing plan k is the number of people choosing plan k under different preset scenarios, the number of people choosing plan k under different passenger attributes, and the number of people choosing plan k under different travel characteristics;

[0013] The preset scenario variables at least include: travel origin and destination stations, passenger status, travel purpose, event occurrence time period, and interruption duration;

[0014] The passenger attribute variables at least include: gender, age group, education level, monthly income, occupation, and monthly working hours;

[0015] The travel characteristic variables at least include: passengers' travel choice behavior during normal subway operation, subway interruption experience, information attention degree, and risk travel attitude;

[0016] Step S3: Using the impact data obtained in step S2, further obtain the set K of alternative travel options selected by passengers and calculate the average working time O of passengers. k , Average monthly income of passengers S k , the travel time of plan k is The objective probability of occurrence p is calculated to calculate the passenger reference point time Passenger time cost coefficient θ k 、The profit and loss value y of the passenger choosing travel plan k k , and then calculate the passenger value function g(y k ), passenger revenue decision weight function ω + (p i ), passenger loss decision weight function ω - (p j );

[0017] Step S4: Using the passenger value function g(y k ), passenger revenue decision weight function ω + (p i ), passenger loss decision weight function ω - (p j ), calculate the cumulative benefit decision weight function Cumulative loss decision weight function Further calculate the cumulative prospect value V(f i ) + , the cumulative prospect value of the loss part V(f j ) - ;

[0018] Step S5: Using the impact data obtained in step S2 and the profit part obtained in step S4, the cumulative prospect value V(f i ) + , the cumulative prospect value of the loss part V(f j ) - , calibrate the parameter β corresponding to the zth variable in travel plan k k , and then calculate the cumulative prospect value V(f k ) and the probability P of the passenger choosing the kth travel plan in the travel alternative set K k , thereby constructing a passenger travel choice behavior model based on subway disruption;

[0019] Step S6: Using the passenger travel choice behavior model based on subway interruption constructed in step S5, further establish the up and down passenger travel cost F p , Time-based emergency bus operation cost T b , Emergency bus operation cost S based on mileage bA combined emergency bus transfer scheduling model that takes into account passengers' travel choice behavior to minimize the travel costs of passengers and the operating costs of emergency buses under constraints;

[0020] Step S7: For the established combined emergency bus transfer scheduling model considering passengers' travel choice behavior, use an optimization algorithm to solve it, obtain the optimal solution of the decision variables of the scheduling model, and use the optimal solution as the basis for controlling the emergency bus departure frequency, the required number of buses, and the bus load factor during transfer, so as to complete the emergency bus transfer scheduling method considering passengers' travel choice behavior under sudden subway interruptions.

[0021] Further, step S2 specifically includes the following steps:

[0022] Step S21: For the influence data of different sudden interruption scenarios on passengers' travel choice behavior, calculate the probability values of each subclass selection scheme k: the probability value of selection scheme k under different preset scenarios, the probability value of selection scheme k under different passenger attributes, and the probability value of selection scheme k under different travel characteristics;

[0023] Step S22: Use the calculated probability values of each subclass selection scheme k to draw a distribution ratio diagram of each subclass travel scheme k;

[0024] Step S23: Use the drawn distribution ratio diagram of each subclass travel scheme k to compare and obtain the preset scenario variables, passenger attribute variables, and travel characteristic variables that affect passengers' travel choice behavior.

[0025] Further, step S3 specifically includes the following steps:

[0026] Step S31: For the obtained influence data, construct a set of alternative travel plans K = {K1, K2, K3, K4, K5, K6, K7} for passengers to choose from, where K1 is to choose to wait for rescue in place when the subway is affected by a sudden interruption event; K2 is to choose to take a standard transfer bus when the subway is affected by a sudden interruption event; K3 is to choose to take a direct bus when the subway is affected by a sudden interruption event; K4 is to choose to take an interval bus when the subway is affected by a sudden interruption event; K5 is to choose to take a skip-stop bus when the subway is affected by a sudden interruption event; K6 is to choose to take a network car when the subway is affected by a sudden interruption event; K7 is to choose to take a shared bicycle when the subway is affected by a sudden interruption event; calculate the average working time O of passengers for 7 alternative travel plans k and the average monthly income S of passengers k and the objective occurrence probability p that the travel time of plan k is ;

[0027] Step S32: According to the average working time O of passengers calculated in step S31 k and the average monthly income S of passengersk , further calculate the passenger time cost coefficient θ k , the formula is:

[0028]

[0029] Step S33: According to the passenger time cost coefficient θ k , calculate the passenger reference point time The formula is:

[0030]

[0031] In the formula, is the possible travel time of travel plan k between OD pairs rs obtained in step S1;

[0032] Step S34: According to the calculated passenger reference point time and the possible travel time of travel plan k between OD pairs rs calculate the profit and loss value y of the passenger choosing travel plan k k , the formula is:

[0033]

[0034] In the formula, considering that there are m + n + 1 possible travel times for the kth travel plan under uncertain conditions, denoted as arranged in descending order as λ is the function risk aversion coefficient;

[0035] Step S35: According to the profit and loss value y of the passenger choosing travel plan k calculated in step S34 k , further calculate the passenger value function g(y k ), the formula is:

[0036]

[0037] In the formula, α and β are the risk preference coefficients of the function, where α represents the risk aversion degree when there is a gain, and β represents the risk preference degree when there is a loss.

