Bus Intelligent Dynamic Scheduling Optimization Method Based on Real-Time Data

Abstract

This application belongs to the field of bus dispatching technology, and specifically relates to a bus intelligent dynamic dispatching optimization method based on real-time data. Through an improved minimum-cost maximum-flow model, this method simultaneously optimizes the load factor and the generalized transportation cost. By means of collaborative optimization, weights are assigned to each objective. Through numerical experiments, the weights for optimally improving bus operation efficiency are selected. A model is established based on the multi-objective optimization method and solved through a genetic algorithm to achieve the purpose of dynamically adjusting the departure interval and vehicle type selection, so as to efficiently allocate resources and significantly reduce the passenger waiting time.

Description

Technical Field

[0001] This application belongs to the technical field of bus dispatching, and specifically relates to a bus intelligent dynamic dispatching optimization method based on real-time data. Background Art

[0002] With the rapid growth of the global population and the acceleration of the urbanization process, urban traffic problems have become increasingly severe. Especially during peak hours, traffic congestion has become an important factor affecting the travel efficiency of urban residents, the quality of public transportation services, and energy consumption. At the same time, environmental pollution and traffic safety problems have also become increasingly serious, and the pressure on urban traffic management has been continuously increasing. Most traditional traffic management models rely on fixed departure schedules and simple traffic flow predictions. These methods are difficult to cope with complex traffic dynamic changes, especially in the case of emergencies or unpredictable traffic flow fluctuations, often resulting in waste of traffic resources or failure to meet passenger needs.

[0003] In this context, a public transportation intelligent dispatching optimization method based on real-time data has emerged. The intelligent transportation system can monitor and analyze traffic flow in real time by applying advanced technologies such as the Internet of Things, big data, and artificial intelligence, and can dynamically adjust traffic signals and departure schedules to achieve a more efficient, greener, and safer urban transportation system. Therefore, as a new idea, the stochastic optimization model based on real-time data can more precisely address these challenges, improve the overall efficiency of the traffic system and resource utilization rate, and provide theoretical support and technical guarantee for the further development of the intelligent transportation system. Summary of the Invention

[0004] In this invention, the bus intelligent dynamic dispatching method mainly relies on real-time data. The main research idea is as follows: Based on the collected real-time data, the peak and trough periods of passenger flow in a day are divided. The traffic flow within the road section is obtained through real-time GPS monitoring data of the road to judge the degree of road congestion, so as to predict the arrival time of the bus. In addition, multi-objective optimization is also carried out based on passenger flow data to achieve the purpose of improving vehicle operation efficiency and passenger satisfaction by selecting the vehicle type and departure interval. The specific technical solution is as follows:

[0005] A bus intelligent dynamic dispatching optimization method based on real-time data, comprising the following steps:

[0006] S1. The total cost of passenger travel includes the passenger waiting time and the cost spent by the passenger. Construct the objective function of the total cost of passenger travel:

[0007] ;

[0008] ;

[0009] In the formula, is the total cost of passenger travel; is the waiting time for passengers, is a dimensionless function of is the monetary cost, is a dimensionless function of is the bus operation cost, is the cost incurred by passengers;

[0010] S2: Taking the maximum vehicle load factor as the objective function and considering the vehicle type and headway, the specific objective function is as follows:

[0011] ;

[0012] In the formula, is the vehicle load factor, is the vehicle capacity corresponding to different vehicle types, is an index variable representing whether vehicle type is selected when the line , is the passenger volume borne by the station to the station ; is the total number of stations;

[0013] S3: Using the genetic algorithm to search for the optimal solution of the objective functions in S1 and S2, and finding the optimal solution through survival of the fittest; the total objective function is:

[0014] ;

[0015] In the formula, represent the weights of the sub-objectives respectively, satisfying .

