Dynamic passenger flow-oriented urban rail train departure timetable energy-saving optimization method

Abstract

The invention provides a dynamic passenger flow-oriented urban rail train timetable energy-saving optimization method, and belongs to the technical field of urban rail transit operation. The invention provides a dynamic passenger flow-oriented urban rail train timetable energy-saving optimization method. Designing a passenger flow demand analysis and calculation expression by analyzing the passenger flow dynamic characteristics of the urban rail transit; based on the passenger flow dynamic demand and the train departure time table, constructing an accumulated waiting time calculation model of all passengers; establishing a train energy consumption calculation model based on the running time according to the line condition between adjacent stations on the running line of the rail train and the train traction mass; based on a dynamic passenger flow model and a train energy consumption model, establishing a departure timetable energy-saving dual-target optimization model by taking the shortest passenger accumulated waiting time and the lowest train comprehensive energy consumption as targets; designing an NSGA-II of a self-adaptive neighborhood search strategy to solve the model; cooperative optimization of passenger waiting time and operation energy consumption is realized by adjusting the departure time of the train and the inter-station operation time. According to the method, the Pareto optimal solution set of the service quality and the transportation cost can be provided, decision support is provided for formulating an operation scheme of a subway enterprise, and the method has relatively high use value and application prospect.

Description

Technical Field

[0001] The present invention relates to the field of urban rail transit operation technology, and in particular, to a comprehensive optimization method for improving rail transit operation quality and reducing operation costs. Background Art

[0002] With the development of my country's urbanization process, the population size and transportation demand of large and medium-sized cities have continued to rise, resulting in an increase in the number of household cars and increasing urban road traffic congestion, making urban rail transit the first choice for citizens to travel. Due to its specific operating mode, rail transit plays an irreplaceable role in alleviating the pressure of urban transportation. While facilitating citizens' travel, it is also of great value to the establishment of green and energy-saving cities. How to further give play to the high efficiency and energy-saving characteristics of urban rail transit is an urgent problem that government departments and operating companies need to solve, and it is also a research hotspot in the field of rail transit operation decision-making and operational optimization.

[0003] Urban rail transit is becoming an increasingly popular choice for citizens due to its fast, punctual and safe features. Travelers want to get on the train quickly and minimize waiting time and travel time, while urban rail transit operators want to reduce operating costs while meeting passenger travel needs. In order to alleviate the contradiction between passenger travel needs and cost reduction and efficiency improvement of operators, a new method for optimizing the operation plan of urban rail trains is proposed, which has important practical significance for improving the service quality of enterprises and reducing their operating costs.

[0004] From the perspective of urban rail transit operating companies and passenger travel needs, taking into account factors such as the temporal and spatial distribution characteristics of passenger flow, train operation restrictions, and the impact of inter-station running time on energy consumption, the train departure schedule and the running time between adjacent stations are optimized, and the matching relationship between dynamic passenger flow demand and train departure schedule and inter-station running time is explored to achieve the coordinated optimization of the service quality and operating energy consumption of the operating company. This invention has important theoretical significance and application value for the high-quality operation and intelligent management and control of urban rail transit. Summary of the invention

[0005] The purpose of the present invention is to provide an energy-saving optimization method for urban rail train departure timetables facing dynamic passenger flow. The train departure timetable is formulated by calculating the dynamic passenger flow of urban rail transit, and a model for minimizing the cumulative waiting time of passengers is constructed; and a train energy consumption calculation model based on the running time is designed according to the line conditions between adjacent stations and the vehicle traction quality; a departure timetable energy-saving dual-objective optimization model is established with the goal of minimizing the cumulative waiting time of passengers and minimizing the comprehensive energy consumption of trains; and an NSGA-II with an adaptive neighborhood search strategy is designed to solve the model and obtain a collaborative optimization plan for service quality and operating energy consumption. Provide technical support for the optimization of train operation plans of rail transit operating companies.