[0038] Furthermore, step S4 specifically includes the following steps:

[0039] Step S41: According to the objective occurrence probability p that the travel time of plan k calculated in step S31 is , further calculate the cumulative gain decision weight function cumulative loss decision weight function The formula is:

[0040]

[0041]

[0042] In the formula, ω + (p i ) is the decision-making weight function of passenger revenue, and ω - (p j ) is the decision-making weight function of passenger loss; γ is the revenue attitude coefficient, and δ is the loss attitude coefficient; p n represents the probability of occurrence of the nth revenue travel state; p -m represents the probability of occurrence of the mth loss travel state. Denote the objective probability of its occurrence corresponding to the travel time T as P = {p -m , …, p -1 , p0, p1, …, p n};

[0043] Step S42: According to the passenger value function g(y k ) calculated in step S35 and the cumulative revenue decision-making weight function and the cumulative loss decision-making weight function calculated in step S41, further calculate the cumulative prospect value V(f i ) + of the revenue part and the cumulative prospect value V(f j ) - of the loss part. The formula is:

[0044]

[0045] Further, step S5 specifically includes the following steps:

[0046] Step S51: Construct the calculation formula of the cumulative prospect value V(f k ) corresponding to the special attribute under different schemes and the probability P k of the kth travel scheme selected by the passenger in the travel alternative scheme set K:

[0047]

[0048] ν k = V(f k ) + ε k

[0049]

[0050] In the formula, X kz is the zth variable in the utility of travel scheme k, and β z is its corresponding coefficient; ν k is the selection utility of travel scheme k, and ε k is the prospect error term, which follows the Gumbel distribution of independent and identical distribution for each travel scheme;

[0051] Step S52: Using the obtained impact data, calculate the cumulative prospect value V(f i ) + of the profit part and the cumulative prospect value V(f j ) - of the loss part for each plan in the questionnaire, and the probability P k of choosing the k-th travel plan. Use the multiple regression method for parameter calibration to obtain the parameter values β z corresponding to the utility variables of each travel plan;

[0052] Step S53: According to the calculation formula constructed in Step S51, calculate the cumulative prospect value V(f k ) corresponding to the special attribute under different plans and the probability P k that the passenger chooses the k-th travel plan in the travel alternative plan set K, and complete the construction of the passenger travel choice behavior model based on subway interruption.

[0053] Furthermore, the specific steps of Step S6 are as follows:

[0054] Step S61: Determine that the transfer scheduling mode is a combined emergency bus transfer scheduling mode; the mode is divided into two types of routes. One type of route is the whole-line route L1, that is, the emergency transfer bus stops at each interruption site; the other is the large-station route L2, which only stops at the interruption sites with large passenger flow;

[0055] Step S62: According to the probability P k that the passenger chooses the k-th travel plan in the travel alternative plan set K calculated in Step S53, combined with the historical passenger flow OD data during the interruption duration of the subway sudden interruption section, calculate the number of people Q uv who need to transfer from u to v in the interruption section;

[0056] Step S63: According to the number of people Q uv who need to transfer from u to v in the interruption section, calculate the waiting time T h of passengers under the combined emergency transfer, the running time T zz between stations, the waiting time T zd of large-station passenger flow, and the waiting time T zq of the whole-line bus passenger flow, and further calculate the total travel cost F p of passengers under the combined emergency bus transfer. The formula is:

[0057]

[0058] F p = μ1T h + μ2(T zz + T zd + Tzq )

[0059] In the formula, f q is the departure frequency of emergency buses on the entire connecting route, f d is the departure frequency of emergency buses on major station shuttle routes, Q represents the total number of stranded people who choose emergency shuttle bus routes, and x u 、x v 、x k A 0-1 variable indicating whether the emergency bus stops at the station. When the emergency bus stops at the station, x u =x v =x k =1; t uv represents the emergency bus running time, which is obtained by dividing the interval travel distance by the average travel speed of the emergency bus, Q kv represents the number of passengers boarding the shuttle line at station k, Q uk represents the number of passengers getting off at station k on the entire station connecting line; represents the average time for a passenger to get on the bus, t represents the average time for a passenger to get off the bus; μ1 represents the unit waiting time cost, and μ2 represents the unit time cost on the bus;

[0060] Step S64: Calculate the bus operation cost F under the combined emergency bus connection b , based on the time-based bus operating cost T b and the mileage-based bus operating cost S b It consists of two parts, and the calculation formula is as follows:

[0061]

[0062] β d =f d / (f q +f d )

[0063] S b =(f q +f d )L

[0064] F b =ζ1T b +ζ2S b

[0065] Where, T o is the total operation time of emergency public transport; β d is the proportion of passengers who choose to take the bus at the main station; L is the single trip mileage of the emergency bus;

[0066] Step S65: Based on the calculated total passenger travel cost F p and emergency bus operating costs F b, and by using the three constraints that the number of scheduled emergency shuttle buses cannot exceed the maximum number of standby emergency shuttle buses, the average full-load rate of emergency buses is not less than the set value, and the departure frequency should be greater than the specified minimum departure frequency and less than the maximum departure frequency, the minimum total cost of the combined emergency bus shuttle scheduling is obtained: min(δ1F p +δ2F b ), and the construction of the combined emergency bus shuttle scheduling model considering passengers' travel choice behavior is completed.