[0016] Preferably, in step S1, the time cost of the passengers waiting for train at station in the upward route is expressed by the following formula:

[0017] ;

[0018] The time cost of the passengers waiting for train at station in the downward route is expressed by the following formula:

[0019]

[0020] ;

[0021] In the formula, is the waiting time for all passengers, represents the upward direction, represents the downward direction, is the arrival rate of passengers waiting on the upward line, is the arrival rate of passengers waiting on the downward line, which is a function of ; represents the trips in the upward direction departing from the station at the moment of departure, represents the trips in the downward direction departing from the station at the moment of departure, represents the set of all upward trips, represents the set of all downward trips, represents the set of all upward stations, represents the set of all downward stations.

[0022] Preferably, the transportation cost and the passenger cost are respectively:

[0023] ;

[0024] ;

[0025] In the formula, is the transportation cost, is the passenger cost, is a 0-1 variable for judging whether the line is the line with the lowest bus operation cost. When the th line is the line with the lowest operation cost from the station to the station , is 1, otherwise it is 0; is the transportation cost spent on taking the line from the station to the station . In particular, when , let , and when and there is no direct line between them, ; is the average passenger cost spent on taking the line from the station to the station ; is the passenger flow volume borne by the section between the station and the station .

[0026] Preferably, the relationship equation between the vehicle travel time between stations and the arrival time at adjacent stations is as follows:

[0027] ;

[0028] ;

[0029] To avoid repeated route selection, the constraints are as follows:

[0030] ;

[0031] In the formula, and are respectively functions of the travel time of the vehicle on the sections between adjacent stations in the up and down directions.

[0032] Preferably, in step S2, a vehicle type selection constraint function is established, and its formula is as follows:

[0033] ;

[0034] In the formula, is the index variable for selecting vehicle type for line , which is selected as 1 if selected and 0 if not selected.

[0035] Preferably, in step S2, the passenger flow carried should satisfy:

[0036] ;

[0037] In the formula, is the passenger flow from station to station , which is a non - negative quantity.

[0038] Preferably, each individual in step S1 is sequentially coded. For the up direction:

[0039] ;

[0040] Among them, is the departure interval between adjacent trips, , is the set of trips in the up direction;

[0041] For the down direction:

[0042] ;

[0043] Among them, is the departure interval between adjacent trips, , It is the set of trips in the downward direction;

[0044] The departure time plan for the bus lines at the double - station is the set of the upward - direction and downward - direction plans, expressed as:

[0045] ;

[0046] An evaluation function is established to assign probabilities to each chromosome, and its formula is as follows:

[0047] , ;

[0048] Wherein, is any chromosome, is a real number, , is the population size.

[0049] Preferably, in step S3, the roulette - wheel method is used to select chromosomes for cultivating the new generation, and its steps are as follows:

[0050] Step 1: Calculate the cumulative probability of each chromosome , that is:

[0051] ;

[0052] ;

[0053] In the formula, is any chromosome except chromosome ;

[0054] Step 2: Generate a random number , if , then chromosome is selected. Repeat this step times, then chromosomes are obtained.

[0055] Preferably, in step S3, a chromosome is randomly selected as the parent. Assuming the crossover probability is , a random number is generated from . If , for the part of the chromosome belonging to the upward direction, two non - overlapping gene segments with equal lengths are randomly selected for exchange. The operation for the downward - direction chromosome is the same, and a new population is to be generated.

[0056] Preferably, in step S3, assuming the mutation probability is , from Randomly generated numbers If For the part of the chromosome that belongs to the upward direction, replace the gene segment of length with a newly generated gene segment of the same length; randomly generate a gene segment of length in the interval The length of the interval is:

[0057] ;

[0058] respectively represent the chromosomes at both ends of the gene covering the gene segment.

