Cooperative scheduling method for public transport vehicles and drivers

Through the coordinated scheduling method of bus vehicles and drivers, and the coordinated scheduling network and multi-charging mode optimization scheduling, the problems of vehicle and driver coordination difficulties and improper charging management in the traditional scheduling method are solved, and operational efficiency and service quality are improved.

CN118917594APending Publication Date: 2024-11-08CHINA ACAD OF TRANSPORTATION SCI +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410962185.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-18
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Traditional bus and driver dispatch methods cannot effectively coordinate vehicles and drivers, and cannot flexibly respond to road conditions and charging modes, resulting in low operational efficiency, high cost and poor service quality.

Method used

The coordinated scheduling method of bus vehicles and drivers is adopted, and by establishing a coordinated scheduling network, time expansion map, Markov train number chain and delay communication model, combining multi-charging mode and responsibility chain, scheduling decisions are optimized to reduce operational costs and improve service quality.

Benefits of technology

It has achieved flexible response to road conditions changes, improved the efficiency and punctuality of bus operations, reduced charging costs, improved service quality, and met passengers' travel needs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118917594A_ABST
    Figure CN118917594A_ABST
Patent Text Reader

Abstract

The invention discloses a public transport vehicle and driver cooperative scheduling method, and relates to the field of electric public transport vehicle scheduling and driver scheduling, and the method comprises the following steps: building a cooperative scheduling network of drivers and electric public transport vehicles in multiple parking lots, establishing a time expansion graph for the collaborative scheduling network corresponding to each parking lot; establishing a connection arc set in the time expansion graph and classifying the connection arc set; establishing an electric bus energy consumption model and a driver work model based on a responsibility chain; constructing a Markov train number chain to obtain the influence of the road condition on the train number travel time; combining a delay accumulation effect and a delay propagation effect on the train number chain to construct a delay transfer model, and designing a delay propagation chain selection algorithm; and establishing a regional collaborative scheduling general model of the driver and the electric bus. The method not only can comprehensively consider the cooperative scheduling problem of the vehicle and the driver, but also can flexibly deal with the random travel time caused by the road condition change.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of electric bus dispatching and driver dispatching, and in particular to a method for coordinated dispatching of a bus and a driver. Background Art

[0002] Today's cities are expanding rapidly, and bus operations are an important part of ensuring the travel needs of urban residents, reducing traffic congestion, and reducing environmental pollution. However, traditional bus vehicle and driver scheduling methods have many limitations, which not only increase operating costs, but also affect operating efficiency and service quality. Traditional methods often separate vehicle scheduling and driver scheduling, which leads to difficulties in coordination between vehicles and drivers. For example, the number of vehicles may be excessive in some periods and insufficient in other periods, which requires flexible optimization of vehicle and driver scheduling; in addition, existing scheduling methods usually fail to fully consider the impact of road conditions changes on operation plans, which may result in the inability to accurately execute the timetable and thus affect passengers' travel experience and trust in bus companies.

[0003] With the popularity of electric buses, charging management has become a key issue. Traditional charging management methods often only consider the charging needs of vehicles, but do not fully consider the impact of the charging process on bus operations. For example, the determination of the charging mode is often based on the charging needs of the vehicle, without considering the impact of the charging process of various charging modes on vehicle scheduling and operation plans; in addition, existing technologies often do not consider the flexible switching between different charging modes, as well as the integration between charging plans and vehicle scheduling plans, which will lead to increased charging costs and reduced operational efficiency.

[0004] In summary, the traditional bus vehicle and driver dispatching methods and charging management methods have many problems and limitations. In order to solve these problems, it is necessary to provide an integrated method that comprehensively considers vehicle and driver dispatching. This method needs to consider the impact of road conditions on the operation plan and can flexibly select different charging modes to minimize the operating costs of bus companies and ultimately ensure the efficient execution of bus operation plans. Summary of the invention

[0005] 1. Technical issues to be solved

[0006] In view of the shortcomings of the prior art, the present invention provides a method for coordinated scheduling of public buses and drivers, which has the advantage of comprehensively considering the coordinated scheduling of vehicles and drivers, thereby solving the problem of not considering the impact of the charging process of various charging modes on vehicle scheduling and operation plans.