[0067] Furthermore, step S7 specifically includes the following steps:

[0068] Step S71: Perform gene coding on the emergency bus shuttle scheduling model considering passengers' travel choice behavior. After initializing the population, calculate the fitness, perform roulette wheel selection with elite retention, crossover operation, and mutation operation in sequence to generate the next generation population and update the population;

[0069] Step S72: Determine whether the total cost of the emergency bus shuttle scheduling model considering passengers' travel choice behavior is the minimum. If so, output the best shuttle scheduling plan; otherwise, go to step S73;

[0070] Step S73: If the total cost of the emergency bus shuttle scheduling model considering passengers' travel choice behavior has not reached the minimum, continue to calculate the fitness, and determine whether the new fitness is less than the old fitness. If so, accept the new population according to the Metropolis criterion; otherwise, directly accept the new population;

[0071] Step S74: Perform roulette wheel selection with elite retention, crossover operation, and mutation operation based on the accepted new population to generate the next generation population and update the population until the minimum value of the total cost of the emergency bus shuttle scheduling model is obtained, and output the best shuttle scheduling plan.

[0072] Compared with the prior art, the present invention has the following beneficial effects:

[0073] By designing and collecting questionnaires, it obtains the basic attribute values of travel alternative plans, the influence data of different sudden interruption scenarios on passengers' travel choice behavior, the preset scenario variables, passenger attribute variables, and travel characteristic variables that affect travel choice behavior, calculates the value function value, decision weight function value, and the cumulative prospect value corresponding to special attributes under different plans, constructs a passenger travel choice behavior model based on subway interruption, further establishes an emergency bus shuttle scheduling model considering passengers' travel choice behavior, and uses an optimization algorithm to obtain the optimal solution of the decision variables of the scheduling model. The present invention can accurately analyze passengers' travel mode choice behavior, improve the efficiency of emergency bus shuttle scheduling, and provide an effective technical means for the emergency evacuation management of passengers under subway interruption.

[0074] Compared with the prior art, the present invention and its preferred embodiments have the following beneficial effects:

[0075] (1) By collecting passenger data through questionnaires, the present invention analyzes the preset scenario attributes, passenger attributes, and passenger travel characteristic attributes that affect passengers' travel choice behaviors under subway disruptions from multiple perspectives, and calculates the proportion of passengers who choose to take emergency shuttle buses to complete their subsequent trips during sudden subway disruptions, which can meet passengers' emergency travel needs at the source and breaks through the limitations of the prior art solutions where the connection demand is unclear, resulting in low connection efficiency;

[0076] (2) The present invention enables the choice of passengers' travel modes to reflect relatively bounded rationality, describes a relatively simple and direct choice process, which is more in line with people's choice habits and patterns, and improves the drawbacks of the prior art where, due to following the compensation principle, the relationship of mutual compensation is not satisfied among all attributes of each scheme;

[0077] (3) The optimal connection scheduling scheme calculated by the present invention can reduce passengers' travel costs and the operating costs of emergency buses, provide a theoretical basis for the bus operation management department to formulate efficient bus operation timetables and bus operation plans, and makes up for the deficiencies of the prior art in inaccurately depicting connection demands, low connection efficiency, and low load factors of connection vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 is a flowchart of the method implementation of an embodiment of the present invention;

[0079] Figure 2 is a flowchart of obtaining preset scenario variables, passenger attribute variables, and travel characteristic variables that affect passengers' travel choice behaviors after a sudden subway disruption event in an embodiment of the present invention;

[0080] Figure 3 is a flowchart of constructing a passenger travel choice behavior model based on subway disruptions in an embodiment of the present invention;

[0081] Figure 4 is a flowchart of constructing a combined emergency bus connection scheduling model considering passengers' travel choice behaviors in an embodiment of the present invention;

[0082] Figure 5 is a flowchart of obtaining the optimal solution of the decision variables of the scheduling model using an optimization algorithm in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0083] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0084] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs.

[0085] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0086] This embodiment provides an emergency bus connection scheduling method considering passengers' travel choice behavior under sudden subway interruptions. This method obtains the basic attribute values of travel alternative plans, the impact data of different sudden interruption scenarios on passengers' travel choice behavior, the preset scenario variables, passenger attribute variables, and travel characteristic variables that affect travel choice behavior, calculates the value function value, decision weight function value, and the cumulative prospect value corresponding to special attributes under different plans, constructs a passenger travel choice behavior model based on subway interruptions, further establishes an emergency bus connection scheduling model considering passengers' travel choice behavior, and uses an optimization algorithm to obtain the optimal solution of the decision variables of the scheduling model. As Figure 1 shown, the specific implementation steps of this method are as follows:

[0087] Step S1: Design and collect a questionnaire to obtain the basic attribute values of travel alternative plans.

[0088] The travel alternative plans at least include: waiting in place, standard connection bus, direct connection bus, skip-stop connection bus, regular bus, taxi / online car-hailing, and shared bicycle. The basic attribute values of the travel alternative plans at least include: the possible travel time of travel plan k between OD pairs rs and the number of transfers required for travel plan k between OD pairs rs.

