Rail transit scheduling method based on multi-computer cluster parallel offline solution

Abstract

The invention relates to a rail transit scheduling method based on multi-computer cluster parallel offline solution, and the method comprises the steps: generating a to-be-updated urban rail train arrival time set and a to-be-updated urban rail train departure time set according to an urban rail train arrival time initial value, an urban rail train departure time initial value and a preset value range; forming a power supply optimization model according to the to-be-updated urban rail train arrival time set and the to-be-updated urban rail train departure time set; forming a passenger flow optimization model according to the passenger flow number, the to-be-updated urban rail train arrival time set and the to-be-updated urban rail train departure time set; a host acquires a to-be-updated urban rail train arrival time set and a to-be-updated urban rail train departure time set, and sends the sets to a plurality of slaves; and the plurality of slaves carry out parallel calculation on the power supply optimization model and the passenger flow optimization model. According to the method and the device, the problems of low speed and low efficiency of urban rail train operation scheduling optimization are solved, and rapid optimization of operation scheduling is realized.

Description

Technical Field

[0001] This application relates to the technical field of rail transit, and particularly to a rail transit scheduling method based on parallel offline solution of a multi-computer cluster. Background Art

[0002] Currently, in rail transit, the train operation diagram is extremely important. It is the key to ensuring the orderly operation of trains. By accurately setting the arrival and departure times of each train at stations and the connection of line transfers, it avoids train conflicts and optimizes the turnover. At the same time, it reasonably arranges the entry and exit of the depot; for passenger services, it provides support for the accuracy of travel information, which is conducive to passengers planning their trips; in terms of operation management and resource allocation, it is the basis for human resource management, which can reasonably arrange the working hours of station and train crew members, and is also closely related to the utilization and maintenance of equipment resources, assisting the efficient operation of systems such as signals and power supplies and facilitating the arrangement of equipment maintenance. It plays an indispensable role in the stable and efficient operation of the entire rail transit system.

[0003] In the prior art (CN 118863146 A), an optimization method for the integration of urban rail transit power supply and operation is disclosed Figure 1 including obtaining basic data; calculating the mechanical power spatio-temporal distribution diagram of all trains on the whole line by solving the energy-saving operation speed curve of a single train interval; constructing the topological structure of an urban rail DC traction power supply system with a bidirectional converter device; performing power flow calculation to obtain the network voltage of all trains on the whole line, the voltage, current, and power of all substations; constructing and optimizing a DC traction power supply system energy-saving and safety optimization model that minimizes the electricity cost and network voltage fluctuation; solving the optimized DC traction power supply system energy-saving and safety optimization model to obtain the optimal solution.

[0004] In the related art, only the energy-saving and safety optimization model of the DC traction power supply system is constructed, without considering the influence of other energy consumptions and passenger flow on train operation, and the speed of updating the operation scheduling of urban rail trains is slow and the efficiency is low. Summary of the Invention

[0005] The embodiments of this application provide a rail transit scheduling method based on parallel offline solution of a multi-computer cluster to at least solve the problem of slow speed and low efficiency in updating the operation scheduling of urban rail trains in the related art.

[0006] In a first aspect, the embodiments of this application provide a rail transit scheduling method based on parallel offline solution of a multi-computer cluster. The operation scheduling of urban rail trains includes adjusting the train tracking interval and the stop time. The method includes:

[0007] S1: Generate a set of arrival times of urban rail trains to be updated and a set of departure times of urban rail trains to be updated according to the initial arrival time value of the urban rail train, the initial departure time value of the urban rail train, and the preset value range.

[0008] S2: Based on the initial value of the comprehensive optimization index of the urban rail traction power supply system and the initial value of the allowable delay time limit index of all urban rail trains on the operating line, according to the initial arrival time value of the urban rail train, the initial departure time value of the urban rail train, the set of arrival times of urban rail trains to be updated, and the set of departure times of urban rail trains to be updated, calculate the comprehensive optimization index of the urban rail traction power supply system to be updated and the allowable delay time limit index of all urban rail trains on the operating line to be updated. The comprehensive optimization index of the urban rail traction power supply system to be updated and the allowable delay time limit index of all urban rail trains on the operating line to be updated are set with equal weights, and a normalized objective function is established to form a power supply optimization model.

[0009] S3: According to the number of passengers, the set of arrival times of urban rail trains to be updated, and the set of departure times of urban rail trains to be updated, calculate the allowable delay time limit index for the operation of the train to be updated in an interval and the traction energy consumption index for the operation of the train to be updated in an interval, and set a screening function for screening the minimum values of the two indexes. The screening function for screening the minimum values of the two indexes forms a passenger flow optimization model.

[0010] S4: A host obtains the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated and sends them to multiple slaves. After receiving them, the multiple slaves calculate the power supply optimization model and the passenger flow optimization model in parallel, iteratively optimize the data in the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated, obtain the optimal set of train tracking intervals and the optimal set of train stop times, and adjust the train tracking intervals and stop times according to the two optimal sets to complete the operation scheduling of urban rail trains.

[0011] By constructing a power supply optimization model and a passenger flow optimization model, the rationality and flexibility of the operation plan are improved. Through the parallel calculation of the host and the slaves, the speed of the operation scheduling of urban rail trains is accelerated.

[0012] In some embodiments, S2 includes:

[0013] According to the active power under the rectification condition, the active power under the inversion condition, and the power returned by the main substation, through power flow calculation, obtain the traction energy consumption of the train under the rectification condition, the feedback energy under the inversion condition, and the power returned by the main substation.

[0014] According to the initial arrival time and initial departure time of urban rail trains, the set of urban rail train arrival times to be updated and the set of urban rail train departure times to be updated, the traction energy consumption when the train is in the rectification condition, the feedback energy when the train is in the inversion condition, and the power returned by the main substation, the comprehensive optimization index of the urban rail traction power supply system to be updated is calculated;

[0015] According to the initial arrival time and initial departure time of urban rail trains, the set of urban rail train arrival times to be updated and the set of urban rail train departure times to be updated, the number of trains and the number of stations, the allowable delay time limit index of all urban rail trains on the operating line to be updated is calculated;

[0016] According to the initial value of the comprehensive optimization index of the urban rail traction power supply system and the initial value of the allowable delay time limit index of all urban rail trains on the operating line, the comprehensive optimization index of the urban rail traction power supply system to be updated and the allowable delay time limit index of all urban rail trains on the operating line to be updated, a normalized objective function is established to obtain a power supply optimization model.

[0017] By dynamically calculating the traction energy consumption and feedback energy, not only can the energy loss be minimized to the greatest extent, but also the energy feedback efficiency of the main substation can be improved, and at the same time, a reasonable arrangement of the operation time can be achieved.