[0007] (II) Technical solution

[0008] In order to achieve the advantages of comprehensively considering the coordinated scheduling problem of vehicles and drivers, the specific technical solutions adopted by the present invention are as follows:

[0009] A method for coordinating the dispatching of a public transport vehicle and a driver, the method comprising the following steps:

[0010] Establish a coordinated dispatching network between drivers and electric buses in multiple parking lots, and establish a time expansion diagram for the coordinated dispatching network corresponding to each parking lot;

[0011] Establish and classify the connection arc sets in the time expansion graph, and set cost functions for different types of connection arcs in the collaborative scheduling network;

[0012] Use the collaborative dispatching network to connect and limit the coverage of each bus under multiple parking lots; establish an electric bus energy consumption model and a driver work model based on the responsibility chain;

[0013] Construct a Markov train chain to obtain the impact of road conditions on train travel time; construct a delay transfer model based on the delay accumulation effect and delay propagation effect on the train chain, and design a delay propagation chain selection algorithm;

[0014] Based on the delay propagation chain selection algorithm and the scheduling chain generation algorithm, the overall model of regional collaborative scheduling between drivers and electric buses is established and output to complete the scheduling process.

[0015] Furthermore, establishing a coordinated dispatching network between drivers and electric buses in multiple parking lots includes:

[0016] Based on the regional dispatch characteristics of drivers and electric buses, a corresponding collaborative dispatch network layer of drivers and electric buses is established for each depot, and its expression is:

[0017]

[0018] N k =o k ∪s k ∪T k

[0019] In the formula, G k Represents the collaborative dispatching network layer between drivers and electric buses;

[0020] o k represents the departure station in the current layer network k;

[0021] N k represents the set of train numbers and parking lot nodes in the current layer network k;

[0022] s k represents the departure station in the current layer network k;

[0023] T k represents the set of tasks that all electric buses and drivers belonging to the current layer network k should complete;

[0024] A k Represents a valid connection formed between two nodes in the connection arc set.

[0025] Furthermore, establishing and classifying a set of connection arcs in the time extension graph, and setting cost functions for different types of connection arcs in the collaborative scheduling network include the following steps:

[0026] Define the key parameters of each train in multiple yards to ensure that the time and location logic between trains meet the constraints;

[0027] The connection arcs are classified according to the key parameters of each train and the execution activities of the connection arc set in the time expansion diagram, and a cost function is set for each type of connection arc.

[0028] Furthermore, the types of connection arcs include circulation arcs, exit arcs, entry arcs, service arcs, waiting arcs and empty arcs;

[0029] Among them, the loop arc has no actual scheduling significance and is used to maintain the flow balance in the coordinated scheduling network. The starting point of the loop arc is the receiving yard and the end point is the departure yard.

[0030] The departure arc indicates that the electric bus starts to perform service from the depot. The starting point of the departure arc is the departure depot where the electric bus is currently located, and the end point is any departure station.

[0031] The entry arc indicates that the electric bus returns to the depot to end its service. The starting point of the entry arc is any receiving station, and the end point is the receiving depot where the current electric bus is located.

[0032] The service arc means that the electric bus runs two trips on the timetable continuously without waiting or idling in between.

[0033] The waiting arc indicates that the electric bus is waiting for the next trip. The starting point of the waiting arc is the stop of the previous trip, and the end point is the departure point of the next trip.

[0034] The idle arc represents the idle running activity of electric buses between different stops. The starting point of the idle arc is the arrival stop of the previous bus, and the end point is the departure stop of the next bus.

[0035] Furthermore, establishing an electric bus energy consumption model includes the following steps:

[0036] Collect the initial battery status of each electric bus before starting service, and dynamically adjust the energy consumption forecast and charging requirements of each electric bus on the route based on the initial battery status;

[0037] Obtain the battery energy status, set the power status limit and set the upper and lower limits of the battery energy status to ensure that the power status of each electric bus remains within the optimal range during operation;

[0038] The charging time of the electric bus under each battery charging mode is calculated according to the required supplementary power of the electric bus, and an electric bus energy consumption model is established to constrain the battery charging and discharging state. At the same time, the energy consumption of each electric bus is predicted through the electric bus energy consumption model.

[0039] Furthermore, establishing a driver work model based on the responsibility chain includes the following steps:

[0040] Build a driver work model that maps the driver's work behavior; connect various dispatch nodes based on the preset responsibility chain to clarify the division of responsibilities between the driver and the electric bus;

[0041] Use the driver work model to dynamically control the driver's work behavior and manage and enforce the constraints on the driver's work behavior;

[0042] Among them, the driver's work behavior includes the driver's working hours, the driver's continuous driving time, and the driver's rest and dining behavior.