[0089] Step S2: Use the questionnaire collected in Step S1 to obtain the impact data of different sudden interruption scenarios on passengers' travel choice behavior, and obtain the preset scenario variables, passenger attribute variables, and travel characteristic variables that affect passengers' travel choice behavior after a sudden subway interruption event through data analysis.

[0090] The impact data at least includes: the travel time of selected plan k is The number of people, the number of people choosing option k under different preset scenarios, the number of people choosing option k under different passenger attributes, and the number of people choosing option k under different travel characteristics. The preset scenario variables at least include: origin and destination stations of the trip, passenger status, travel purpose, time period when the event occurs, and interruption duration. The passenger attribute variables at least include: gender, age group, education level, monthly income, occupation, and monthly working hours. The travel characteristic variables at least include: passenger travel choice behavior during normal subway operation, subway interruption experience, degree of information attention, and attitude towards risky travel.

[0091] In this embodiment, the flowchart for implementing this step is as Figure 2 shown, and specifically includes the following steps:

[0092] Step S21: For the influence data of different sudden interruption scenarios on the passenger travel choice behavior, calculate the probability values of each subclass choosing option k: the probability value of choosing option k under different preset scenarios, the probability value of choosing option k under different passenger attributes, and the probability value of choosing option k under different travel characteristics.

[0093] Step S22: Use the calculated probability values of each subclass choosing option k to draw the distribution ratio diagram of each subclass travel option k.

[0094] Step S23: Use the drawn distribution ratio diagram of each subclass travel option k to compare and obtain the preset scenario variables, passenger attribute variables, and travel characteristic variables that affect the passenger travel choice behavior.

[0095] Step S3: Using the influence data obtained in step S2, further obtain the set K of alternative travel options chosen by passengers, calculate the average working time O of passengers k , the average monthly income S of passengers k , the objective occurrence probability p that the travel time of option k is , calculate the reference point time of passengers the passenger time cost coefficient θ k , the profit and loss value y of the passenger choosing travel option k k , and then calculate the passenger value function g(y k ), the passenger revenue decision weight function ω + (p i ), the passenger loss decision weight function ω - (p j ).

[0096] In this embodiment, the step S3 specifically includes the following steps:

[0097] Step S31: Based on the acquired impact data, construct a set of passenger travel alternatives K=K1, K2, K3, K4, K5, K6, K7}, where K1 is to choose to wait for rescue on the spot when the subway is affected by a sudden interruption; K2 is to choose standard shuttle bus when the subway is affected by a sudden interruption; K3 is to choose direct bus when the subway is affected by a sudden interruption; K4 is to choose interval bus when the subway is affected by a sudden interruption; K5 is to choose skip-stop bus when the subway is affected by a sudden interruption; K6 is to choose online car-hailing when the subway is affected by a sudden interruption; K7 is to choose shared bicycle when the subway is affected by a sudden interruption; calculate the average working time O of the 7 passengers who choose the travel alternatives k , Average monthly income of passengers S k , the travel time of plan k is The objective probability of occurrence p.

[0098] Step S32: The average working time O of passengers calculated in step S31 k , Average monthly income of passengers S k , further calculate the passenger time cost coefficient θ k , the formula is:

[0099]

[0100] Step S33: Based on the passenger time cost coefficient θ k , calculate the passenger reference point time The formula is:

[0101]

[0102] In the formula, It is the possible travel time between the OD pair rs of the travel plan k obtained in step S1.

[0103] Step S34: Based on the calculated passenger reference point time The possible travel time between OD pairs rs and travel plan k Calculate the profit and loss value y of the passenger choosing travel plan k k , the formula is:

[0104]

[0105] In the formula, considering that under uncertain conditions, the kth travel plan has m+n+1 possible travel times, recorded as Sort in descending order λ is the risk aversion coefficient of the function, which is calibrated to λ=2.25 in this embodiment.

[0106] Step S35: According to the profit and loss value y of the travel plan k selected by the passenger calculated in Step S34 k , further calculate the passenger value function g(y k ), and the formula is:

[0107]

[0108] In the formula, α and β are the risk preference coefficients of the function, where α represents the degree of risk aversion when there is a gain, and β represents the degree of risk preference when there is a loss. In this embodiment, it is calibrated that α = β = 0.88.

[0109] Step S4: Utilize the passenger value function g(y k ), the passenger gain decision weight function ω + (p i ), and the passenger loss decision weight function ω - (p j ) obtained in Step S3 to calculate the cumulative gain decision weight function cumulative loss decision weight function Further calculate the cumulative prospect value V(f i ) + and the cumulative prospect value of the loss part V(f j ) - .