[0059] Compared with the prior art, the beneficial effects of this application are as follows:

[0060] 1. Traditional methods usually fixedly use one type of vehicle and do not consider flexibly selecting the vehicle type according to the real-time passenger flow or the needs of different sections. This method lacks flexibility and the ability to cope with changing traffic conditions. During peak hours, the fixed vehicle type may be overloaded, and during off-peak hours, there may be empty vehicles running, thus affecting the overall efficiency. This application introduces a strategy of mixed departure of multiple vehicle types and dynamically selects a suitable vehicle type according to real-time passenger flow data. Large-capacity vehicle types can be selected during peak hours, while small vehicles can be used during off-peak hours, minimizing unnecessary resource waste to the greatest extent, while improving the vehicle utilization rate and operation efficiency. This not only improves the service quality but also optimizes the transportation cost.

[0061] 2. Traditional scheduling optimization methods mostly rely on historical data for modeling and optimization. Historical data provides information such as past traffic patterns and passenger flows and can predict the passenger flow in each period. However, its limitation is that the data cannot reflect the immediately changing traffic environment. For example, factors such as special weather and traffic accidents may cause sudden changes in traffic flow, and historical data cannot respond to this change in a timely manner. This patent can monitor variables such as traffic flow and passenger demand in real time by introducing real-time data and adjust the scheduling strategy accordingly to ensure the high efficiency and flexibility of transportation.

[0062] 3. In previous studies, optimization algorithms usually solve problems based on fixed conditions and assumptions and adopt a global optimization strategy. The algorithm usually performs scheduling optimization within a relatively large time range, considering the operation of the entire transportation system, but may not be able to quickly obtain the optimal solution in a short time and often requires a large amount of computing resources to process large-scale data. Through a dynamic adjustment and real-time feedback mechanism in this application, the optimization process can be quickly carried out within a smaller time range and the optimization target can be adjusted in real time. Through this method, the system can significantly improve the computing efficiency while ensuring the optimization quality and adapt to the changing traffic conditions. Brief Description of the Drawings

[0063] Figure 1 is a flow chart,

[0064] Figure 2 is a flow chart of a genetic algorithm,

[0065] Figure 3 is a schematic diagram of the line operation of a bus double - terminal station,

[0066] Figure 4 is the optimization result diagram of the departure interval at each time period. Specific implementation manners

[0067] The technical solution of the present application will be described in detail below through specific embodiments and the accompanying drawings. It should be understood that the specific features in the embodiments of the present application are detailed descriptions of the technical solution of the present application, rather than limitations on the technical solution of the present application. The specific technical features can be combined with each other.

[0068] A bus intelligent dynamic scheduling optimization method based on real - time data includes the following steps:

[0069] By establishing a traffic flow model, according to the traffic flow monitored by real - time data, judge the traffic congestion degree to perform vehicle scheduling.

[0070] Consider the one - way flow of cars on a section of road. To simplify the model, assume that the length of this section of road is much longer than the distance between vehicles, the length of each vehicle is the same, and the traffic flow is regarded as a continuous quantity.

[0071] Take the axis in the traffic flow direction of the road, and the car moves along the positive direction of the axis. According to the number of vehicles in the section, the change in the total number of vehicles in the time period is equal to the number of vehicles entering this area minus the number of vehicles leaving this area, and we get:

[0072] ;

[0073] In the formula, is the function of traffic flow density (the number of vehicles per unit road length), is the traffic flow of the vehicle.

[0074] Divide both ends of the formula by and after arrangement, we get:

[0075] ;

[0076] Take the limit of both ends simultaneously to get:

[0077] ;

[0078] From Obtained:

[0079] ;

[0080] Starting from the characteristics of the traffic flow, the relationship model between density and speed is established as:

[0081] ;

[0082] In the formula, is a function of density. When the density is small to 0, it reaches the maximum speed limit of the road. Therefore, is the maximum speed limit of the road. is the maximum traffic flow density of the road.

[0083] Accordingly, based on the real-time monitored road density, the vehicle speed can be obtained, and then the vehicle arrival time can be predicted.