[0018] In some of the embodiments, the set of urban rail train arrival times to be updated includes the times when train i to be updated arrives at each station, and the set of urban rail train departure times to be updated includes the times when train i to be updated departs from each station; the comprehensive optimization index of the urban rail traction power supply system to be updated is obtained through a first calculation formula, and the first calculation formula includes:

[0019]

[0020] where f1 is the comprehensive optimization index of the urban rail traction power supply system to be updated, is the traction energy consumption when the train is in the rectification condition, is the feedback energy when the train is in the inversion condition, is the power returned by the main substation, is the initial value of the time when train i arrives at each station, are the initial values of the times when train i departs from each station respectively, is the time when train i to be updated arrives at each station, is the time when train i to be updated departs from each station.

[0021] Through this calculation formula, the energy states of each station and each train at a specific moment can be comprehensively analyzed, and the timeliness and accuracy of the calculation results of the optimization index can be improved.

[0022] In some of these embodiments, the allowable delay time limit index for all urban rail trains on the operation line to be updated is obtained through a second calculation formula, and the second calculation formula includes:

[0023]

[0024] where f2 is the allowable delay time limit index for all urban rail trains on the operation line to be updated, is the initial value of the arrival time of train i at station s, is the initial value of the departure time of train i from station s, is the arrival time of train i to be updated at station s, is the departure time of train i to be updated from station s, N is the total number of stations, and M is the total number of trains.

[0025] Through this formula, the time optimization effect of train operation can be accurately quantified, ensuring that the punctuality rate of the train is not affected while optimizing the energy consumption.

[0026] In some of these embodiments, the power supply optimization model includes:

[0027]

[0028] where f 1,0 is the initial value of the comprehensive optimization index of the urban rail traction power supply system, and f 2,0 is the initial value of the allowable delay time limit index for all urban rail trains on the operation line.

[0029] Through normalization processing, this model can make a reasonable trade-off between different optimization objectives, improving the efficiency and robustness of the optimization calculation.

[0030] In some of these embodiments, the delay time limit index for a train to be updated during operation in an interval includes:

[0031] f T,i,g = |T i,g (λ i,g , v i,g ) - T i,g |;

[0032] where λ i,g is the inert control coefficient, v i,g is the target speed of train i during operation in interval g, T i,g is the planned operation time of train i in the gth interval, and T i,g (λ i,g , v i,g ) is the actual time of train i during operation in the gth interval at λ i,g , v i,g .

[0033] By introducing variables such as the target speed and the planned running time, more accurate calculations can be obtained, and further optimized to obtain results that better meet the conditions.

[0034] In some of these embodiments, the traction energy consumption index of the train to be updated during operation in a section includes:

[0035]

[0036] Among them, μ i,g is the traction force utilization coefficient of train i in section g, M i,g is the vehicle weight of train i during operation in section g, M i,g is related to the number of passengers, f max,i,g (v i,g ,M i,g ) is the maximum traction force when the train speed is vi,g and the vehicle weight is Mi,g, and η i is the electromechanical efficiency of train i.

[0037] By introducing the traction energy consumption index, the energy consumption level of the train during operation can be intuitively reflected, providing a direct energy consumption target for the optimization model, thereby achieving the optimization effect.

[0038] In some of these embodiments, in S4, a host obtains the set of arrival times of the urban rail trains to be updated and the set of departure times of the urban rail trains to be updated, and sends them to multiple slaves. Further, it includes:

[0039] Receiving the initial values of the arrival times of the urban rail trains and the initial values of the departure times of the urban rail trains, the preset value range, the initial value of the comprehensive optimization index of the urban rail traction power supply system, the initial value of the allowable delay time limit index of all urban rail trains on the operation line, and the number of passengers;

[0040] According to the initial values of the arrival times of the urban rail trains and the initial values of the departure times of the urban rail trains and the preset value range, generating the set of arrival times of the urban rail trains to be updated and the set of departure times of the urban rail trains to be updated; generating multi-computer cluster parallel solution instructions, intelligent algorithm individual information, and multiple calculation tasks; wherein, the parallel solution instructions include the number of slaves, the set of arrival times of the urban rail trains to be updated and the set of departure times of the urban rail trains to be updated, the initial values of the arrival times of the urban rail trains and the initial values of the departure times of the urban rail trains, and the preset value range, and the individual information includes the initial value of the comprehensive optimization index of the urban rail traction power supply system and the initial value of the allowable delay time limit index of all urban rail trains on the operation line;

[0041] Evenly distributing and sending multiple calculation tasks to multiple slaves, and sending the number of passengers, the parallel solution instructions, and the individual information to each slave.

[0042] The host is responsible for task allocation and data aggregation, improving the adjustment efficiency of urban rail train operation scheduling.

[0043] In some of these embodiments, in S4, after receiving by multiple slave machines, they calculate the power supply optimization model and the passenger flow optimization model in parallel, and iteratively optimize the data in the set of to-be-updated arrival times of urban rail trains and the set of to-be-updated departure times of urban rail trains, further including:

[0044] Receive partial calculation tasks, the number of passengers, the parallel solution instruction, and the individual information sent by the host; according to the order of the issued calculation tasks, obtain the corresponding to-be-updated arrival times of train i at each station and the to-be-updated departure times of train i from each station in the set of to-be-updated arrival times of urban rail trains and the set of to-be-updated departure times of urban rail trains;

[0045] According to the initial arrival time of the urban rail train, the initial departure time of the urban rail train, the initial comprehensive optimization index of the urban rail traction power supply system, the initial allowable delay time limit index of all urban rail trains on the operation line, the corresponding to-be-updated arrival times of train i at each station and the to-be-updated departure times of train i from each station, calculate the to-be-updated comprehensive optimization index of the urban rail traction power supply system and the to-be-updated allowable delay time limit index of all urban rail trains on the operation line, and run the power supply optimization model;

[0046] According to the number of passengers, the initial arrival time of the urban rail train, the initial departure time of the urban rail train, the corresponding to-be-updated arrival times of train i at each station and the to-be-updated departure times of train i from each station, calculate the to-be-updated delay time limit index for a train running in a section and the to-be-updated traction energy consumption index for a train running in a section, and run the passenger flow optimization model;

[0047] Among them, the power supply optimization model and the passenger flow optimization model in any one slave machine operate simultaneously to obtain an optimized train tracking interval and train stop time;

[0048] Perform multiple iterative optimizations. During the iteration process, according to the optimized train tracking interval and train stop time obtained in the Nth optimization, the preset value range, and the individual information of the intelligent algorithm, update the data in the set of to-be-updated arrival times of urban rail trains and the set of to-be-updated departure times of urban rail trains; according to the optimized train tracking interval and train stop time obtained in the Nth optimization, update the initial arrival time of the urban rail train, the initial departure time of the urban rail train, the initial comprehensive optimization index of the urban rail traction power supply system, and the initial allowable delay time limit index of all urban rail trains on the operation line used in the (N + 1)th optimization; control the power supply optimization module and the passenger flow optimization module to perform the (N + 1)th optimization according to the updated data, and obtain the optimized train tracking interval and train stop time obtained in the (N + 1)th optimization.