[0110] In this embodiment, Step S4 specifically includes the following steps:

[0111] Step S41: According to the objective occurrence probability p of the travel time of the plan k calculated in Step S31 being , further calculate the cumulative gain decision weight function cumulative loss decision weight function The formula is:

[0112]

[0113] In the formula, ω + (p i ) is the passenger gain decision weight function, and ω - (p j ) is the passenger loss decision weight function; γ is the gain attitude coefficient, and δ is the loss attitude coefficient. In this embodiment, it is calibrated that γ = 0.61 and δ = 0.69; p n represents the probability of the occurrence of the nth gain travel state; p -m represents the probability of the occurrence of the mth loss travel state. Denote the objective probability corresponding to the travel time T as P = {p -m , …, p -1 , p0, p1, …, p n};

[0114] Step S42: According to the passenger value function g(y k ) calculated in Step S35 and the cumulative gain decision weight function cumulative loss decision weight function calculate the cumulative prospect value V(f i ) of the gain part further + and the cumulative prospect value V(f j ) of the loss part - . The formula is:

[0115]

[0116] Step S5: Use the impact data obtained in Step S2 and the cumulative prospect value V(f i ) of the gain part and the cumulative prospect value V(f + ) of the loss part obtained in Step S4 j to calibrate the parameter β corresponding to the z-th variable in the travel plan k - , and then calculate the cumulative prospect value V(f k ) corresponding to the special attribute under different plans and the probability P of the k-th travel plan selected by the passenger in the travel alternative plan set K k , so as to construct a passenger travel choice behavior model based on subway interruption. k In this embodiment, Step S5 specifically includes the following steps:

[0117] Step S51: Construct a calculation formula for the cumulative prospect value V(f

[0118] ) corresponding to the special attribute under different plans and the probability P of the k-th travel plan selected by the passenger in the travel alternative plan set K k : k The formula is:

[0119]

[0120] ν k = V(f k ) + ε k

[0121]

[0122] where X kz is the z-th variable in the utility of travel plan k, and β z is its corresponding coefficient; ν k is the choice utility of travel plan k, and ε k is the prospect error term, which follows an independent and identically distributed Gumbel distribution for each travel plan.

[0123] Step S52: Using the obtained impact data, calculate the cumulative prospect value V(f i ) + of the benefit part and the cumulative prospect value V(f j ) - of the loss part for each solution in the questionnaire, and the probability P k of selecting the k-th travel solution. Use the multiple regression method for parameter calibration to obtain the parameter values β z corresponding to the utility variables of each travel solution.

[0124] Step S53: According to the calculation formula constructed in Step S51, calculate the cumulative prospect value V(f k ) corresponding to the special attribute under different solutions and the probability P k that the passenger selects the k-th travel solution in the travel alternative set K, and complete the construction of the passenger travel choice behavior model based on subway interruption.

[0125] In this embodiment, the implementation process of constructing the passenger travel choice behavior model based on subway interruption is as Figure 3 shown.

[0126] Step S6: Using the passenger travel choice behavior model based on subway interruption constructed in Step S5, further establish a combined emergency bus transfer scheduling model that takes the travel cost F p of up and down passengers, the emergency bus operation cost T b based on time, and the emergency bus operation cost S b based on mileage as constraints to minimize the travel cost of passengers and the emergency bus operation cost considering the passenger travel choice behavior.

[0127] In this embodiment, the implementation flowchart of this step is as Figure 4 shown, and specifically includes the following steps:

[0128] Step S61: Determine the transfer scheduling mode as the combined emergency bus transfer scheduling mode; the mode is divided into two types of routes. One route is the whole journey route L1, that is, the emergency transfer bus stops at each interruption site; the other is the express route L2, which only stops at the interruption sites with large passenger flow.

[0129] Step S62: According to the probability P k that the passenger selects the k-th travel solution in the travel alternative set K calculated in Step S53, combined with the historical passenger flow OD data during the interruption duration of the subway sudden interruption section, calculate and obtain the number of people Q uv who need to be transferred from u to v in the interruption section.

[0130] Step S63: According to the number of people Q uvCalculate the waiting time T of passengers under combined emergency connection h , the running time T between stations zz , the waiting time T of large station passenger flow zd and the waiting time T of all-station bus passenger flow zq , and further calculate the total travel cost F of passengers under combined emergency bus connection p , the formula is:

[0131]

[0132] F p =μ1T h +μ2(T zz +T zd +T zq )

[0133] In the formula, f q is the departure frequency of the emergency bus on the all-station connection line, f d is the departure frequency of the emergency bus on the large station connection line, Q represents the total number of stranded people choosing the emergency connection bus line, x u , x v , x k represents the 0-1 variable indicating whether to stop at this station. When the emergency bus stops at this station, x u =x v =x k =1; t uv represents the running time of the emergency bus, which is obtained by dividing the interval driving distance by the average driving speed of the emergency bus, Q kv represents the number of people getting on the bus at station k on the all-station connection line, Q uk represents the number of people getting off the bus at station k on the all-station connection line; represents the average boarding time per passenger, t represents the average alighting time per passenger; μ1 represents the unit waiting time cost, and μ2 represents the unit in-vehicle time cost.

[0134] Step S64: Calculate the bus operation cost F under combined emergency bus connection b , which consists of two parts: the time-based bus operation cost T b and the mileage-based bus operation cost S b . The calculation formula is as follows:

[0135]

[0136] β d =f d / (f q +f d )

[0137] S b =(fq +f d )L

[0138] F b = ζ1T b + ζ2S b

[0139] Wherein, T o is the total operation time of the emergency bus; β d is the proportion of passengers choosing the large-stop transfer bus; L is the single-trip driving mileage of the emergency bus.

[0140] Step S65: According to the calculated total passenger travel cost F p and the emergency bus operation cost F b , and using the three constraints that the number of dispatched emergency transfer buses cannot exceed the maximum number of standby emergency transfer buses, the average full-load rate of the emergency bus is not less than 85% (set value in this embodiment), and the departure frequency should be greater than the specified minimum departure frequency and less than the maximum departure frequency, the minimum value of the combined emergency bus transfer scheduling total cost is obtained: min(δ1F p + δ2F b ), and the construction of the combined emergency bus transfer scheduling model considering the passenger travel choice behavior is completed.