[0006] The above technical objectives of the present invention are achieved through the following technical solutions:

[0007] The present invention is an energy-saving optimization method for urban rail train departure timetables for dynamic passenger flow, which is characterized by analyzing the dynamic change characteristics of passenger flow at each station during operation for a single operating line of a rail transit train, optimizing the train departure timetable to reduce passenger waiting time; and then analyzing the relationship between the energy consumption and running time of the train between adjacent stations, and the running time between stations affects the number of passengers and passenger waiting time, so it is necessary to coordinately optimize the train departure timetable and the running time between each station to improve the enterprise service quality and reduce the train operation energy consumption. To ensure the applicability of the present invention, the following relevant assumptions are made:

[0008] (1) Passengers at each station follow the principle of "first come, first served";

[0009] (2) The train’s inter-station running time does not include the time spent at stations;

[0010] (3) The train’s stop times at each station are fixed and known;

[0011] (4) The inter-station energy consumption of a train is the minimum energy consumption corresponding to the inter-station running time of the train;

[0012] (5) There is no overtaking or giving way during the operation of the train.

[0013] S1: Design the dynamic demand calculation expression of urban rail transit passenger flow:

[0014] When analyzing passenger flow characteristics, minutes or longer time periods are usually chosen as the granularity, and the passenger flow density within the time period is assumed to obey a certain specific distribution; however, the actual operation process of the train is based on seconds as the time granularity, resulting in the traditional passenger flow statistics model not being able to accurately match the train operation schedule, so the passenger flow data statistics are not accurate enough, and the formulated train operation plan lacks rationality; the present invention designs a new method for calculating passenger flow demand within any time period with seconds as the time granularity.

[0015] S1.1: Number the stations on a single rail transit line S = {1, 2, ..., N}, u, v ∈ S represents a single platform; k represents the interval number of adjacent stations, in the interval (u, u + 1), let k = u; T = [0, T] is the daily operating time interval, δ represents the unit time period, the train departure time t ∈ T, and t = nδ, n is a positive integer;

[0016] S1.2: Any time period [t 1 ,t 2 ] to calculate the cumulative passenger flow demand from station u to station v (v>u). Let f u,v(t) is the passenger flow density from u to v at time point t, F u,v (t) is the cumulative passenger flow from u to v before time point t, Then [t 1 ,t 2 ] The cumulative passenger flow from u to v

[0017] S1.3: Draw the cumulative passenger flow demand curve between stations according to the dynamic characteristics of passenger flow, and perform data fitting; within the operating time interval T, the cumulative passenger flow demand between u and v is regarded as a continuous curve composed of multiple passenger flow peaks and flat peaks, divided into I (I ≥ 1) segments, where I is also the number of time periods with passenger flow peaks; each time period is represented by Represents; Fitting each curve with the Logistic function can be obtained The cumulative passenger flow demand within. The Logistic function expression is: in and p i is the parameter to be fitted, t is the time period time point within.

[0018] S2: Based on expression (2), a model for calculating the number of waiting passengers and the cumulative waiting time can be designed.

[0019] S2.1: Calculation of the number of waiting passengers at each station.

[0020] S2.1.1: Based on the I value, F u,v (t) and the passenger flow curve, the Logistic function can be used to obtain each The fitting parameter value of the cumulative passenger flow curve within the period. The fitting parameter value can be used to calculate the value of the cumulative passenger flow curve at any time point [t 1 ,t 2 ] passenger flow demand from u to v.

[0021] S2.1.2: Consider [t 1 ,t 2 ], if Then [t 1 ,t 2 ]The number of passengers departing from u for v is:

[0022] S2.1.3: If [t 1 ,t 2 ] respectively in and within, that is season and Represent the start and end time points of time period i respectively, then in [t 1 ,t 2 ]The number of passengers from u to v is:

[0023] S2.1.4: Calculate [t 1 ,t 2 ] Cumulative passenger flow from u:

[0024] S2.1.5: Assume that the passenger is at time If you arrive before , you will have the opportunity to board train j, and if you arrive before , you will have the opportunity to board train j. Passengers arriving later can only wait for train j+1.