[0043] Furthermore, a Markov train chain is constructed to obtain the impact of road conditions on train travel time; a delay transfer model is constructed by combining the delay accumulation effect and delay propagation effect on the train chain, and a delay propagation chain selection algorithm is designed, including the following steps:

[0044] The travel time between trains is regarded as a random variable, and the time interval between trains is used to estimate the probability distribution of the random variable to construct a Markov train chain.

[0045] The Markov train chain is used to calculate the delay transmission probability from the current train to the next train, and the impact of road conditions on the train travel time is analyzed based on the delay transmission probability;

[0046] The delay expectation calculation technology is combined with the cumulative effect of delays on the train chain to obtain the probability of waiting time under different delay times, and a delay transfer model is constructed based on the delay propagation effect and occurrence probability.

[0047] The delay transfer model is used to determine the actual probability distribution of each train number, and the probability distribution of the train number is calculated by integrating it with the impact results of the train number's travel time. The maximum value of the probability distribution is selected as the basis for designing the delay propagation chain selection algorithm.

[0048] Furthermore, establishing a regional coordinated dispatching model of drivers and electric buses includes the following steps:

[0049] Define the cost parameters of the overall model of regional cooperative dispatch of drivers and electric buses;

[0050] Define the constraints of the overall model of regional coordinated dispatch of drivers and electric buses;

[0051] Based on cost parameters and constraints, a general model of regional collaborative scheduling of drivers and electric buses is established.

[0052] Furthermore, the cost parameters include bus vehicle costs, driver salary costs, and penalty costs;

[0053] The constraints include vehicle dispatch constraints, electric bus range constraints, driver dispatch constraints, and delay and waiting time constraints.

[0054] Furthermore, the cost function of each type of connection arc is expressed as:

[0055]

[0056] In the formula, represents the vehicle dispatching cost of the connecting arc (i, j) in parking lot k; C f is the fixed cost per vehicle; C m Represents the unit vehicle driving cost; C w represents the unit vehicle waiting cost; Indicates the departure time. Indicates the time of car collection; is the stop for train number i To the departure station of train number j The idle time of arc (i, j) is the waiting time of arc (i, j).

[0057] (III) Beneficial effects

[0058] Compared with the prior art, the present invention provides a method for coordinated dispatching of public transport vehicles and drivers, which has the following beneficial effects:

[0059] (1) The method for coordinated dispatching of public transport vehicles and drivers provided by the present invention can not only comprehensively consider the coordinated dispatching of vehicles and drivers, but also flexibly respond to the random travel time caused by changes in road conditions, thereby ensuring the efficiency and punctuality of public transport operations. As a result, public transport operators can more flexibly arrange the operation plans of vehicles and drivers, minimize charging costs, improve operational efficiency, improve service quality, and meet the travel needs of passengers.

[0060] (2) By constructing a time extension network and a multi-layer time graph, the present invention can more accurately simulate and predict the scheduling of vehicles and drivers, and set corresponding costs for different types of connection arcs in the network, so that scheduling decisions can be optimized based on cost-effectiveness. At the same time, by considering multiple charging modes and driver responsibility chains, the scheduling method can flexibly respond to various emergencies, so that delays can be effectively predicted and managed, reducing their negative impact on the entire network. Through accurate scheduling models and algorithms, the reliability and time accuracy of bus services can be guaranteed, thereby improving passenger satisfaction and overall service quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0062] Figure 1 is a flow chart of a method for coordinated dispatching of public transport vehicles and drivers according to an embodiment of the present invention;

[0063] Figure 2 is a diagram describing the problem of coordinated dispatching of vehicles and drivers considering delay propagation under multiple depots in a coordinated dispatching method of public buses and drivers according to an embodiment of the present invention;

[0064] Figure 3 It is a multi-type arc description diagram under a multi-layer time extension network in a method for cooperative scheduling of public transportation vehicles and drivers according to an embodiment of the present invention;

[0065] Figure 4 is a driver time expansion scheduling network diagram based on the responsibility chain in the public transportation vehicle and driver collaborative scheduling method according to an embodiment of the present invention;

[0066] Figure 5 is a delay impact diagram of multiple scenarios in a method for cooperative dispatching of public buses and drivers according to an embodiment of the present invention;

[0067] Figure 6 is a definition diagram of delay time and waiting time in a method for coordinated dispatching of public transport vehicles and drivers according to an embodiment of the present invention;

[0068] Figure 7 It is a delay impact diagram on the responsibility chain in the method for coordinated dispatching of public buses and drivers according to an embodiment of the present invention. DETAILED DESCRIPTION

[0069] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in the field should be able to understand other possible implementations and advantages of the present invention. The components in the figures are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0070] According to an embodiment of the present invention, a method for coordinated dispatching of public transportation vehicles and drivers is provided.