[0084] The common bus line is a double-terminal station bus line, that is, a line with an up-line station and a down-line station. At this time, the waiting time of passengers should be the sum of the waiting time of passengers in the up-line direction and the waiting time of passengers in the down-line direction. Let the arrival rate of passengers at each station in the up and down directions be a function of time , and represent, which means that the passenger arrival rate is a continuous function of time, that is, for any given time, the passenger arrival rate of the corresponding station can be obtained. For the up-line direction, let the train arrive at the station at the time of , , , then the arrival time interval between two adjacent trains is . Therefore, the waiting time of passengers taking the train at the station is:

[0085] ;

[0086] The total waiting time of all passengers at all stations in the up-line direction for all train trips is:

[0087] ;

[0088] Similarly, the total waiting time of all passengers at all stations in the down-line direction for all train trips is:

[0089] ;

[0090] S1. The total travel cost of passengers includes the waiting time of passengers and the cost spent by passengers. Construct the objective function of the total travel cost of passengers:

[0091] ;

[0092] ;

[0093] The normalization process is as follows:

[0094] ;

[0095] ;

[0096] ;

[0097] ;

[0098] Among them, is the minimum waiting time for passengers, with a minimum of 0, is the maximum waiting time for passengers, which is determined according to the actual departure interval data. is the minimum cost, which is the minimum when no bus departs, with a minimum of 0, is the maximum cost, which is determined according to the actual number of departures.

[0099] For the passengers waiting at station for train on the upward route, the time cost is expressed by the following formula:

[0100] ;

[0101] For the passengers waiting at station for train on the downward route, the time cost is expressed by the following formula:

[0102] ;

[0103] ;

[0104] In the formula, is the waiting duration of all passengers, represents the upward direction, represents the downward direction, is the arrival rate of passengers waiting on the upward route, is the arrival rate of passengers waiting on the downward route, which is a function of , represents the departure time of the th trip in the upward direction from station , represents the departure time of the th trip in the downward direction from station . represents the set of all upward trips, represents the set of all downward trips, represents the set of all upward stations, represents the set of all downward stations, is the total travel cost for passengers; The magnitude of is determined by the passenger waiting time and the passenger travel cost, is the passenger waiting time, is the monetary cost, is the bus operation cost, is the passenger spending cost.

[0105] Transportation cost and passenger spending are respectively:

[0106] ;

[0107] ;

[0108] In the formula, is the transportation cost, is the passenger spending cost, is a 0-1 variable for judging whether the line is the line with the lowest bus operation cost. When the th line is the line with the lowest operation cost from station to station , is 1, otherwise it is 0; is the transportation cost spent on taking line from station to station . In particular, when , let . When there is no direct line between and , ; is the average spending cost of passengers taking line from station to station ; is the passenger flow volume borne by the section between station and station .

[0109] Generally speaking, during the operation of a line, after passing through the passing stations, the remaining passenger capacity on the vehicle will change. Denote the vehicle when running between station and station , the current capacity of the bus is There are three possible initial vehicle capacities (e.g., small vehicles with a passenger capacity of less than 20 people, medium-sized vehicles with a capacity of 20 - 50 people, and large vehicles with a capacity of more than 50 people), denoted as ( ).

[0110] When passing through a station, the vehicle capacity is updated as:

[0111] ;

[0112] In the formula, is the time interval during which the number of passengers waiting at station boarding the vehicle is, is the time interval during which the number of passengers boarding vehicle passing through station and getting off is.

[0113] S2: Taking the maximum vehicle load factor as the objective function, considering the vehicle type and departure interval, the specific objective function is as follows:

[0114] ;

[0115] In the formula, is the vehicle load factor, is the vehicle capacity corresponding to different vehicle types, is the index vector, representing whether vehicle type is selected when the line departs, is the passenger flow volume borne by station to station ; is the total number of stations;

[0116] The relationship equation between the vehicle running time between stations and the arrival time at adjacent stations is as follows:

[0117] ;

[0118] ;

[0119] To avoid repeated route selection, the constraints are as follows:

[0120] ;

[0121] In the formula, and are respectively functions of the running time of each section of the vehicle between adjacent stations in the up and down directions.