[0141] Step S7: For the established combined emergency bus transfer scheduling model considering the passenger travel choice behavior, use the optimization algorithm to solve it, obtain the optimal solution of the decision variables of the scheduling model, and use the optimal solution as the control basis for the emergency bus departure frequency, the required number of buses, and the bus full-load rate during transfer, so as to complete the emergency bus transfer scheduling method considering the passenger travel choice behavior under the sudden interruption of the subway.

[0142] In this embodiment, the implementation flowchart of this step is as Figure 5 shown, and specifically includes the following steps:

[0143] Step S71: The emergency bus transfer scheduling model considering the passenger travel choice behavior encodes genes with travel alternative plans. After the population initialization, the fitness is calculated, the roulette wheel selection with elite retention, the crossover operation, and the mutation operation are performed in sequence, and then the next generation population is generated and the population is updated.

[0144] Step S72: Judge whether the total cost of the emergency bus transfer scheduling model considering the passenger travel choice behavior is the smallest. If so, output the best transfer scheduling plan; otherwise, go to step S73.

[0145] Step S73: If the total cost of the emergency bus transfer scheduling model considering passengers' travel choice behavior has not reached the minimum, continue to calculate the fitness, and determine whether the new fitness is less than the old fitness. If so, accept the new population according to the Metropolis criterion; otherwise, directly accept the new population.

[0146] Step S74: Generate the next generation population and update the population according to the accepted new population after roulette wheel selection, crossover operation, and mutation operation with elite retention until the minimum value of the total cost of the emergency bus transfer scheduling model is obtained, and output the optimal transfer scheduling plan.

[0147] In summary, the present invention provides an emergency bus transfer scheduling method considering passengers' travel choice behavior under subway sudden interruption. The emergency bus transfer scheduling method considering passengers' travel choice behavior under subway sudden interruption of the present invention obtains basic data through questionnaire surveys, calculates the cumulative prospect values corresponding to special attributes under different travel plans, constructs a passenger travel choice behavior model based on subway interruption and an emergency bus transfer scheduling model considering passengers' travel choice behavior, and uses an optimization algorithm to obtain the optimal solution of the decision variables of the scheduling model, providing an effective technical means for passenger emergency evacuation management under subway interruption; the design method of the present invention breaks through the limitation of the existing technical solution that the transfer demand is not clear, resulting in low transfer efficiency, improves the drawback that all attributes of each plan do not satisfy the mutual compensation relationship due to following the compensation principle in the existing technology, and makes up for the deficiencies of the existing technology in inaccurate description of transfer demand, low transfer efficiency, and low full load rate of transfer vehicles.

[0148] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0149] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for realizing in the process Figure 1 each process or multiple processes and / or blocks Figure 1a device for the functions specified in one or more boxes.

[0150] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device that implements the functions specified in one Figure 1 process or more processes and / or boxes Figure 1 a box or more boxes.

[0151] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 process or more processes and / or boxes Figure 1 a box or more boxes.