[0071] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the method for coordinated dispatching of a public transport vehicle and a driver according to an embodiment of the present invention, the method for coordinated dispatching of a public transport vehicle and a driver comprises the following steps:

[0072] S1. Establish a collaborative dispatching network between drivers and electric buses in multiple parking lots, and establish a time expansion diagram for the collaborative dispatching network corresponding to each parking lot.

[0073] Among them, the establishment of a coordinated dispatching network for drivers and electric buses in multiple parking lots includes:

[0074] Based on the regional dispatch characteristics of drivers and electric buses, a corresponding collaborative dispatch network layer of drivers and electric buses is established for each depot, and its expression is:

[0075]

[0076] Where N k represents the set of train and parking lot nodes in the current layer network k, which contains three types of nodes, namely N k =o k ∪s k ∪T k , where G k Represents the collaborative dispatching network layer between drivers and electric buses; o k represents the departure station in the current layer network k; s k represents the departure station in the current layer network k; T k is the set of all bus trips i in a single day, and is also the most important component of the node, representing the set of tasks that all electric buses and drivers belonging to the current layer network k should complete; A k It represents a valid connection between two nodes in the connection arc set, that is, there is a corresponding vehicle or driver serving bus number i and bus number j continuously. In particular, the so-called "valid" means that there is enough time between the two corresponding service buses to meet the conditions for the bus or driver to continue.

[0077] S2. Establish and classify a set of connection arcs in the time expansion graph, and set cost functions for different types of connection arcs in the collaborative scheduling network.

[0078] Among them, establishing and classifying a set of connection arcs in the time extension graph and setting cost functions for different types of connection arcs in the collaborative scheduling network include the following steps:

[0079] Define the key parameters of each train in multiple yards to ensure that the time and location logic between trains meet the constraints.

[0080] It should be noted that for any train node It contains four parameters, showing the corresponding specific train information; among them, is the departure station (s means start), is the departure time, accordingly, is the stop for the vehicle (e means start), For the closing time, they must follow the following time restrictions:

[0081]

[0082] In the formula, for any parking lot Any arc of In terms of Indicates the time of pickup. Indicates the stop for train number i To the departure station of train number j of idle time (dh means "idle").

[0083] The connection arcs are classified according to the key parameters of each train and the execution activities of the connection arc set in the time expansion diagram, and a cost function is set for each type of connection arc.

[0084] Among them, the type of connecting arc (i, j) ∈ A k Including circulation arc, exit arc, entry arc, service arc, waiting arc and empty arc;

[0085] It should be noted that the loop arc: This train has no practical significance and is added for the sake of structural integrity. It can balance the flow at any point in the diagram. The starting point is the receiving yard s k , the end point is the departure yard o k .

[0086] Departure arc: indicates that the bus starts service from the parking lot, and the starting point is the departure lot where the vehicle is located. k , the destination can be any departure station. Each train can correspond to a departure arc, and this train is selected for execution.

[0087] Entry arc: indicates that the bus returns to the depot and ends its service, and the end point is the depot where the vehicle is located. k , the starting point can be any train stop. Each train can correspond to an entry arc, which is selected for execution.

[0088] Service arc: It means that the vehicle executes two trips on the timetable consecutively without waiting or idle driving in between.

[0089] Waiting arc: indicates that the vehicle is waiting for the next train. This arc is generated only when the departure point is the same as the arrival point of the previous train, but there is still a gap between the departure time and the arrival time of the previous train.

[0090] Empty arc: represents the empty running activity of the vehicle, which can realize the scheduling of the vehicle between different stations; the starting point is the receiving station of the previous train, and the end point is the departure station of the next consecutive train. Unlike the waiting arc, the empty arc requires that the departure station and the receiving station are not the same actual station.