[0025] S2.1.6: Considering the changes in passenger flow demand at each station, when the passenger flow is small during off-peak hours and there are no passengers staying at the station, calculate the number of passengers waiting to take train j at station u:

[0026] S2.1.7: During peak passenger flow, if there are stranded passengers at station u, the capacity of train j Less than passenger flow demand hour( represents the number of passengers in train j after it leaves station u), calculate the number of passengers who can successfully board train j at station u:

[0027] S2.1.8: If there are stranded passengers after train j leaves station u, some passengers need to wait for train j+1. The number of stranded passengers is Therefore, if train j-1 has delayed passengers at station u, calculate the cumulative number of passengers waiting for train j:

[0028] S2.2: Calculation of the passenger’s cumulative waiting time.

[0029] S2.2.1: The departure time interval of trains on a single line determines the tracking interval. The departure time interval between train j and train j+1 at the starting station must satisfy

[0030] S2.2.2: Calculate the time it takes for train j to arrive at station u: In formula (9) is the time when train j leaves station u-1, is the running time of train j from station u-1 to station u;

[0031] S2.2.3: Calculate the departure time of train j from station u: In formula (10) is the stop time of train j at station u;

[0032] S2.2.4: Calculate the number of passengers on train j getting off at station v: In formula (11) is the number of passengers on train j from station u to station v;

[0033] S2.2.5: During off-peak hours, all All passengers arriving at station u before can successfully board train j, and the number of stranded passengers In this scenario, calculate the cumulative waiting time of passengers waiting for train j at station u:

[0034] S2.2.6: During peak hours, there are stranded passengers. Calculate the cumulative waiting time of stranded passengers:

[0035] S2.2.7: Combining S2.2.4 and S2.2.5, calculate the time a passenger waits for train j at station u:

[0036] S2.2.8: Calculate the cumulative waiting time f of all passengers within the operating time interval T t : In formula (15), is a decision variable. If train j departs at time i, then The value is 1, otherwise the value is 0.

[0037] 3. According to the line conditions and train traction quality between different stations, a train energy consumption calculation model based on running time is established, which is characterized by: time-energy consumption relationship analysis between train stations, and energy consumption model design based on operating time. The specific operations include:

[0038] S3: Construct a train energy consumption calculation model based on running time.

[0039] S3.1: Analysis of the time-energy consumption relationship of train operation between stations.

[0040] S3.1.1: During operation, the energy consumption Q of the traction train in section k k Usually divided into traction operation energy consumption (including self-use energy consumption) and self-use energy consumption during idling, braking and stopping And the energy consumption of shunting operations in and out of the section and during the journey Three parts. Here we only consider and Calculate the energy consumption Q of the train in section k k :

[0041] S3.1.2: Calculate the energy consumption of traction operation Divided into high load traction operation energy consumption and partial load traction operation energy consumption In formula (18), represents the high-load traction time of the train in section k, is the average active current during high load traction, γ y is the traction force utilization coefficient; is the active current for own use; Indicates the accumulated time of high-load traction condition between certain stations; In formula (19), Indicates the partial load traction time. Indicates the accumulated time of partial load traction condition.

[0042] S3.1.3: Calculate the train's own energy consumption during idling, braking and stopping In formula (20), represents the time of coasting, braking and stopping of the train in section k, It is the accumulated time of coasting, braking and stopping conditions.

[0043] S3.1.4: Calculate the running time t of the train in section k k : t k =t k1 +t k2 +t k3 (twenty one)

[0044] S3.2: Train energy consumption calculation model based on running time.

[0045] S3.2.1: Establish the function expression of running time and energy consumption for segment k: f(t k )=Q k ·λ (22) In formula (22), λ is the influence coefficient of traction mass on train energy consumption, which is Among them, M 0 is the mass of the train, and M is the mass of the passengers on board. Taking into account the individual differences of passengers, 60 kg is used as the average mass of passengers, then the mass of passengers on board M = 60·E j,u , where E j,u is the number of passengers in train j after it leaves station u.

[0046] S3.2.2: Based on the allowed running time of the train within the station k and the energy consumption corresponding to different times, generate the train running time-energy consumption relationship curve between stations k, and generate the corresponding traction mass-energy consumption curve under the given running time.