[0049] Through the parallel computing ability of the computer cluster, the efficiency and accuracy of the optimization of urban rail train operation scheduling are greatly improved.

[0050] In some of these embodiments, in S4, an optimal set of train tracking intervals and an optimal set of train stop times are obtained, and based on the two optimal sets, the train tracking intervals and stop times are adjusted to complete the operation scheduling of urban rail trains. Further, it includes: setting an iteration count threshold, when the iteration count reaches the iteration count threshold, stopping the iterative calculation, receiving the optimized train tracking intervals and train stop times obtained by multiple slave machines in the last optimization, and forming an optimal set of train tracking intervals and an optimal set of train stop times, and based on the two optimal sets, adjusting the train tracking intervals and stop times to complete the operation scheduling of urban rail trains.

[0051] By setting the iteration count threshold, the complexity of the calculation process is effectively controlled, ensuring that the optimization results have high timeliness and availability.

[0052] Compared with the related technology, the rail transit scheduling method based on parallel offline solution of multiple computer clusters provided by the embodiments of the present application constructs a power supply optimization model and a passenger flow optimization model, combines the influence of power supply energy consumption and passenger flow on trains, and through parallel operation of the computer cluster, solves the problems of slow optimization speed and low efficiency in the operation scheduling of urban rail trains, and realizes fast and efficient optimization to meet the actual needs.

[0053] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more concise and understandable. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The drawings described herein are used to provide a further understanding of the present application, and constitute a part of the present application. The illustrative embodiments and descriptions thereof of the present application are used to explain the present application and do not constitute an improper limitation of the present application. In the drawings:

[0055] Figure 1 is a flowchart of a rail transit scheduling method based on parallel offline solution of multiple computer clusters according to an embodiment of the present application;

[0056] Figure 2 is another flowchart of a rail transit scheduling method based on parallel offline solution of multiple computer clusters according to an embodiment of the present application;

[0057] Figure 3 is another flowchart of a rail transit scheduling method based on parallel offline solution of multiple computer clusters according to an embodiment of the present application;

[0058] Figure 4It is another flowchart of a rail transit scheduling method based on parallel offline solution of multiple computer clusters according to an embodiment of the present application. Detailed implementation manners

[0059] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be described and explained below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments provided in the present application without creative efforts belong to the scope of protection of the present application.

[0060] Obviously, the accompanying drawings in the following description are only some examples or embodiments of the present application. For those of ordinary skill in the art, the present application can also be applied to other similar scenarios based on these drawings without creative efforts. In addition, it can also be understood that although the efforts made in this development process may be complex and lengthy, for those of ordinary skill in the art related to the content disclosed in the present application, some design, manufacturing or production changes made based on the technical content disclosed in the present application are only conventional technical means and should not be understood as insufficient disclosure of the content of the present application.

[0061] Referring to "embodiments" in the present application means that specific features, structures or characteristics described in connection with the embodiments can be included in at least one embodiment of the present application. The phrase appears in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those of ordinary skill in the art explicitly and implicitly understand that the embodiments described in the present application can be combined with other embodiments without conflict.

[0062] Unless otherwise defined, the technical terms or scientific terms involved in this application shall have the ordinary meanings understood by those with ordinary skills in the technical field to which this application belongs. The words such as "a", "an", "one kind", "the" and the like involved in this application do not indicate a quantity limitation and may represent a singular or plural number. The terms "comprise", "include", "have" and any variations thereof involved in this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may further include steps or units not listed, or may further include other steps or units inherent to these processes, methods, products or devices. The words such as "connect", "be connected", "couple" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "plurality" involved in this application means two or more. "And / or" describes the association relationship of associated objects and indicates that three relationships may exist. For example, "A and / or B" may represent: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally represents an "or" relationship between the front and back associated objects. The terms "first", "second", "third" and the like involved in this application are only used to distinguish similar objects and do not represent a specific order for the objects.

[0063] Currently, the operation diagram optimization models of rail transit lines at home and abroad generally use swarm intelligence algorithms for serial solution, such as genetic algorithms, gravitational search algorithms, salp swarm algorithms, etc.

[0064] When using swarm intelligence algorithms for serial solution of the operation diagram optimization model, the simulation time often lasts about 100 hours. When optimizing the operation diagram with the goal of optimizing system energy consumption, it is necessary to formulate an optimized operation diagram one week in advance, which is difficult to meet the requirements of urban rail transit operation departments for the timeliness of train diagram adjustment.

[0065] From the perspective of the timeliness of train diagram adjustment by the operation department, the simulation time of using swarm intelligence algorithms for serial solution of the operation diagram optimization model is too long to verify the optimization effect of the optimized operation diagram in a timely manner. If an optimized operation diagram cannot be formulated effectively in a timely manner, it will seriously affect the optimization of system energy consumption and cause unnecessary energy waste;

[0066] Currently, the solution methods of the operation diagram optimization model in the rail transit field still mainly rely on serial solution. However, sometimes due to the excessive complexity of the optimization model, even the time used for single-machine multi-threaded parallel computing is too long, and the model solution cost is too large. It cannot effectively utilize the advantage of parallel solution of multi-computer clusters to accelerate the solution speed, nor can the simulation time be compressed within 6 hours.

[0067] To solve the above problems, this embodiment provides a rail transit scheduling method based on parallel offline solution of multiple computer clusters.

[0068] Figure 1 It is a flowchart of the rail transit scheduling method based on parallel offline solution of multiple computer clusters according to an embodiment of the present application. Urban rail train operation scheduling includes adjusting the train tracking interval and stop time. The process includes the following steps:

[0069] S1: Generate a set of arrival times of urban rail trains to be updated and a set of departure times of urban rail trains to be updated according to the initial values of the arrival times and departure times of urban rail trains and the preset value range.

[0070] S2: Based on the initial value of the comprehensive optimization index of the urban rail traction power supply system and the initial value of the allowable delay time limit index of all urban rail trains on the operating line, calculate the comprehensive optimization index of the urban rail traction power supply system to be updated and the allowable delay time limit index of all urban rail trains on the operating line to be updated according to the initial values of the arrival times and departure times of urban rail trains, the set of arrival times of urban rail trains to be updated, and the set of departure times of urban rail trains to be updated. Set the same weight for the comprehensive optimization index of the urban rail traction power supply system to be updated and the allowable delay time limit index of all urban rail trains on the operating line to be updated, establish a normalized objective function, and form a power supply optimization model.

[0071] S3: Calculate the allowable delay time limit index of the train to be updated during operation in an interval and the traction energy consumption index of the train to be updated during operation in an interval according to the number of passengers and the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated, and set a screening function for screening the minimum values of the two indexes. The screening function for screening the minimum values of the two indexes forms a passenger flow optimization model.