[0122] Establish a vehicle type selection constraint function, and its formula is as follows:

[0123] ;

[0124] In the formula, is the riding route Select vehicle type is the index variable. If selected, it is 1; if not selected, it is 0.

[0125] In step S2, the passenger flow borne should satisfy:

[0126] ;

[0127] In the formula, is the passenger flow from station to station , and this is a non - negative quantity.

[0128] S3: Use the genetic algorithm to search for the optimal solution of the objective function in S1 and S2, and find the optimal solution through survival of the fittest; the total objective function is:

[0129] ;

[0130] respectively represent the weights of sub - objectives, and satisfy .

[0131] Perform sequential coding on each individual in step S1. For the up - bound direction:

[0132] ;

[0133] Among them, is the departure interval between adjacent trips, , is the set of trips in the up - bound direction;

[0134] For the down - bound direction:

[0135] ;

[0136] Among them, is the departure interval between adjacent trips, , is the set of trips in the down - bound direction;

[0137] The departure time plan for the bus line with double - terminal stations is the set of the up - bound direction and down - bound direction plans, expressed as:

[0138] ;

[0139] An evaluation function is established to assign probabilities to each chromosome, and the formula is as follows:

[0140] , ;

[0141] where, is any chromosome, that is, any number in the set of the upstream direction and downstream direction schemes, is a real number, , is the population size.

[0142] In step S3, the roulette wheel method is used to select chromosomes for cultivating a new generation, and the steps are as follows:

[0143] Step 1: Calculate the cumulative probability of each chromosome , that is:

[0144] ;

[0145] ;

[0146] In the formula, is any chromosome except chromosome .

[0147] Step 2: Generate a random number , if , then chromosome is selected, and this step is repeated times, then chromosomes are obtained.

[0148] In step S3, a chromosome is randomly selected as the parent. Assuming the crossover probability is , a number is randomly generated from . If , for the part of the chromosome belonging to the upstream direction, two non-overlapping gene segments of equal length are randomly selected for exchange. The same operation is performed on the downstream direction chromosome to generate a new population.

[0149] In step S3, assuming the mutation probability is , a number is randomly generated from . If , for the part of the chromosome belonging to the upstream direction, a gene segment with a length of is replaced with a newly generated gene segment of the same length; a gene segment with a length of is randomly generated in the interval The interval length is:

[0150] ;

[0151] They cover The gene fragments are located on the chromosomes at both ends of the gene.

[0152] In step S3 of the present invention, preferably, the following assumptions are made:

[0153] 1. When the target bus arrives, all passengers will choose to get on the bus, that is, the bus capacity can be greater than 100%;

[0154] 2. The vehicle stays at the station for a fixed time and does not fluctuate greatly with the number of passengers getting on and off the vehicle;

[0155] 3. The fixed costs (such as driver wages, vehicle maintenance, etc.) and variable costs (such as fuel consumption) of each trip are simplified to fixed values ​​for each trip;

[0156] 4. Assume that the load factor is proportional to the number of passengers on each trip;

[0157] 5. Maximize total revenue (assuming each passenger pays a fixed fare) minus total cost, which is maximum profit.

[0158] Based on the passenger flow data collected by S1, mainly including the number of people waiting and the waiting time in each time period at each station, the genetic algorithm is used to solve the problem. The principle of the genetic algorithm is to simulate the natural evolution process in the biological world and find the optimal solution through the survival of the fittest. Due to the evolutionary characteristics of the algorithm, the genetic algorithm imitates the process of biological evolution in nature, including initializing the population, calculating fitness, chromosome selection, chromosome crossover, and chromosome mutation. The genetic algorithm is robust in finding high-quality non-dominated solutions and running time. When applying the genetic algorithm, an initial population is first generated. The population is composed of chromosomes. Each chromosome is a feasible chromosome generated according to certain rules, which is the initial feasible solution. Then, according to the principle of survival of the fittest, the chromosomes in the initial population are selected, crossed, mutated, and other operations are continuously iterated until a solution close to the optimal is generated. According to the objective function of the problem to be solved, an evaluation function is designed to evaluate each generation of the population, that is, chromosomes with higher fitness are usually selected, thereby generating a new population. The iterative process is a process to ensure that high-quality chromosomes are left behind. Through iteration, the quality of the population is gradually improved. The steps are as follows: Figure 1 shown.