[0047] S3.2.3: Based on the running time between stations on the operating line and the comprehensive curve of train traction quality-energy consumption, the least squares method is used to fit and obtain the energy consumption of train j in interval k. The total energy consumption Q of train j on the operating line can be calculated j :

[0048] S3.2.4: Calculate the cumulative energy consumption f of all trains running on the operating line during the operation period T c :

[0049] 4. Based on the passenger flow demand model and the train energy consumption model, a dual-objective optimization model is established with the goal of minimizing the cumulative waiting time of passengers and minimizing the energy consumption of train operation; the characteristics are:

[0050] S4: Construct a dual-objective optimization model.

[0051] S4.1: Based on the cumulative waiting time and operating energy consumption expressions designed in steps 2 and 3, a dual-objective optimization model is established: Z=min(f t ,f c ) (25)

[0052] S4.2: Due to the limitations of train calculation speed and maximum speed between different stations, it is necessary to set constraints on the running time between adjacent stations: In formula (24), is the minimum and maximum train running times allowed within the interval k.

[0053] S4.3: Since the running times of each interval are uncertain, it is necessary to set the safety tracking interval constraints for trains at each station:

[0054] S4.4: To ensure service quality, set the maximum time interval constraints for trains at each station:

[0055] S4.5: It is necessary to set the service capacity constraint for departing trains:

[0056] 5. Design a non-dominated sorting genetic algorithm with an adaptive neighborhood search strategy, determine the double-layer coding structure according to the problem characteristics and construct the initial solution set; design genetic operations that conform to the model characteristics based on the adaptive neighborhood search strategy;

[0057] S5: Design a double-layer coding method: Design a double-layer structure for the chromosome. The upper layer is the coding of the train departure time. Considering the discreteness of the departure time, use the 0-1 coding method, where 0 means no train departs at this time and 1 means a train departs at this time; the lower layer is the running time t between each station k , and use the real number coding method;

[0058] S5.1: Construct the initial solution set: Set the upper layer structure of each chromosome as an integer array G with a length of n, and initialize all elements in G to 0; let j←0, G[j]←1, ε←randInt(I min , I max ), when j + ε < n, let j←j + ε, G[j]←1; set the lower layer structure of the chromosome as a real number array with a length of |N| - 1, and let k←k + 1 until k = |N| - 1; repeat the construction process of the chromosome with the double-layer structure Popsize times to generate the initial population;

[0059] S5.2: Design of crossover operation: Use the roulette wheel strategy to select two chromosomes from the population for crossover, and design single-point crossover and multi-point crossover strategies for the upper and lower layer structures respectively. Judge whether the chromosomes after crossover meet the constraints and perform chromosome repair;

[0060] S5.3: Propose an adaptive neighborhood search strategy in the mutation operation:

[0061] (1) Destruction strategy: Select the chromosome c to be mutated, check the load factor of each train j according to the upper structure G[j], adjust the departure interval of the train j with the lowest load factor and its train j-1 and train j+1, and increase the departure interval by deleting the train number; check the passenger flow section of each station u according to the lower structure g[k], and increase the running time of m sections with small passenger flow sections; the purpose of this operation is to reduce energy consumption;

[0062] (2) Repair strategy: For the selected chromosome c, check the full load rate of each train j according to the upper structure G[j], adjust the departure interval of the train j with the highest full load rate and its train j-1 and train j+1, and reduce the departure interval by increasing the number of trains; check the passenger flow section of each station u according to the lower structure g[k], and reduce the running time of the m sections with large passenger flow sections; the purpose of this operation is to reduce the waiting time of passengers;

[0063] S5.4: Selection operation: When generating a new population, calculate the level number r(·) and crowding distance d(·) of each chromosome. Based on r(·) and d(·), the advantages and disadvantages of chromosomes can be compared, and the fast non-dominated sorting strategy can be used to select the better chromosomes to enter the next generation.

[0064] An electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

[0065] A computer-readable storage medium stores a computer program, which executes the steps of the above method when executed by a processor.