[0072] S4: A host obtains the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated and sends them to multiple slaves; after receiving, the multiple slaves calculate the power supply optimization model and the passenger flow optimization model in parallel, iteratively optimize the data in the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated, obtain the optimal set of train tracking intervals and the optimal set of train stop times, and adjust the train tracking interval and stop time according to the two optimal sets to complete the operation scheduling of urban rail trains.

[0073] Through the above steps, by constructing a power supply optimization model and a passenger flow optimization model, the optimization objective of the traction power supply system is combined with the optimization of train operation time, which not only effectively reduces the traction energy consumption, but also significantly improves the rationality and flexibility of the operation plan. The optimization efficiency is greatly improved through parallel computing, adapting to the complex and changeable operation conditions of the urban rail system. At the same time, the obtained optimal tracking interval and stop time can meet the passenger flow demand and improve the travel experience of passengers. By adopting the method of one master computer and multiple slave computers for parallel operation calculation, the optimization speed is greatly accelerated, and the optimization efficiency of urban rail train operation dispatching is improved to meet the actual needs.

[0074] The weights of the traction energy consumption index and the delay time index can be adjusted according to different line conditions and passenger flow characteristics to adapt to specific power supply energy consumption priority or punctuality priority strategies.

[0075] According to the actual operation situation of the train, the number of iterations can be dynamically adjusted or real-time data can be introduced to update the optimization process to further improve the applicability.

[0076] According to actual needs, a preset value range can be set to limit the data sizes in the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated.

[0077] In some of these embodiments, as Figure 2 shown, S2 includes:

[0078] S201, according to the active power under the rectification condition, the active power under the inversion condition, and the power fed back by the main substation, through power flow calculation, obtain the traction energy consumption of the train in the rectification condition, the feedback energy in the inversion condition, and the power fed back by the main substation;

[0079] S202, according to the initial values of the arrival times of urban rail trains and the initial values of the departure times of urban rail trains, the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated, the traction energy consumption of the train in the rectification condition, the feedback energy in the inversion condition, and the power fed back by the main substation, calculate the comprehensive optimization index of the urban rail traction power supply system to be updated;

[0080] S203, according to the initial values of the arrival times of urban rail trains and the initial values of the departure times of urban rail trains, the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated, the number of trains and the number of stations, calculate the allowable delay time limit index of all urban rail trains on the operation line to be updated;

[0081] S204. Based on the initial values of the comprehensive optimization indexes of the urban rail traction power supply system and the initial values of the allowable delay time limit indexes of all urban rail trains on the operating line, as well as the comprehensive optimization indexes of the to-be-updated urban rail traction power supply system and the allowable delay time limit indexes of all urban rail trains on the to-be-updated operating line, a normalized objective function is established to obtain a power supply optimization model.

[0082] Through the above steps, considering the energy flow characteristics of the power supply system, the optimization results are more in line with the actual working conditions. By dynamically calculating the traction energy consumption and feedback energy, not only can the energy loss be minimized to the greatest extent, but also the energy feedback efficiency of the main substation can be improved, and at the same time, the reasonable arrangement of the operation time can be realized. Considering the energy flow characteristics of the power supply system, the optimization results are more in line with the actual working conditions. By dynamically calculating the traction energy consumption and feedback energy, not only can the energy loss be minimized to the greatest extent, but also the energy feedback efficiency of the main substation can be improved, and at the same time, the reasonable arrangement of the operation time can be realized.

[0083] In some of these embodiments, the set of arrival times of the to-be-updated urban rail trains at stations includes the times when the to-be-updated train i arrives at each station, and the set of departure times of the to-be-updated urban rail trains from stations includes the times when the to-be-updated train i departs from each station; the comprehensive optimization indexes of the to-be-updated urban rail traction power supply system are obtained through a first calculation formula, and the first calculation formula includes:

[0084]

[0085] where f1 is the comprehensive optimization index of the to-be-updated urban rail traction power supply system, is the traction energy consumption when the train is in the rectification working condition, is the feedback energy when the train is in the inversion working condition, is the power sent back by the main substation, is the initial value of the time when train i arrives at each station, are respectively the initial values of the times when train i departs from each station, is the time when the to-be-updated train i arrives at each station, is the time when the to-be-updated train i departs from each station.

[0086] The relationship between the arrival and departure times of the train and the traction energy consumption, feedback energy, and the power sent back by the main substation is quantified. By introducing variables of the to-be-updated time, the calculation accuracy of the optimization index is further improved. Through this calculation formula, the energy states of each station and each train at a specific moment can be comprehensively analyzed, and the timeliness and accuracy of the calculation results of the optimization index can be improved. At the same time, this formula can be applied to different line and train conditions and has strong versatility.

[0087] where, It is the traction energy consumption when the train is in the rectification mode, that is, the sum of all active power degrees of traction in the rectification mode throughout the line. It is the feedback energy when the train is in the inversion mode, that is, the sum of all active power degrees of traction in the inversion mode throughout the line. It is the power sent back by the main substation, that is, the power sent back by the main substations throughout the line. N r It is the number of traction substations throughout the line, N m It is the number of main substations throughout the line, T r,rec It is the duration of the rectification mode of the r-th traction substation, T r,inv It is the duration of the inversion mode of the r-th traction substation, τ is the simulation duration, P r,rec P(t) is the active power of the r-th traction substation under the rectification mode at time t, P r,inv P(t) is the active power of the r-th traction substation under the inversion mode at time t, P m P(t) is the power sent back by the m-th main substation at time t. P r,rec P(t), P r,inv P(t), P m P(t) are all obtained by power flow calculation. The power flow calculation adopts AC-DC alternating iteration, the DC uses the node voltage method, and the AC uses the Newton-Raphson method.

[0088] In some of these embodiments, the allowable delay time limit index for all urban rail trains on the operation line to be updated is obtained through a second calculation formula, and the second calculation formula includes:

[0089]

[0090] Among them, f2 is the allowable delay time limit index for all urban rail trains on the operation line to be updated, It is the initial arrival time of train i at station s, It is the initial departure time of train i from station s, It is the updated arrival time of train i at station s, It is the updated departure time of train i from station s, N is the total number of stations, and M is the total number of trains.

[0091] Through the second calculation formula, the difference between the optimized and non-optimized arrival and departure times of the train is quantified to calculate the overall delay time limit index on the line. The variables of the number of stations and trains are introduced in the formula, fully considering the impact of line complexity on time limits. Through this formula, the time optimization effect of train operation can be accurately quantified to ensure that the punctuality rate of the train is not affected while optimizing energy consumption. The introduction of the delay time limit index provides more comprehensive constraint conditions for the optimization model.

[0092] In some of these embodiments, the power supply optimization model includes:

[0093]

[0094] Among them, f 1,0 is the initial value of the comprehensive optimization index of the urban rail transit traction power supply system, and f 2,0 is the initial value of the allowable delay time limit index for all urban rail trains on the operating line.