[0091] Among them, the expression of the cost function of each type of connection arc (i, j) is:

[0092]

[0093] In the formula, represents the vehicle dispatching cost of the connecting arc (i, j) in parking lot k; the fixed cost is mainly related to the number of vehicles used, C f (f stands for “fixed”) is the fixed cost per vehicle (yuan / vehicle); the driving cost includes fuel costs, maintenance costs and other costs incurred by vehicle driving, C m (m stands for “driving”) represents the unit vehicle driving cost (yuan / minute); waiting cost is the penalty imposed by humans for wasting time while waiting for the next bus, C w (w stands for “waiting”) represents the unit vehicle waiting cost (yuan / minute); Indicates the departure time. Indicates the time of car collection; is the stop for train number i To the departure station of train number j The idle time of arc (i, j) is the waiting time of arc (i, j).

[0094] It should be noted that the exit arc represents a bus starting from the depot to run a trip, so it is necessary to calculate the vehicle fixed cost and the idle driving cost from the depot to the departure point of the first train in the train chain; the corresponding entry arc represents the end of the bus returning to the depot, and the corresponding cost is the execution cost of the last train in the train chain and the idle driving cost returning to the depot; the service arc only needs to calculate the driving cost of the current executing trip; compared with the service arc, the idle driving arc also needs to add the idle driving cost from the current train's terminal to the next train's departure point; the waiting arc is relatively special, and the waiting cost of the waiting time θ(i, j) needs to be considered in the cost item.

[0095] S3. Use the collaborative dispatching network to connect and limit the coverage of each bus in multiple parking lots; establish an electric bus energy consumption model and a driver work model based on the responsibility chain.

[0096] It should be noted that the restrictions on the driver's coverage of the trains are taken into account on the basis of the train connection, and the train coverage and connection model of the electric bus cooperative driver is established, as shown below:

[0097]

[0098]

[0099] In the formula, denote decision variables, where Indicates whether the connection arcs (i, j) of the k-th layer network are connected, that is, whether train numbers i and j are executed consecutively by a certain vehicle; Indicates the departure time. Indicates the time of pickup. is the stop for train number i To the departure station of train number j of idle time, indicates whether the parking lot k has allocated vehicles from the departure station o to the network node j, cap k represents the number of vehicles that parking lot k can accommodate, T d represents the set of trips that driver d can perform, D(k) represents the set of drivers in parking lot k, A k It represents the effective connection formed between two nodes in the connection arc set, i and j represent the train number, and d represents the driver.

[0100] Formula (3) is the single execution constraint of the train, and each train node has a single vehicle to complete it; Formula (4) is the flow balance constraint, and each train node has its corresponding forward train and backward train to take over; the setting of loop arcs also makes the departure station and the arrival station meet the above two constraints, ensuring the integrity of the model; Formula (5) is the feasible constraint of train connection, which is a necessary condition for the continuous execution of two trains; Formula (6) sets the capacity upper limit of the yard, and the capacity of each yard is limited and cannot exceed the fixed number of vehicles; Formula (7) explains the decision variables is a 0-1 variable. When its value is 1, it means that the connection arc (i, j) is connected at the kth layer, and the train number i and the train number j are executed continuously by the vehicle; Formula (8) means that each train task will have and only one driver assigned to complete it; Formula (9) connects different tasks performed by the same driver to form a complete task chain; Formula (10) represents the decision variable is a 0-1 variable. When its value is 1, it means that the d-th driver belonging to the k-th layer continuously executes task (i, j). In order to associate the network flow problem with bus scheduling, formula (11) restricts vehicles to execute trips only when there is a driver. Formula (12) will allocate at least one driver to cover the existing vehicle scheduling plan to ensure that all trips are executed by a driver.

[0101] Among them, establishing the electric bus energy consumption model includes the following steps:

[0102] Collect the initial battery status of each electric bus before starting service, and dynamically adjust the energy consumption forecast and charging requirements of each electric bus on the route based on the initial battery status;

[0103] Obtain the battery energy status, set the power status limit and set the upper and lower limits of the battery energy status to ensure that the power status of each electric bus remains within the optimal range during operation;

[0104] The charging time of the electric bus under each battery charging mode is calculated according to the required supplementary power of the electric bus, and an electric bus energy consumption model is established to constrain the battery charging and discharging state. At the same time, the energy consumption of each electric bus is predicted through the electric bus energy consumption model.