[0066] Compared with the existing technologies in this research field, the beneficial results of the present invention are:

[0067] 1. According to the passenger flow variation characteristics of different stations, a method for calculating passenger flow demand in any period with seconds as the time granularity is proposed; and the difference in passenger flow demand during off-peak and peak hours is fully considered, and a calculation model for the cumulative waiting time of all passengers under different scenarios is designed;

[0068] 2. Considering the composition of vehicle energy consumption during inter-station operation, a train energy consumption calculation method based on inter-station operation time and train traction quality is proposed. The matching of passenger flow demand and train energy consumption can be achieved by adjusting the inter-station operation time;

[0069] 3. According to the dynamic demand of passenger flow at each station, the coordinated optimization strategy of train departure schedule and inter-station running time was designed, and a dual-objective optimization model was constructed to minimize the cumulative waiting time of passengers and the comprehensive energy consumption of train operation, so as to improve service quality and reduce operating costs;

[0070] 4. In the design of the solution method, an improved NSGA-II is adopted to design a double-layer chromosome, which can not only characterize the characteristics of the train departure schedule, but also realize the control of the running time between adjacent stations. In order to improve the solution efficiency of the algorithm, an adaptive neighborhood search strategy is designed. By checking the train load factor and passenger flow section, the train departure schedule and the running time between stations are improved to speed up the solution efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is the overall flow chart of the present invention.

[0072] Figure 2 This is a diagram showing the relationship between energy consumption and time between different stations described in the present invention.

[0073] Figure 3 This is a diagram showing the relationship between energy consumption, time and traction quality between stations in the present invention.

[0074] Figure 4 It is a specific implementation flow chart of the solution algorithm in the present invention.

[0075] Figure 5 It is the Pareto solution set of the train operation plan obtained by the algorithm in the present invention. DETAILED DESCRIPTION

[0076] In this embodiment, the research object is first selected, the operating line is determined and the basic data is set; secondly, the passenger flow data is counted, analyzed and curve-fitted according to the passenger flow distribution of each station on weekdays, and the Logistic parameters of the passenger flow getting on and off at each station in different time periods are determined; then, the relationship curve between the running time, traction quality and energy consumption is determined according to the line conditions between the stations, and the fitting data of different curves are obtained by the least squares method; then a dual-objective optimization model with the minimum cumulative waiting time for passengers and the lowest overall operating energy consumption is established; finally, the designed algorithm is used to solve the model and analyze the results. Steps are as follows: Figure 1 shown.

[0077] In the specific implementation, Wuhan Metro Line 2 was selected as the research object, and the section from Changqing Garden to Wuhan East Station was selected, with a total of 25 stations and a total length of 31.24km, and the down direction was selected. The station spacing and the established running time are shown in Table 1.

[0078]

[0078] Table 1. Operating time and distance between stations in the designated interval

[0078] The trains on this line are of type A (4 motors and 2 trailers) with 6 cars. The maximum speed of the passenger cars is 80 km / h and the train capacity is 1,440 people. The relevant parameters of the train operation are shown in Table 2.

[0079]

[0080] Table 2 Train operation related parameters

[0082] In the embodiment, the initial departure time of the train is set according to the train tracking interval, and the 0-1 variable is set at a time interval of 30 seconds; the designed example code is shown in Table 3.

[0083]

[0084] Table 3 Chromosome coding structure

[0085] In the embodiment, the genetic iteration operation with the adaptive neighborhood search strategy is performed in combination with the coding structure of Table 2. After the iteration is terminated, the Pareto solution obtained is as follows: Figure 5 shown.

[0086] The optimal value of each objective is taken from the obtained non-dominated solution set, which corresponds to the optimal solution with the minimum cumulative waiting time of passengers (solution 1) and the lowest energy consumption of train operation (solution 2).

[0087]

[0088] Table 4 Non-dominated solutions corresponding to minimum waiting time and minimum train energy consumption

[0089] The number of trains before optimization was 330, and 278 trains were run in non-dominated solution 1, with a 15.75% reduction in the number of trains. This was mainly due to the large departure intervals during the last trains in the morning and evening, and the small departure intervals during the morning and evening peaks when passenger demand was high. The waiting time for passengers was reduced from 9.15×10 6 min increased to 9.21×10 6 min, the cumulative waiting time increased by 0.66%; the average single train running time after optimization increased by 14s, which was mainly reflected in the general increase in the running time in short-distance sections and the reduction in the running time in some long-distance sections; the average single-trip operating energy consumption of the train on the entire line was reduced by 8.871kWh.