[0152] As described above, it is only the preferred embodiment of the present invention, and it is not a limitation to the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes. However, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution content of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. An emergency bus connection scheduling method considering passengers' travel choice behavior under sudden subway interruptions, characterized in that, Obtain the basic attribute values of travel alternative plans, the influence data of different sudden interruption scenarios on passengers' travel choice behaviors, the preset scenario variables, passenger attribute variables, and travel characteristic variables that affect travel choice behaviors, calculate the value function values, decision weight function values, and the cumulative prospect values corresponding to special attributes under different plans, construct a passenger travel choice behavior model based on subway interruptions, further establish an emergency bus connection scheduling model considering passengers' travel choice behaviors, and use an optimization algorithm to obtain the optimal solution of the decision variables of the scheduling model; the specific implementation steps are as follows: Step S1: Collect questionnaires to obtain the basic attribute values of travel alternative plans; The travel alternative plans at least include: waiting in place, standard connection buses, direct connection buses, skip-stop connection buses, regular buses, taxis / online car-hailing, and shared bicycles; The basic attribute values of the travel alternative at least include: the possible travel time of travel plan k between OD pair rs and the number of transfers required for travel plan k between OD pair rs; Step S2: Use the questionnaires collected in Step S1 to obtain the influence data of different sudden interruption scenarios on passengers' travel choice behaviors, and through data analysis, obtain the preset scenario variables, passenger attribute variables, and travel characteristic variables that affect passengers' travel choice behaviors after a subway sudden interruption event; The influence data at least includes: the number of people whose travel time for option k is , the number of people who choose option k under different preset scenarios, the number of people who choose option k under different passenger attributes, and the number of people who choose option k under different travel characteristics; The preset scenario variables at least include: origin and destination stations of travel, passenger status, travel purpose, event occurrence time period, and interruption duration; The passenger attribute variables at least include: gender, age group, education level, monthly income, occupation, and monthly working hours; The travel characteristic variables at least include: passengers' travel choice behaviors during normal subway operation, subway interruption experience, information attention level, and risk travel attitude; Step S3: Using the impact data obtained in Step S2, further obtain the set K of alternative travel options selected by passengers, and calculate the average working time O of passengers k , the average monthly income S of passengers k , the objective occurrence probability p that the travel time of option k is , calculate the reference point time of the passenger time cost coefficient θ k , the profit and loss value y of the passenger's choice of travel option k k , and then calculate the passenger value function g(y k ), the passenger revenue decision weight function ω + (p i ), the passenger loss decision weight function ω - (p j ); Step S4: Using the passenger value function g(y k ), the passenger revenue decision weight function ω + (p i ), the passenger loss decision weight function ω - (p j ), calculate the cumulative revenue decision weight function the cumulative loss decision weight function Further calculate the cumulative prospect value V(f i ) + of the revenue part, and the cumulative prospect value V(f j ) - ; Step S5: Using the impact data obtained in Step S2 and the cumulative prospect value V(f i ) + of the profit part and the cumulative prospect value V(f j ) - of the loss part obtained in Step S4, calibrate the parameter β k corresponding to the z-th variable in the travel plan k, and then calculate the cumulative prospect value V(f k ) corresponding to the special attribute under different plans and the probability P k of the k-th travel plan selected by passengers from the set K of travel alternative plans, so as to construct a passenger travel choice behavior model based on subway interruption; Step S6: Using the passenger travel choice behavior model based on subway interruptions constructed in Step S5, further establish the combined emergency bus transfer scheduling model that takes into account passenger travel choice behavior with the up and down passenger travel cost F p , the emergency bus operation cost T based on time b , the emergency bus operation cost S based on mileage b as constraints to minimize the combined emergency bus transfer scheduling model of passenger travel cost and emergency bus operation cost considering passenger travel choice behavior; Step S7: For the established combined emergency bus connection scheduling model considering passengers' travel choice behaviors, use an optimization algorithm to solve it to obtain the optimal solution of the decision variables of the scheduling model, and use the optimal solution as the basis for controlling the emergency bus departure frequency, required bus vehicles, and bus load factor during connection, so as to complete the emergency bus connection scheduling method considering passengers' travel choice behaviors under subway sudden interruptions; The specific steps of Step S3 include the following steps: Step S31: Based on the acquired impact data, construct a set of passenger travel alternatives K = {K1, K2, K3, K4, K5, K6, K7}, where K1 is to choose to wait for rescue on the spot when the subway is affected by a sudden interruption; K2 is to choose standard shuttle bus when the subway is affected by a sudden interruption; K3 is to choose direct bus when the subway is affected by a sudden interruption; K4 is to choose interval bus when the subway is affected by a sudden interruption; K5 is to choose skip-stop bus when the subway is affected by a sudden interruption; K6 is to choose online car-hailing when the subway is affected by a sudden interruption; K7 is to choose shared bicycle when the subway is affected by a sudden interruption; calculate the average working time O of the 7 passengers who choose the travel alternatives k , Average monthly income of passengers S k , the travel time of plan k is The objective probability of occurrence p; Step S32: Based on the average working time O of passengers calculated in Step S31 k , the average monthly income S of passengers k , further calculate the passenger time cost coefficient θ k , and the formula is: Step S33: Calculate the passenger reference point time according to the passenger time cost coefficient θ k , The formula is: Wherein, is the possible travel time of travel plan k obtained in step S1 between OD pair rs; Step S34: Based on the calculated passenger reference point time and the possible travel time of travel plan k between OD pair rs calculate the profit and loss value y of the passenger choosing travel plan k k , and the formula is: In the formula, considering that there are m + n + 1 possible travel times for the k-th travel plan under uncertain conditions, denoted as Sorted in descending order as λ is the function risk aversion coefficient; Step S35: Based on the profit and loss value y of the travel plan k selected by the passenger calculated in Step S34 k , further calculate the passenger value function g(y k ), and the formula is: In the formula, α and β are the risk preference coefficients of the function, where α represents the degree of risk aversion when there is a gain, and β represents the degree of risk preference when there is a loss; The specific steps of Step S4 include the following steps: Step S41: Based on the travel time of Plan k calculated in Step S31, which is the objective occurrence probability p, further calculate the cumulative gain decision weight function the cumulative loss decision weight function