[0105] It should be noted that the present invention uses battery life to characterize the energy state of charge (SOC) of the battery. This parameter can be understood as the length of time that the battery can discharge at a certain rated power e. It is assumed that all vehicles are in a fully charged state before leaving the depot to start the trip. After a period of driving, the battery energy will inevitably decrease. In order to protect the battery life as much as possible, avoid the negative effects of overcharging and discharging of the battery, and help the electric bus to run stably, the parameter τ is set to characterize the lower limit of SoC, and the parameter σ is used to represent the upper limit of SoC; that is, the battery life after the vehicle performs any trip needs to be limited to a certain range E. i ∈[τE rated , σE rated ], where E i It is used to refer to the remaining endurance time of the corresponding vehicle after completing trip i, and is used to express the constraint of battery capacity on vehicle operation, where E rated is the rated capacity of the battery.

[0106] Considering that operators will require buses to use the interval between buses as much as possible to recharge in order to reduce costs and minimize the number of buses, in addition to the conventional slow charging method, the present invention also adds fast charging and battery replacement technology to meet the needs of fast charging, thereby further coordinating the structural contradiction between vehicle charging and scheduling plans. u and They respectively refer to the charging power and charging time of the corresponding charging modes. Compared with fast charging, slow charging has a lower charging power and thus requires a relatively longer charging time. The time required for battery replacement is generally a shorter fixed time, and compared with the other two charging modes, it can be regarded as infinitely close to instant completion.

[0107] A charging time calculation method for multiple charging modes is designed, and the charging power r 3 The value is set to the rated capacity of the battery. 3 =E rated , so the corresponding "charging time" of battery replacement It can be approximated to one minute; analyzed from another perspective, the time used for the battery replacement operation is less than the minimum time interval, so it can definitely be completed within the train interval.

[0108] There is a mathematical model of energy consumption of electric buses under regional coordinated dispatch considering multiple charging modes, and its specific form is as follows:

[0109]

[0110] In the formula, represents the state of charge of the vehicle in parking lot k when it departs from the departure station o, σE ratedrepresents the upper limit of the vehicle's state of charge, e represents the rated power of the bus battery, represents the maximum waiting time between consecutive trains i and j (gap represents "waiting interval"), u represents the charging mode, and r u , They represent the charging power and charging time corresponding to charging mode u, r 3 Indicates the charging power.

[0111] Formula (13) ensures that the vehicle is fully charged before leaving the parking lot; Formula (14) is the power update formula, where the current train SoC is the previous train SoC minus the train energy consumption plus the supplementary energy of the train; Formula (15) limits the train SoC from dropping to a too low level and causing a large loss to the battery; Formula (16) limits the train SoC from being overcharged; the intermediate variables in Formula (17) are It represents the remaining time between any two nodes, which is defined as the maximum waiting time between trains in the present invention. Formula (18) is the calculation of the time required for the three charging methods. At the same time, the charging time will not exceed the maximum interval time between trains, ensuring the feasibility of the train chain. Formula (19) ensures that the charging opportunity will only appear at the connection with the train. The function sgn(·) in formula (19) is a symbolic function in mathematics, also known as a return function. The function can return a value of 1 for any non-zero input greater than zero, and a value of 0 for an input value equal to zero. Therefore, the function can indicate whether the bus at train number j is charged.

[0112] Among them, establishing a driver work model based on the responsibility chain includes the following steps:

[0113] Build a driver work model that maps the driver's work behavior; connect various dispatch nodes based on the preset responsibility chain to clarify the division of responsibilities between the driver and the electric bus;

[0114] Use the driver work model to dynamically control the driver's work behavior and manage and enforce the constraints on the driver's work behavior;

[0115] Among them, the driver's work behavior includes the driver's working hours, the driver's continuous driving time, and the driver's rest and dining behavior.

[0116] It should be noted that the driver coordination dispatch is included and a parallel "chain of responsibility" is established, as follows:

[0117] Based on relevant regulations, a time-expanded scheduling network for drivers is constructed, and a “responsibility chain” is defined to describe the collaborative scheduling process between drivers and vehicles. Finally, a labor law mathematical model for driver scheduling is established, as follows:

[0118] The first is the driver's daily cumulative working time limit model:

[0119]

[0120]

[0121] In the formula, Indicates that driver d starts from the departure station o d Initial working hours at departure, represents the single-day working hours (including rest time) of driver d when completing trip j, represents the single-day working hours (including rest time) of driver d when completing trip i, α max Indicates the upper limit of working hours in a single day. Indicates that the driver d returns to the pickup station s d The single-day working hours are: d represents the driver, o represents d Indicates the corresponding departure yard of driver d.