[0090] The non-dominated solution 2 corresponds to 235 trains, with a 28.79% decrease in the number of trains. This is mainly due to the reduction of train operation time during peak hours in the morning and evening, and the increase of train operation intervals during off-peak hours. Compared with the existing train operation diagram, the passenger waiting time is reduced by 9.15×10 6 min increased to 9.93×10 6 min, the cumulative waiting time increased by 8.52%; compared with before optimization, the average single running time of the train increased by 40s, and the average single running energy consumption of the train decreased by 17.955kWh.

[0091] In this embodiment, an electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

[0092] In this embodiment, a computer-readable storage medium stores a computer program on the computer-readable storage medium, and the computer program executes the steps of the above method when executed by a processor.

Claims

1. Energy-saving optimization method for urban rail train timetable for dynamic passenger flow, Features: ① Analyze the dynamic characteristics of urban rail transit passenger flow, design passenger flow demand analysis and calculation expressions; ② Based on the dynamic passenger flow demand and train departure schedule, build a calculation model for the accumulated waiting time of all passengers; ③According to the line conditions between adjacent stations on the rail train operation line and the vehicle load situation, a train energy consumption calculation model based on the running time is established; ④Based on the dynamic passenger flow model and the train energy consumption model, a dual-objective energy-saving optimization model for the departure timetable is established with the goals of minimizing the cumulative waiting time of passengers and minimizing the comprehensive energy consumption of trains; ⑤NSGA-II with an adaptive neighborhood search strategy is designed to solve the model; the coordinated optimization of service quality and operating energy consumption is achieved by adjusting the train departure time and the running time between stations.

2. According to the analysis of the dynamic characteristics of urban rail transit passenger flow in claim 1, establish a passenger flow demand analysis and calculation model, Features: Dynamic passenger flow characteristics analysis; Design of expression for calculating passenger flow demand in any period of time; specific operations include: S1: dynamic passenger flow feature analysis and calculation. S1.1: Number the stations on a single rail transit line S = {1, 2, ..., N}, u, v ∈ S represents a single platform; k represents the interval number of adjacent stations, in the interval (u, u + 1), let k = u; T = [0, T] is the daily operating time interval, δ represents the unit time period, the train departure time t ∈ T, and t = nδ, n is a positive integer; S1.2: Any time period [t 1 ,t 2 ] to calculate the cumulative passenger flow demand from station u to station v (v>u). Let f u,v (t) is the passenger flow density from u to v at time point t, F u,v (t) is the cumulative passenger flow from u to v before time point t, Then [t 1 ,t 2 ] The cumulative passenger flow from u to v S1.3: Draw the cumulative passenger flow demand curve between stations according to the dynamic characteristics of passenger flow, and perform data fitting; within the operating time interval T, the cumulative passenger flow demand between u and v is regarded as a continuous curve composed of multiple passenger flow peaks and flat peaks, divided into I (I ≥ 1) segments, where I is also the number of time periods with passenger flow peaks; each time period is represented by Represents; Fitting each curve with the Logistic function can be obtained The cumulative passenger flow demand within. The Logistic function expression is: in and p i is the parameter to be fitted, t is the time period time point within.