[0095] Through the power supply optimization model, the comprehensive optimization index and the delay time limit index are normalized to form a unified optimization goal. This model synthesizes the weights of different indexes to ensure that the optimization result reaches a balance between energy consumption optimization and punctuality. Through normalization, this model can make a reasonable trade-off between different optimization goals, improving the efficiency and robustness of the optimization calculation. The adjustable parameter design of the model makes it applicable to the line optimization requirements of different scales and complexities.

[0096] In the power supply optimization model, this optimization model is a multi-objective optimization model, and the algorithm used is a single-objective optimization algorithm. Therefore, the multi-objective optimization is transformed into a single-objective optimization by normalization. By dividing the two objectives, the comprehensive optimization index f1 and the delay time limit index f2, by their respective reference values, respectively, and converting them into summation of the same order of magnitude, the objective function synthesizes the two optimization goals.

[0097] f 1,0 has the same calculation formula as the first calculation formula and is obtained through power flow calculation. f 2,0 has the same calculation formula as the second calculation formula and is obtained by calculating the second calculation formula. 1,0 Although f and f1 have the same calculation formula, the parameters used in the formula are different. Similarly, for f 2,0 and f2, although the calculation formulas are the same, the parameters used in the formula are also different. For example, in the Nth optimization, f 1,0 and f 2,0 are f1 and f2 calculated in the (N - 1)th time respectively, and f1 and f2 obtained from the (N - 2)th optimization are used in the calculation processes of f 1,0 and f 2,0 .

[0098] The power supply optimization model also sets constraint conditions, including:

[0099]

[0100] Among them, is the minimum stop time of train i at station s, is the maximum stop time of train i at station s, is the minimum tracking interval between train i and train i + 1, is the maximum tracking interval between train i and train i + 1, Δh is the secondary adjustment amount of the tracking interval, Δh ∈ [0, 5], T i,s,s+1 is the running time of train i in the section s, s + 1, is the minimum value of the total stop time of all trains at each station, is the maximum value of the total stop time of all trains at each station, is the minimum value of the total tracking interval of all trains, is the maximum value of the total tracking interval of all trains. S = {1, 2, …, s, …, N} is the set of stations, and I = {1, 2, …, i, M} is the set of trains.

[0101] The train tracking interval is the interval time between two trains running in the same direction on the same operating line, between the trailing train and the leading train.

[0102] Among them, D i is the delay time of train i, is the minimum value of the delay time of train i, is the maximum value of the delay time of train i, D total is the total sum of the delay times of all trains, is the minimum value of the total sum of the delay times of all trains, is the maximum value of the total sum of the delay times of all trains.

[0103] In some application scenarios, the maximum and minimum values of the parameters in the formula are set according to actual requirements, and a certain value is added or subtracted on the original basis. T i,s is the stop time of the train at the station, set to 30s, then is the minimum value of the sum of the stop time delays of the whole line, set that a station can be delayed by at most 5s, then is the number of stations multiplied by 5s. is the minimum tracking interval between train i and train i + 1, set the departure interval to 180s, then is the minimum value of the sum of the departure interval delays of the whole line, set that the minimum departure interval delay time for a train to follow the previous train is 10s, then it is the number of trains multiplied by 10s, then is the number of trains multiplied by a value exceeding 10s.

[0104] In some of these embodiments, the delay time limit indicators for the train to be updated during operation in an interval include:

[0105] f T,i,g = |T i,g (λ i,g , v i,g ) - T i,g |;

[0106] Wherein, λ i,g is the inert control coefficient, v i,g is the target speed of train i running in section g, T i,g is the planned running time of train i in the g-th section, T i,g (λ i,g , v i,g ) is the actual time of train i running in the g-th section at λ i,g , v i,g .

[0107] The inert control coefficient changes during operation. According to engineering experience values, its range is 0 - 0.25.

[0108] The planned running time can be found according to the planned operation diagram given by the subway side.

[0109] By the delay time limit index, the time difference of train running in a section is quantified. Variables such as target speed and planned running time are introduced in the formula, reflecting the running state at different speeds and times. This index can accurately reflect the punctuality and energy consumption optimization characteristics of train operation, making the optimization of urban rail train operation scheduling more accurate. By controlling the delay time, abnormal situations during operation can be significantly reduced.

[0110] In some of these embodiments, the traction energy consumption index of the train to be updated during operation in a section includes:

[0111]

[0112] Wherein, μ i,g is the traction force utilization coefficient of train i in section g, 0 ≤ u i,g ≤ 1, M i,g is the vehicle weight of train i during operation in section g, M i,g is related to the number of passengers, f max,i,g (v i,g , M i,g ) is the maximum traction force when the train speed is vi,g and the vehicle weight is Mi,g, η i is the electromechanical efficiency of train i, is the time when train i arrives at station s + 1, is the time when train i leaves station s. The passenger flow is detected and provided by the subway side of the station. The vehicle weight of train i during operation in section g = empty vehicle weight + average weight of adults * number of passengers of train i during operation in section g.

[0113] The calculation formula of the traction energy consumption index comprehensively considers parameters such as the speed, weight, and traction force of the train, and realizes the optimal allocation of energy consumption through dynamic adjustment, further improving the accuracy of the calculation. By introducing the traction energy consumption index, the energy consumption level of the train during operation can be intuitively reflected, providing a direct energy consumption target for optimizing the model, thereby achieving the goal of energy consumption optimization.

[0114] In some of these embodiments, the passenger flow optimization model includes:

[0115] where δ max is the maximum deviation allowed between the actual running time and the planned running time, which is set according to the actual situation, v lim (l i,g ) is the speed limit at position l i,g , a i,g (t) is the acceleration of train i at time t in the g-th interval, a i,g (t - 1) is the acceleration of train i at time t - 1 in the g-th interval, Δt is the simulation interval, Δa max is the maximum value of the acceleration change rate, a i,g (t) and a i,g (t - 1) are set according to the train requirements.

[0116] In some of these embodiments, as Figure 3 shown, in S4, a host obtains the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated, and sends them to multiple slaves. It further includes:

[0117] S301, receiving the initial values of the arrival times of urban rail trains and the initial values of the departure times of urban rail trains, the preset value range, the initial value of the comprehensive optimization index of the urban rail traction power supply system, the initial value of the allowable delay time limit index of all urban rail trains on the operation line, and the number of passengers;

[0118] S302, according to the initial values of the arrival times of urban rail trains and the initial values of the departure times of urban rail trains and the preset value range, generating the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated; generating multi-computer cluster parallel solution instructions, individual information of intelligent algorithms, and multiple calculation tasks; wherein, the parallel solution instructions include the number of slaves, the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated, the initial values of the arrival times of urban rail trains and the initial values of the departure times of urban rail trains, and the preset value range, and the individual information includes the initial value of the comprehensive optimization index of the urban rail traction power supply system and the initial value of the allowable delay time limit index of all urban rail trains on the operation line;

[0119] In S303, multiple computing tasks are evenly distributed and sent to multiple slave machines, and the passenger flow number, the parallel solution instruction, and the individual information are sent to the multiple slave machines.