The formula is: where ω + (p i ) is the decision-making weight function for passenger benefits, and ω - (p j ) is the decision-making weight function for passenger losses; γ is the benefit attitude coefficient, and δ is the loss attitude coefficient; p n represents the probability of the nth benefit travel state occurring; p -m represents the probability of the mth loss travel state occurring. Denote the objective probability of the travel time T corresponding to its occurrence as P = {p -m , …, p -1 , p0, p1, …, p n}; Step S42: According to the passenger value function g(y k ) calculated in step S35 and the cumulative gain decision weight function cumulative loss decision weight function calculate the cumulative prospect value V(f i ) + of the gain part and the cumulative prospect value V(f j ) - of the loss part further. The formula is as follows: The specific steps of Step S5 include the following steps: Step S51: Construct the calculation formula for the cumulative foreground value V(f k ) corresponding to the special attribute under different scenarios and the probability P k of the k-th travel option in the set K of passenger travel alternative options: where X kz is the z-th variable in the utility of travel plan k, and β z is its corresponding coefficient; ν k is the choice utility of travel plan k, and ε k is the prospect error term, which follows an independent and identically distributed Gumbel distribution for each travel plan; Step S52: Using the obtained impact data, calculate the cumulative prospect value V(f i ) + of the benefit part for each solution in the questionnaire, the cumulative prospect value V(f j ) - of the loss part, and the probability P k of selecting the k-th travel solution. Then, use the multiple regression method for parameter calibration to obtain the parameter values β z corresponding to the utility variables of each travel solution; Step S53: Calculate the cumulative foreground value V(f k ) corresponding to the special attribute under different scenarios and the probability P k of the k-th travel option in the set K of alternative travel options selected by passengers, and complete the construction of the passenger travel choice behavior model based on subway interruptions; The specific steps of Step S6 include the following steps: Step S61: Determine the connection scheduling mode as the combined emergency bus connection scheduling mode; the mode is divided into two types of routes. One route is the whole journey route L1, that is, the emergency connection bus stops at each interruption site; the other is the large station route L2, which only stops at interruption sites with large passenger flows; Step S62: According to the probability P of the k-th travel plan in the set K of travel alternative plans selected by passengers calculated in step S53, combined with the historical passenger flow OD data during the interruption duration of the subway sudden interruption section, calculate the number of people Q who need to transfer from u to v within the interruption section k , and calculate the number of people Q who need to transfer from u to v within the interruption section uv ; Step S63: Calculate the waiting time of passengers under the combined emergency connection according to the number of passengers Q to be connected from u to v within the interruption interval uv Calculate the waiting time of passengers under the combined emergency connection T h 、Running time T between stations zz 、Waiting time T of passengers at large stations zd and waiting time T of through - bus passengers zq , and further calculate the total travel cost F of passengers under the combined emergency bus transfer p , and the formula is: F p = μ1T h + μ2(T zz + T zd + T zq ) In the formula, f q is the departure frequency of emergency buses on the entire connecting route, f d is the departure frequency of emergency buses on major station shuttle routes, Q represents the total number of stranded people who choose emergency shuttle bus routes, and x u 、x v 、x k A 0-1 variable indicating whether the emergency bus stops at the station. When the emergency bus stops at the station, x u =x v =x k =1; t uv represents the emergency bus running time, which is obtained by dividing the interval travel distance by the average travel speed of the emergency bus, Q kv represents the number of passengers boarding the shuttle line at station k, Q uk represents the number of passengers getting off at station k on the entire station connecting line; It represents the average time for a passenger to board the bus. t represents the average time for a passenger to get off the bus; μ1 represents the unit waiting time cost, and μ2 represents the unit time cost on the bus; Step S64: Calculate the bus operation cost F under the combined emergency bus connection b , which consists of the time-based bus operation cost T b and the mileage-based bus operation cost S b . The calculation formula is as follows: β d = f d / (f q + f d ) S b = (f q + f d )L F b = ζ1T b + ζ2S b where T o is the total operation time of the emergency bus; β d is the proportion of passengers choosing the large-stop transfer bus; L is the one-way driving mileage of the emergency bus; Step S65: According to the calculated total passenger travel cost F p and the emergency bus operation cost F b , and by using the three constraints that the number of dispatched emergency shuttle buses cannot exceed the maximum number of standby emergency shuttle buses, the average full-load rate of emergency buses should be not less than the set value, and the departure frequency should be greater than the specified minimum departure frequency and less than the maximum departure frequency, the minimum value of the combined emergency bus shuttle scheduling total cost is obtained: min(δ1F p +δ2F b ), and the construction of the combined emergency bus shuttle scheduling model considering passengers' travel choice behavior is completed; The specific steps of Step S7 include the following steps: Step S71: Perform gene coding on the emergency bus connection scheduling model considering passengers' travel choice behaviors. After initializing the population, calculate the fitness, roulette wheel selection with elite retention, crossover operation, and mutation operation in sequence to generate the next generation population and update the population; Step S72: Judge whether the total cost of the emergency bus connection scheduling model considering passengers' travel choice behaviors is the smallest. If so, output the best connection scheduling plan; otherwise, go to Step S73; Step S73: If the total cost of the emergency bus transfer scheduling model considering passengers' travel choice behavior has not reached the minimum, continue to calculate the fitness, and determine whether the new fitness is less than the old fitness. If so, accept the new population according to the Metropolis criterion; otherwise, directly accept the new population. Step S74: Generate the next generation population and update the population according to the accepted new population after roulette wheel selection with elite retention, crossover operation, and mutation operation until the minimum value of the total cost of the emergency bus transfer scheduling model is obtained, and output the optimal transfer scheduling plan.

2. The emergency bus connection scheduling method considering passengers' travel choice behavior under sudden subway interruptions according to claim 1, wherein The specific steps of step S2 are as follows: Step S21: Calculate the probability values of each subclass selection scheme k for the impact data of passengers' travel choice behavior selected under different sudden interruption scenarios: the probability value of selection scheme k under different preset scenarios, the probability value of selection scheme k under different passenger attributes, and the probability value of selection scheme k under different travel characteristics. Step S22: Use the calculated probability values of each subclass selection scheme k to draw the distribution ratio diagram of each subclass travel scheme k. Step S23: Use the drawn distribution ratio diagram of each subclass travel scheme k to compare and obtain the preset scenario variables, passenger attribute variables, and travel characteristic variables that affect passengers' travel choice behavior.

Images