3. According to claim 1, a calculation model for the accumulated waiting time of all passengers is constructed based on the dynamic passenger flow demand and train departure schedule; Features: Calculate the number of waiting passengers at each station; calculate the cumulative waiting time of all passengers; specific operations include: S2: Model for calculating the number of waiting passengers and the cumulative waiting time. S2.1: Calculation of the number of waiting passengers: S2.1.1: Based on the I value, F u,v (t) and the passenger flow curve, the Logistic function can be used to obtain each The fitting parameter value of the cumulative passenger flow curve within the period. The fitting parameter value can be used to calculate the value of the cumulative passenger flow curve between any two time points [t 1 ,t 2 ]The number of passenger flow demands from u to v. S2.1.2: Consider [t 1 ,t 2 ] distribution, if Then [t 1 ,t 2 ]The number of passengers departing from u to v is: S2.1.3: If [t 1 ,t 2 ] respectively in and within, that is season and Represent the start and end time points of time period i respectively, then in [t 1 ,t 2 ]The number of passengers from u to v is: S2.1.4: Calculate [t 1 ,t 2 ] Cumulative passenger flow from u: S2.1.5: Assume that the passenger is at time If you arrive before then you will have the opportunity to board train j, and if you arrive before then Passengers arriving later can only wait for train j+1. S2.1.6: Considering the changes in passenger flow demand at each station, when the passenger flow is small during off-peak hours and there are no passengers staying at the station, calculate the number of passengers waiting to take train j at station u: S2.1.7: During peak passenger flow, if there are stranded passengers at station u, the capacity of train j Less than passenger flow demand hour( represents the number of passengers in train j after it leaves station u), calculate the number of passengers who can successfully board train j at station u: S2.1.8: If there are stranded passengers after train j leaves station u, some passengers need to wait for train j+1. The number of stranded passengers is Therefore, if train j-1 has delayed passengers at station u, calculate the cumulative number of passengers waiting for train j: S2.2: Passenger waiting time calculation: S2.2.1: The departure time interval of trains on a single line determines the tracking interval. The departure time interval between train j and train j+1 at the starting station must satisfy S2.2.2: Calculate the time it takes for train j to arrive at station u: in is the time when train j leaves station u-1, is the running time of train j from station u-1 to station u; S2.2.3: Calculate the departure time of train j from station u: in is the stop time of train j at station u; S2.2.4: Calculate the number of passengers on train j getting off at station v: in is the number of passengers on train j from station u to station v; S2.2.5: During off-peak hours, all All passengers arriving at station u before can successfully board train j, and the number of stranded passengers In this scenario, calculate the cumulative waiting time of passengers waiting for train j at station u: S2.2.6: During peak hours, there are stranded passengers. Calculate the cumulative waiting time of stranded passengers: S2.2.7: Combining S2.2.4 and S2.2.5, calculate the time a passenger waits for train j at station u: S2.2.8: Calculate the cumulative waiting time f of all passengers within the operating time interval T t : in is a decision variable. If train j departs at time i, then The value is 1, otherwise the value is 0.

4. According to claim 1, a train energy consumption calculation model based on running time is established according to the inter-station line conditions and vehicle traction quality. Features: Analysis of the time-energy consumption relationship between train stations and design of energy consumption model based on operation time. Specific operations include: S3: Construct a train energy consumption calculation model based on running time. S3.1: Analysis of the time-energy consumption relationship of train operation between stations. S3.1.1: During operation, the energy consumption Q of the traction train in section k k Usually divided into traction operation energy consumption (including self-use energy consumption) and self-use energy consumption during idling, braking and stopping And the energy consumption of shunting operations in and out of the section and during the journey Three parts. Here we only consider and Calculate the energy consumption Q of the train in section k k : S3.1.2: Calculate the energy consumption of traction operation Divided into high load traction operation energy consumption and partial load traction operation energy consumption Among them U w Indicates the contact network voltage, is the average active current during high load traction, γ y is the traction force utilization coefficient; is the active current for own use, represents the high-load traction time of the train in section k; Indicates the accumulated time of high-load traction condition between certain stations; Indicates the average active current during partial load traction. Indicates the partial load traction time. Indicates the accumulated time of partial load traction condition. S3.1.3: Calculate the train's own energy consumption during idling, braking and stopping in represents the time of coasting, braking and stopping of the train in section k, It is the accumulated time of coasting, braking and stopping conditions. S3.1.4: Calculate the running time t of the train in section k k : t k =t k1 +t k2 +t k3 S3.2: Train energy consumption calculation model based on running time. S3.2.1: Establish the function expression of running time and energy consumption for segment k: f(t k )=Q k ·λ Where λ is the influence coefficient of traction mass on train energy consumption, and its value is M 0 is the mass of the train, M is the mass of the passengers on board; considering the individual differences of passengers, 60 kg is used as the average mass of passengers, then the mass of passengers on board M = 60·E j,u , E j,u is the number of passengers in train j after it leaves station u. S3.2.2: Based on the allowed running time of the train within the station k and the energy consumption corresponding to different times, generate the train running time-energy consumption relationship curve between stations k, and generate the corresponding traction mass-energy consumption curve under the given running time. S3.2.3: Based on the running time between stations on the operating line and the comprehensive curve of train traction quality-energy consumption, the least squares method is used to fit and obtain the energy consumption of train j in interval k. The total energy consumption Q of train j on the operating line can be calculated j : S3.2.4: Calculate the cumulative energy consumption f of all trains running on the operating line during the operation period T c :