[0120] The host is responsible for task allocation and data aggregation, which improves the adjustment efficiency of urban rail train operation scheduling. The architecture design of the master-slave cooperation enables the system to have good scalability and is suitable for large-scale optimization tasks of complex lines.

[0121] In some embodiments, as Figure 4 shown, in S4, after receiving, multiple slave machines calculate the power supply optimization model and the passenger flow optimization model in parallel, and iteratively optimize the data in the set of to-be-updated arrival times of urban rail trains at each station and the set of to-be-updated departure times of urban rail trains at each station, further including:

[0122] S401, receive partial computing tasks, the passenger flow number, the parallel solution instruction, and the individual information sent by the host; according to the order of the issued computing tasks, obtain the to-be-updated arrival time of train i at each station and the to-be-updated departure time of train i at each station from the set of to-be-updated arrival times of urban rail trains at each station and the set of to-be-updated departure times of urban rail trains at each station;

[0123] S402, calculate the to-be-updated comprehensive optimization index of the urban rail traction power supply system and the to-be-updated allowable delay time limit index of all urban rail trains on the operation line according to the initial arrival time of the urban rail train, the initial departure time of the urban rail train, the initial comprehensive optimization index of the urban rail traction power supply system, the initial allowable delay time limit index of all urban rail trains on the operation line, the corresponding to-be-updated arrival time of train i at each station, and the to-be-updated departure time of train i at each station, and run the power supply optimization model;

[0124] S403, calculate the to-be-updated allowable delay time limit index for a train to run in a section and the to-be-updated traction energy consumption index for a train to run in a section according to the passenger flow number, the initial arrival time of the urban rail train, the initial departure time of the urban rail train, the corresponding to-be-updated arrival time of train i at each station, and the to-be-updated departure time of train i at each station, and run the passenger flow optimization model;

[0125] Among them, the power supply optimization model and the passenger flow optimization model in any one slave machine operate simultaneously to obtain an optimized train tracking interval and train stop time;

[0126] S404, perform multiple iterations of optimization. During the iteration process, based on the optimized train tracking interval and train dwell time obtained from the Nth optimization, the preset value range, and the individual information of the intelligent algorithm, update the data in the set of arrival times of urban rail trains to be updated and the set of departure times of urban rail trains to be updated; according to the optimized train tracking interval and train dwell time obtained from the Nth optimization, update the initial values of the arrival times of urban rail trains, the initial values of the departure times of urban rail trains, the initial value of the comprehensive optimization index of the urban rail traction power supply system, and the initial value of the allowable delay time limit index of all urban rail trains on the operation line used in the (N + 1)th optimization; control the power supply optimization module and the passenger flow optimization module to perform the (N + 1)th optimization based on the updated data to obtain the optimized train tracking interval and train dwell time obtained from the (N + 1)th optimization.

[0127] Multiple slave machines perform parallel calculations on the power supply optimization model and the passenger flow optimization model. Through the parallel computing ability of the computer cluster, the efficiency and accuracy of the operation scheduling optimization of urban rail trains are greatly improved.

[0128] In some of these embodiments, in S4, obtain the optimal set of train tracking intervals and the optimal set of train dwell times, and based on the two optimal sets, adjust the train tracking intervals and dwell times to complete the operation scheduling of urban rail trains. Further include: set an iteration count threshold. When the iteration count reaches the iteration count threshold, stop the iterative calculation, receive the optimized train tracking interval and train dwell time obtained from the last optimization by multiple slave machines, and form the optimal set of train tracking intervals and the optimal set of train dwell times, and based on the two optimal sets, adjust the train tracking intervals and dwell times to complete the operation scheduling of urban rail trains.

[0129] By setting the iteration count threshold, it is ensured that the optimization process can be completed within a reasonable time, while avoiding excessive consumption of computing resources. After receiving the final optimization result, the host summarizes and adjusts the train operation scheduling. Through the setting of the iteration count threshold, the complexity of the calculation process is effectively controlled, ensuring that the optimization result has high timeliness and usability. At the same time, the finally formed optimal set can be directly used for the operation scheduling of urban rail trains, enhancing the value of practical applications.

[0130] In some of these embodiments, in one optimization, the power supply optimization model and the passenger flow optimization model in any slave machine operate simultaneously to obtain an optimized train tracking interval, train dwell time, and system energy consumption. The system energy consumptions calculated by multiple slave machines are compared to obtain the minimum system energy consumption, and the corresponding train tracking interval and train dwell time are the optimal train tracking interval and the optimal train dwell time.

[0131] By further determining the optimal result through the system energy consumption, the appropriate train tracking interval and the optimal train dwell time can be selected more precisely.

[0132] In some of these embodiments, the computer cluster includes N + 1 computers, and each of the computers includes a machine ID identification unit for obtaining the machine ID corresponding to each computer and identifying the code of the machine ID.

[0133] When the code of the machine ID is an integer greater than N, the computer corresponding to the machine ID is set as the host.

[0134] When the code of the machine ID is an integer greater than or equal to 0 and less than or equal to N, the computer corresponding to the machine ID is set as a slave.

[0135] Through the coding rule of the machine ID, the role assignment of the host and the slave is clarified. The host is responsible for task distribution and result aggregation, and the slave runs the optimization model independently according to the assigned tasks. The management efficiency of the computer cluster is further optimized, and the configuration complexity is reduced. The clear division of the roles of the master and slave makes the system architecture more scalable and fault-tolerant.

[0136] The MQTT message for parallel solving instructions of the multi-computer cluster can be set as:

[0137] "{\"machineid\":%d,\"maxtimeintervalopt\":%d,\"maxdwelltimeopt\":%d,\"curren ttimeinterval\":%d,\"currentdwelltime\":%d,\"popsize\":%d,\"referencesystemenergy\":%.2f,\"isparallelizeopt\":%d,\"isreferencesystem\":%d,\"ismmcps\":%d}"。

[0138] Among them, machineid is the ID value of the slave, maxtimeintervalopt is the departure interval optimization threshold, maxdwelltimeopt is the dwell time optimization threshold, currenttimeinterval is the departure interval of the current train operation diagram, currentdwelltime is the dwell time of the current train operation diagram, popsize is the number of population individuals, referencesystemenergy is the reference system energy consumption, isparallelizeopt is whether it is parallel optimization, and isreferencesystem is whether it is a reference system.

[0139] The MQTT message of the intelligent algorithm individual information can be set as: {"MACHINEID": %d, "RESULT": "true", "TIMEINTERVAL": [%.2f, …, %.2f], "DWELLTIME": [%.2f, …, %.2f], "UNIFIEDOBJECTIVE": %.2f, "ENERGYCONSUMPTION": %.2f, "MULTITRAINTOTALDELAYTIME": %.2f}".