[0159] The present invention can improve the efficiency of public transportation operations:

[0160] 1. By collecting and processing passenger flow data, analyzing real-time traffic and passenger flow data, and making precise arrangements for vehicle departure intervals and vehicle type selections according to different time periods (peak periods, off-peak periods, low-peak periods), the reasonable dispatching of vehicles is achieved. This avoids the problem that fixed timetables cannot adapt to dynamic traffic changes, thereby reducing resource waste and unnecessary empty vehicle operations.

[0161] 2. Select appropriate vehicle types according to passenger flow density to avoid overcrowding caused by too small vehicles and resource waste caused by too large vehicles, and improve the full load rate.

[0162] Through multi-objective optimization and genetic algorithms, the present invention reasonably arranges departure times, reducing the average waiting time of passengers at stations. By dynamically adjusting routes and intervals and establishing a traffic flow model, vehicle delays caused by congestion are alleviated, improving the travel experience of passengers.

[0163] The present invention enhances the flexibility and adaptability of the system. Through real-time data monitoring (such as GPS and card-swipe data) and a traffic density prediction model, the system can quickly respond to emergencies or unpredictable traffic flow fluctuations, ensuring the reliability and stability of the bus system.

[0164] The present invention reduces operating costs:

[0165] 1. Through the minimum cost maximum flow model and full load rate optimization, reasonably select vehicle types and departure frequencies, avoiding the operation of too many empty vehicles, thereby reducing operating costs such as fuel and maintenance.

[0166] 2. Guide passengers to choose appropriate routes, optimize the passenger flow distribution among routes, and further reduce operating costs.

[0167] Based on multi-objective optimization of dynamic bus departures, using a traffic flow model and real-time data, the system can flexibly respond to emergencies or abnormal traffic flows (such as sudden traffic jams or major events), ensuring the continuity of bus services. Based on the Greenshields speed-density relationship model, the traffic flow and congestion conditions on sections are predicted, providing accurate basis for vehicle dispatching.

[0168] Based on multi-objective optimization of dynamic bus departures, this application has good scalability and applicability. The system is based on real-time data collection and existing algorithms, facilitating expansion to bus networks or cities of different scales. Data can be collected and suggestions can be provided through the passenger APP, further enhancing the promotion of the system and public participation.

[0169] Based on multi-objective optimization of dynamic bus departures, in practical applications, it can achieve efficient traffic management, reduce resource waste, and provide strong support for building a smart city.

[0170] The above are only the preferred embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the technical principle of the present application, several improvements and modifications can be made, and these improvements and modifications should also be regarded as the protection scope of the present application.

Claims

1. A bus intelligent dynamic scheduling optimization method based on real-time data, characterized in that: The following steps are involved: S1. The total cost of passenger travel includes the waiting time of passengers and the cost of passengers. The objective function of the total cost of passenger travel is constructed as follows: ； ； In the formula, is the total cost of passenger travel; Waiting time for passengers, for The dimensionless function of For monetary cost, for The dimensionless function of is the bus operating cost, Costs to passengers; S2: Taking the maximum vehicle load rate as the objective function, considering the vehicle type and departure interval, the objective function is as follows: ； Establish the vehicle model selection constraint function, which is as follows: ； In the formula, For ride routes Select Model The indicator variable is 1 if selected and 0 if not selected; is the vehicle load factor, is the vehicle capacity corresponding to different models, For Site To site The passenger flow capacity; is the total number of sites; S3: Use genetic algorithm to search for the optimal solution of the objective function in S1 and S2, and find the optimal solution through survival of the fittest; the overall objective function is: ； In the formula, Represent the sub-goal weights respectively, satisfying ; Sequentially encode each individual in step S1, for the uplink direction: ； in, The interval between adjacent trains. , The trip collection for the upward direction; For the downlink direction: ； in, The interval between adjacent trains. , It is the trip collection for the down direction; The departure time plan of the dual-station bus line is a combination of the up-direction and down-direction plans, expressed as: ； Establish an evaluation function to assign probability to each chromosome, the formula is as follows: , ； in, For any chromosome, is a real number, , For population size.