5. According to claim 1, based on the passenger flow model and the train energy consumption model, a dual-objective optimization model is established with the shortest cumulative waiting time for passengers and the lowest energy consumption for train operation as the goals; Features: S4: Construct a dual-objective optimization model. S4.1: Based on the cumulative waiting time and operating energy consumption expressions designed in steps 2 and 3, a dual-objective optimization model is established: Z=min(f t ,f c ) S4.2: Due to the limitations of train calculation speed and maximum speed between different stations, it is necessary to set constraints on the running time between stations: are the minimum and maximum train running times allowed in interval k. S4.3: Since the running time of each section is uncertain, it is necessary to set the safe tracking interval constraints of the train at each station: S4.4: To ensure service quality, set the maximum time interval constraints for trains at each station: S4.5: It is necessary to set the service capacity constraints of the outgoing trains:

6. The non-dominated sorting genetic algorithm for designing an adaptive neighborhood search strategy according to claim 1, Features: Determine the two-layer coding structure based on the problem characteristics; Construct an initial solution set; Design genetic operations that meet the characteristics of the model; propose an adaptive neighborhood search strategy; the specific process includes: S5: Design a double-layer coding method: Design a double-layer structure for the chromosome. The upper layer is the train departure time code. Considering the discreteness of the departure time, a 0-1 coding method is used, where 0 means no train is leaving at this time, and 1 means a train is leaving at this time. The lower layer is the running time t between each station. k , using real number encoding; S5.1: Construct the initial solution set: Set the upper structure of each chromosome as an integer array G of length n, and initialize all elements in G to 0; Let j←0, G[j]←1, ε←randInt(I min , I max ), when j + ε < n, let j←j + ε, G[j]←1; Set the lower structure of the chromosome as a real number array all with length |N| - 1, let k←1, k←k + 1 until k = |N| - 1; Repeat the construction process of the chromosome with the double - layer structure Popsize times to generate the initial population; S5.2: Crossover operation design: Use the roulette strategy to select two chromosomes from the population for crossover, design single-point crossover and multi-point crossover strategies for the upper and lower structures respectively, determine whether the chromosomes after crossover meet the constraints, and perform chromosome repair; S5.3: Propose an adaptive neighborhood search strategy in the mutation operation: S5.3.1: The destruction strategy involves: selecting the chromosome c to be mutated, checking the load factor of each train j for the upper structure G[j], adjusting the departure interval of the train j with the lowest load factor and its train j-1 and train j+1, and increasing the departure interval by deleting the train number; checking the passenger flow section of each station u for the lower structure g[k], and increasing the running time of m sections with small passenger flow sections; the purpose of this operation is to reduce energy consumption; S5.3.2: The repair strategy involves: for the selected chromosome c, check the full load rate of each train j for the upper structure G[j], adjust the departure interval of the train j with the highest full load rate and its train j-1 and train j+1, and reduce the departure interval by increasing the number of trains; check the passenger flow section of each station u for the lower structure g[k], and reduce the running time of m sections with large passenger flow sections; the purpose of this operation is to reduce the waiting time of passengers; S5.4: Selection operation: When generating a new population, calculate the level number r(·) and crowding distance d(·) of each chromosome. Based on r(·) and d(·), the advantages and disadvantages of chromosomes can be compared, and the fast non-dominated sorting strategy can be used to select the better chromosomes to enter the next generation.

7. An electronic device, comprising a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

8. A computer-readable storage medium having a computer program stored thereon, wherein the computer program executes the steps of any one of the methods of claims 1 to 6 when executed by a processor.

Images