[0140] Among them, MACHINEID is the ID of the slave, RESULT represents whether the message is sent successfully. When it is true, it means the sending is successful. TIMEINTERVAL is the tracking interval of all stations on the line, DWELLTIME is the stop time of all stations on the line, and UNIFIEDOBJECTIVE is the normalized objective function. ENERGYCONSUMPTION is the comprehensive optimization index f1 of the urban rail traction power supply system to be updated, and MULTITRAINTOTALDELAYTIME is the allowable delay time limit index f2 of all urban rail trains on the operating line to be updated.

[0141] In the mqttconnect.ini configuration file, the server field identifies the address of the host in the local area network, the database field identifies the address of the host database, the clientid field identifies the unique device ID of the local computer in the MQTT communication network, and the machineid field identifies the machine ID of the local computer.

[0142] When establishing an MQTT server on one host and multiple slaves, the mosquitto software can be used to build the MQTT server. Modify the configuration file mosquitto.conf to include: allow_anonymous true, which allows different computers under the same local area network to access this machine anonymously through MQTT. listener 1883, which opens the listening port number 1883. This port is used to send and receive MQTT messages.

[0143] In an actual application scenario, the configuration file mqttconnect.ini of 1 host and N slave machines is read simultaneously. N ≥ 3. In the mqttconnect.ini configuration file, the server field identifies the address of the host in the local area network, the database identifies the address of the host database, the clientid identifies the unique device ID of the local computer in the MQTT communication network, and the machineid identifies the machine ID of the local computer. If the value of machineid is any integer in [0, N], the local computer is one of the N slave machines. If the value of machineid is an integer greater than N, the local computer is 1 host.

[0144] When the local computer is the host, 1 host uses a unique clientid to establish an MQTT connection with N slave machines, and uses a unique machineid to send MQTT messages of multi-computer cluster parallel solution instructions and MQTT messages of intelligent algorithm individual information to N slave machines. When the local computer is a slave machine, N slave machines use different clientids to establish an MQTT connection with 1 host. N slave machines use different machineids to receive MQTT messages of multi-computer cluster parallel solution instructions and MQTT messages of intelligent algorithm individual information sent by 1 host, and send the optimized solution set, that is, the intelligent algorithm individuals that have calculated the objective function, to 1 host. N slave machines form a computer cluster and complete M parallel solution tasks, accelerating the solution speed and saving simulation time. 1 host outputs the optimal solution set {h best ,T best}, where h is the tracking interval between multiple trains, and T is the stop time of the train at each station.

[0145] One host assigns M subtasks to N slave machines. N slave machines evenly divide M for calculation. The subtasks assigned to the first slave machine are from 1 to M / N, the subtasks assigned to the i-th slave machine are from i*M / N to (i + 1)*M / N, and the subtasks assigned to the N-th slave machine are from (N - 1)*M / N to N*M / N.

[0146] The numbers of h and T are both related to the number of stations. During the iterative optimization process, the number of randomly generated data is related to the population size of the intelligent algorithm. After the iterative optimization ends, the number of the optimal solution set is determined and related to the number of stations. The number of iterations is set to 500 times. After the number of iterations is completed, the optimal solution is output.

[0147] It should be noted that the steps shown in the above process or the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0148] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as falling within the scope described in this specification.

[0149] The above-described embodiments only represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the appended claims.

Claims

1. A rail transit scheduling method based on parallel offline solution of multiple computer clusters, characterized in that: The urban rail train operation scheduling includes adjusting the train tracking interval and the stop time, and the method includes: S1: Generate a set of arrival times of urban rail trains to be updated and a set of departure times of urban rail trains to be updated according to the initial values ​​of arrival times of urban rail trains and the initial values ​​of departure times of urban rail trains and a preset value range; S2: Based on the initial value of the comprehensive optimization index of the urban rail traction power supply system and the initial value of the allowable delay time limit index of all urban rail trains on the running line, according to the initial value of the arrival time of the urban rail train and the initial value of the departure time of the urban rail train, the arrival time set of the urban rail train to be updated and the departure time set of the urban rail train to be updated, the comprehensive optimization index of the urban rail traction power supply system to be updated and the allowable delay time limit index of all urban rail trains on the running line to be updated are calculated, the comprehensive optimization index of the urban rail traction power supply system to be updated and the allowable delay time limit index of all urban rail trains on the running line to be updated are set with equal weights, a normalized objective function is established, and a power supply optimization model is formed; S3: according to the number of passengers and the set of arrival time of urban rail trains to be updated and the set of departure time of urban rail trains to be updated, calculate the delay time limit index of the train to be updated running in a section and the traction energy consumption index of the train to be updated running in a section, and set a screening function for screening the minimum values ​​of the two indicators, and the screening function for screening the minimum values ​​of the two indicators constitutes a passenger flow optimization model; S4: A host obtains a set of urban rail train arrival times to be updated and a set of urban rail train departure times to be updated, and sends them to multiple slaves; after receiving them, multiple slaves calculate the power supply optimization model and the passenger flow optimization model in parallel, iteratively optimize the data in the set of urban rail train arrival times to be updated and the set of urban rail train departure times to be updated, obtain the optimal set of train tracking intervals and the optimal set of train stop times, and adjust the train tracking intervals and stop times according to the two optimal sets to complete the operation scheduling of urban rail trains.

2. The rail transit scheduling method based on multi-computer cluster parallel offline solution according to claim 1 is characterized in that: The S2 includes: According to the active power under rectification condition, the active power under inverter condition and the return power of the main substation, the traction energy consumption of the train under rectification condition, the feedback energy under inverter condition and the return power of the main substation are obtained through power flow calculation; According to the initial value of the arrival time of the urban rail train and the initial value of the departure time of the urban rail train, the set of arrival time of the urban rail train to be updated and the set of departure time of the urban rail train to be updated, the traction energy consumption of the train in the rectification condition, the feedback energy of the train in the inverter condition and the return power of the main substation, the comprehensive optimization index of the urban rail traction power supply system to be updated is calculated; According to the initial value of the urban rail train arrival time and the initial value of the urban rail train departure time, the urban rail train arrival time set to be updated and the urban rail train departure time set to be updated, the number of trains and the number of stations, the allowable delay time limit index of all urban rail trains on the line to be updated is calculated; According to the initial values ​​of the comprehensive optimization index of the urban rail traction power supply system and the initial values ​​of the allowable delay time limit index of all urban rail trains on the operating lines, the comprehensive optimization index of the urban rail traction power supply system to be updated and the allowable delay time limit index of all urban rail trains on the operating lines to be updated, a normalized objective function is established to obtain the power supply optimization model.