[0122] Formula (20) represents the daily working hours from driver d at the corresponding departure yard o d The calculation starts at the time of departure; Formula (23) indicates that the daily working hours of each driver d cannot exceed the upper limit; Formula (21) and Formula (22) use the Big M method to update the daily working hours. When driver d completes tasks i, j continuously, that is, arc (i, j) exists in the driver network d, Formula Formula (24) indicates that the accumulated working hours per day is a natural number.

[0123] The restable nodes for the driver are defined as follows:

[0124]

[0125] In the formula, Indicates whether driver d took a rest between tasks (i, j), which is a 0-1 decision variable; γmin represents the lower limit of the driver's single rest time.

[0126] Formulas (25) and (26) use 0-1 decision variables It determines whether the driver can take a rest. Formula (25) ensures that the rest process only occurs on arcs with sufficient time and arc (i, j) exists in this layer of the network.

[0127] A definition model for the driver's continuous driving time is established as follows:

[0128]

[0129]

[0130] In the formula, Indicates that driver d starts from the departure station o d Continuous driving time at departure, represents the continuous driving time of driver d when completing trip j (reset after each break), represents the continuous driving time of driver d when completing trip i (reset after each rest), β max Indicates the upper limit of the driver's continuous driving time.

[0131] Formulas (27)-(31) are about the continuous driving time of the driver. Formula (27) indicates that driver d starts from the corresponding departure yard o d The continuous driving time starts to be calculated at the time of departure; Formula (30) indicates that the continuous driving time does not exceed the upper limit of the continuous driving time at any node, and the value is restarted after the rest; When the driver does not take a rest, Formula (28) and Formula (29) ensure that the continuous working time is calculated according to the formula Update; Formula (31) indicates that the continuous driving time is a natural number.

[0132] Restrictions are imposed on drivers’ rest and dining behaviors. The specific mathematical model is as follows:

[0133]

[0134]

[0135] In the formula, Indicates that driver d starts from the departure station o d Total rest time per day at departure, represents the total rest time per day when driver d completes trip j, represents the total rest time per day when driver d completes trip i, γ max Indicates the upper limit of the driver's single rest time. Indicates that the driver returns to the pickup station d The total rest time per day is Indicates whether driver d has completed his meal between tasks (i, j), Indicates the lower limit of the dining time window, Indicates the upper limit of the meal time window.

[0136] Formulas (32) to (36) are about the total rest time of the driver. Formula (32) indicates that driver d starts from its corresponding departure yard o d The rest time at departure is 0; Formula (35) indicates that driver d arrives at the corresponding parking lot s d There will not be too much rest when there is a break; Formulas (33) and (34) indicate that only when there is a break, the total rest time is calculated according to the formula Update; Formula (36) indicates that the total rest time is a natural number.

[0137] Formulas (37)-(41) are aimed at the driver's meal problem. Here, the meal time is not included in the total rest time. Formula (37) ensures that the meal opportunity only appears in the task chain with backward trains; Formula (38) arranges the driver to have a meal before the next task starts when the conditions are met; Formulas (39) and (40) use the big M method to calculate the driver's meal time, and the formula is only valid when the arc exists; Formula (41) shows the decision variable It is a 0-1 variable.

[0138] Formulas (42) and (43) restrict rest opportunities. In any shift of driver d, rest opportunities are prohibited within 90 minutes after starting work and before ending work, making the task arrangement more scientific.

[0139] S4. Construct a Markov train chain to obtain the impact of road conditions on train travel time; construct a delay transfer model based on the delay accumulation effect and delay propagation effect on the train chain, and design a delay propagation chain selection algorithm.

[0140] It should be noted that in order to describe the adverse effects that actual road conditions may have on the punctuality of bus operation, the present invention regards the journey of vehicles between stations as a random process, and uses random variables to describe the travel time of buses from bus number i to bus number j, and then defines δ(i) to represent the accumulated delay time of the bus at bus number i, as follows:

[0141]

[0142] When δ(i) = 0, the vehicle is on time. When δ(i) = m, the vehicle is delayed for m minutes when it starts to run the train number i. The maximum delay time is set to δ M Minutes. Obviously, the delay time of each station is not fixed, and it is impossible to be equal. The delay time probability can be used to characterize the delay of a specific train under a scheduling plan. Define represents the probability that the cumulative delay time of train number i is m minutes, as shown in formula (44). According to the delay state space {0, 1, …, δ M}, available probability vector To represent the probability distribution of the delay time of train number i {p i}, thus quantifying the degree of influence of random trips on the station in the train chain. According to the actual situation of production and life, vehicles generally do not experience delays when they depart from the depot, especially

[0143] definition represents the probability that a bus will be delayed for n minutes when traveling from bus number i, which is delayed for m minutes, to bus number j, as shown in formula (45). The transfer probability from bus number i to bus number j is essentially a conditional probability. According to the total probability formula, the probability distribution of the delay time of bus number j is calculated as {p j}, as in formula (46), further simplification yields formula (47).