2. The method for optimizing bus intelligent dynamic scheduling based on real-time data according to claim 1 is characterized in that: In step S1, at the station in the up route Waiting for the train Passenger time cost , expressed as follows: ； On the down route at the station Waiting for the train Passenger time cost , expressed as follows: ； ； In the formula, Waiting time for all passengers, Represents the upward direction, Represents the downward direction, is the arrival rate of passengers waiting for the uplink route, is the arrival rate of waiting passengers on the downlink line, Indicates the number of trips in the upward direction From the site The time of departure, Indicates the number of trips in the down direction From the site The time of departure, represents the set of all upstream trips, represents the set of all downstream trips, represents the set of all upstream sites, Represents the set of all downstream sites.

3. The method for optimizing public transportation intelligent dynamic scheduling based on real-time data according to claim 1 is characterized in that: Shipping costs and passenger costs They are: ； ； In the formula, For transportation costs, Costs for passengers, To determine the line Is it a 0-1 variable indicating the minimum bus operating cost? The line is from the station To site When the line with the lowest operating cost is is 1, otherwise it is 0; From site To site Take the route between The transportation cost is season ,when and When there is no direct line between ; For ride routes From the site To site The average cost of passengers It is a site To site The passenger flow capacity of the road sections between them.

4. The method for optimizing bus intelligent dynamic scheduling based on real-time data according to claim 2 is characterized in that: The relationship between the vehicle travel time between stations and the time of arrival at adjacent stations is as follows: ； ； To avoid repeated route selection, the constraints are as follows: ； In the formula, and They are functions of the vehicle's travel time on the road section between each adjacent station in the uplink and downlink directions respectively.

5. The method for optimizing bus intelligent dynamic scheduling based on real-time data according to claim 3 is characterized in that: In step S2, the passenger flow capacity should meet the following requirements: ； In the formula, From site To site passenger flow.

6. The method for optimizing public transportation intelligent dynamic scheduling based on real-time data according to claim 1 is characterized in that: In step S3, a roulette wheel method is used to select and cultivate a new generation of chromosomes, and the steps are as follows: Step 1: Calculate the cumulative probability of each chromosome ,Right now: ； ； In the formula, In addition to chromosomes Any chromosome other than Step 2: Generate a random number ,like , then chromosome is selected, repeat the step times, we get chromosomes.

7. The method for optimizing bus intelligent dynamic scheduling based on real-time data according to claim 1 is characterized in that: In step S3, a chromosome is randomly selected Assume that the crossover probability is ,from Randomly generated numbers ,if For the part of the chromosome that belongs to the upward direction, two non-overlapping gene fragments of equal length are randomly selected for exchange. The operation of the chromosome in the downward direction is similar to that to generate a new population.

8. The method for optimizing public transportation intelligent dynamic scheduling based on real-time data according to claim 1 is characterized in that: In step S3, assume that the mutation probability is ,from Randomly generated numbers ,if , for the part of the chromosome that is in the ascending direction, the length will be The gene fragment of is replaced with a newly generated gene fragment of the same length; a randomly generated gene fragment of length The gene fragment is in the interval The interval length is ; They cover The gene fragments are located on the chromosomes at both ends of the gene.

Images