3. The rail transit scheduling method based on multi-computer cluster parallel offline solution according to claim 2 is characterized in that: The urban rail train arrival time set to be updated includes the time when the train i to be updated arrives at each station, and the urban rail train departure time set to be updated includes the time when the train i to be updated leaves each station; The comprehensive optimization index of the urban rail traction power supply system to be updated is obtained by a first calculation formula, which includes: Among them, f1 is the comprehensive optimization index of the urban rail traction power supply system to be updated. is the traction energy consumption of the train in rectification condition, is the feedback energy of the train in inverter condition, The amount of electricity returned to the main substation, is the initial value of the time when train i arrives at each station, are the initial values ​​of the time when train i leaves each station, is the arrival time of train i at each station to be updated, is the time when the train i to be updated leaves each station.

4. The rail transit scheduling method based on multi-computer cluster parallel offline solution according to claim 3 is characterized in that: The allowable delay time limit index of all urban rail trains on the line to be updated is obtained by a second calculation formula, and the second calculation formula includes: Among them, f2 is the allowable delay time limit indicator of all urban rail trains on the line to be updated. is the initial value of the time when train i arrives at station s, is the initial value of the time when train i leaves station s, is the time when train i arrives at station s to be updated, is the time when the train i to be updated leaves the station s, N is the total number of stations, and M is the total number of trains.

5. The rail transit scheduling method based on multi-computer cluster parallel offline solution according to claim 4 is characterized in that: The power supply optimization model includes: Among them, f 1,0 is the initial value of the comprehensive optimization index of the urban rail traction power supply system, f 2,0 It is the initial value of the allowable delay time limit indicator for all urban rail trains on the operating line.

6. The rail transit scheduling method based on multi-computer cluster parallel offline solution according to claim 1 is characterized in that: The delay time limit index of the train to be updated running in a section includes: f T,i,g =|T i,g (λ i,g ,v i,g )-T i,g |; Among them, λ i,g is the inertia control coefficient, v i,g is the target speed of train i in section g, T i,g is the planned running time of train i in the gth section, T i,g (λ i,g ,v i,g ) is the number of train i in the gth interval with λ i,g 、v i,g The actual time at runtime.

7. The rail transit scheduling method based on multi-computer cluster parallel offline solution according to claim 6 is characterized in that: The traction energy consumption index of the train to be updated running in a section includes: Among them, μ i,g is the traction force utilization coefficient of train i in section g, M i,g is the weight of train i when running in section g, M i,g Related to the number of passengers, f max,i,g (v i,g ,M i,g ) is the maximum traction force when the train speed is vi,g and the vehicle weight is Mi,g, η i is the electromechanical efficiency of train i.

8. The rail transit scheduling method based on multi-computer cluster parallel offline solution according to claim 1 is characterized in that: In S4, a host obtains a set of urban rail train arrival times to be updated and a set of urban rail train departure times to be updated, and sends them to multiple slaves, further comprising: Receive the initial value of the arrival time of the urban rail train and the initial value of the departure time of the urban rail train, the preset value range, the initial value of the comprehensive optimization index of the urban rail traction power supply system, the initial value of the allowable delay time limit index of all urban rail trains on the running line, and the number of passengers; According to the initial value of the arrival time of the urban rail train and the initial value of the departure time of the urban rail train and the preset value range, the set of arrival time of the urban rail train to be updated and the set of departure time of the urban rail train to be updated are generated; a multi-computer cluster parallel solution instruction and intelligent algorithm individual information and multiple computing tasks are generated; wherein the parallel solution instruction includes the number of slave machines, the set of arrival time of the urban rail train to be updated and the set of departure time of the urban rail train to be updated, the initial value of the arrival time of the urban rail train and the initial value of the departure time of the urban rail train and the preset value range, and the individual information includes the initial value of the comprehensive optimization index of the urban rail traction power supply system and the initial value of the allowable delay time limit index of all urban rail trains on the running line; The multiple computing tasks are evenly distributed and sent to the multiple slave machines, and the number of passengers, the parallel solution instructions and the individual information are sent to the multiple slave machines.

9. The rail transit scheduling method based on multi-computer cluster parallel offline solution according to claim 8 is characterized in that: In the above S4, after receiving, multiple slaves calculate the power supply optimization model and the passenger flow optimization model in parallel, iteratively optimize the data in the set of arrival time of urban rail trains to be updated and the set of departure time of urban rail trains to be updated, further comprising: Receiving part of the computing tasks, the number of passengers, the parallel solution instruction and the individual information issued by a host; obtaining the corresponding arrival time of train i at each station and the departure time of train i from each station from the set of urban rail train arrival time and the set of urban rail train departure time to be updated in the order in which the computing tasks are issued; According to the initial value of the arrival time of the urban rail train, the initial value of the departure time of the urban rail train, the initial value of the comprehensive optimization index of the urban rail traction power supply system, the initial value of the allowable delay time limit index of all urban rail trains on the operating line, the corresponding arrival time of the train i to be updated at each station and the departure time of the train i to be updated at each station, the comprehensive optimization index of the urban rail traction power supply system to be updated and the allowable delay time limit index of all urban rail trains on the operating line to be updated are calculated to operate the power supply optimization model; According to the number of passengers, the initial value of the arrival time of the urban rail train, the initial value of the departure time of the urban rail train, the corresponding arrival time of the train i to be updated at each station and the departure time of the train i to be updated at each station, the delay time limit index of the train to be updated running in a section and the traction energy consumption index of the train to be updated running in a section are calculated to run the passenger flow optimization model; The power supply optimization model and the passenger flow optimization model in any slave are operated simultaneously to obtain an optimized train tracking interval and train stop time; Perform multiple iterative optimizations. During the iterative process, update the data in the set of urban rail train arrival times to be updated and the set of urban rail train departure times to be updated according to the optimized train tracking intervals and train stop times obtained from the Nth optimization, the preset value range and the individual information of the intelligent algorithm; update the initial value of the urban rail train arrival time, the initial value of the urban rail train departure time, the initial value of the comprehensive optimization index of the urban rail traction power supply system, and the initial value of the allowable delay time limit index of all urban rail trains on the operating line used in the N+1th optimization according to the optimized train tracking intervals and train stop times obtained from the Nth optimization; perform the N+1th optimization according to the updated data to obtain the optimized train tracking intervals and train stop times obtained from the N+1th optimization.

10. The rail transit scheduling method based on multi-computer cluster parallel offline solution according to claim 9, characterized in that: In the S4, an optimal set of train tracking intervals and an optimal set of train stop times are obtained, and according to the two optimal sets, the train tracking intervals and stop times are adjusted to complete the operation scheduling of urban rail trains, further comprising: setting an iteration number threshold, stopping the iterative calculation when the iteration number reaches the iteration number threshold, receiving the optimized train tracking intervals and train stop times obtained by the last optimization performed by multiple slaves, and forming an optimal set of train tracking intervals and an optimal set of train stop times, and according to the two optimal sets, adjusting the train tracking intervals and stop times to complete the operation scheduling of urban rail trains.

Images