[0144]

[0145] In the formula, represents the probability that a bus will be delayed for n minutes when traveling from bus number i, which is delayed for m minutes, to bus number j. represents the probability that the cumulative delay time of train number i is m minutes, represents the probability distribution of the delay time of train number i, p j represents the probability distribution of the delay time of train number j, δ M represents the maximum delay time, δ(i) represents the accumulated delay time of the bus in bus number i, and δ(j) represents the accumulated delay time of the bus in bus number j.

[0146] By observing formula (46), it is not difficult to find that the delay probability of train number j is completely determined by the delay status of train number i in the previous stage, and has nothing to do with other train numbers. This characteristic is very similar to the memorylessness of Markov chain; therefore, the process of accumulating delay time at different stages on the same bus is defined as Markov process, and the corresponding train number chain can be called Markov train number chain. This definition can better describe the process of a train number chain being affected by random travel time and the continuous accumulation of delay time; at the same time, it can greatly facilitate the calculation of delay probability, and the probability distribution of any stage can be calculated by formula (47).

[0147] Among them, the Markov train chain is constructed to obtain the impact of road conditions on train travel time; the delay transfer model is constructed by combining the delay accumulation effect and delay propagation effect on the train chain, and the delay propagation chain selection algorithm is designed, which includes the following steps:

[0148] The travel time between trains is regarded as a random variable, and the time interval between trains is used to estimate the probability distribution of the random variable to construct a Markov train chain.

[0149] The Markov train chain is used to calculate the delay transmission probability from the current train to the next train, and the impact of road conditions on the train travel time is analyzed based on the delay transmission probability;

[0150] The delay expectation calculation technology is combined with the cumulative effect of delays on the train chain to obtain the probability of waiting time under different delay times, and a delay transfer model is constructed based on the delay propagation effect and occurrence probability.

[0151] The delay transfer model is used to determine the actual probability distribution of each train number, and the probability distribution of the train number is calculated by integrating it with the impact results of the train number's travel time. The maximum value of the probability distribution is selected as the basis for designing the delay propagation chain selection algorithm.

[0152] It should be noted that the bus schedule is not completely resistant to delays. A certain interval may be reserved between two trains. Under normal circumstances, the vehicle will wait in this short interval. In the face of small-scale delays, the train can also be adjusted flexibly. Therefore, affected by the randomness of the travel time, the waiting time will also change accordingly, and it also needs to be represented by a random variable. As shown in formula (17), the interval time is a certain value, and various situations will occur depending on the delay situation; for example, when the travel time is too long, it may occupy part of the waiting time, resulting in a reduction or even disappearance of the waiting time. In severe cases, it will cause delays in subsequent trains, resulting in a chain reaction, such as Figure 5 shown.

[0153] According to formula (17), the maximum waiting time for any set of tasks can be obtained The waiting time is defined as θ(i, j), which represents the waiting time of the bus between bus number i and bus number j. It is still treated in the same way as the delay time, and the waiting time is discretized into a set of integers with upper and lower bounds. The corresponding θ(i, j) = 0 means that there is no need to wait for a seamless connection between the two buses, and θ(i, j) = 1 means that the vehicle needs to wait for 1 minute after completing bus number i before starting bus number j.

[0154]

[0155] Similar to the delay time, the waiting time can be characterized by a set of distributions, and the probability matrix can be used to represent the possibility of waiting time under different delay times.

[0156] In the formula, represents the probability that a bus waits for 1 minute between bus number i and bus number j, as shown in formula (48); It represents the probability that a bus needs to wait for l minutes to start bus j after it finishes bus i which is delayed by m minutes, as shown in formula (49).

[0157] In order to obtain the probability distribution of delay time and waiting time for each bus, it is necessary to establish a connection between the travel time and the transfer probability matrix. It is known that the bus is delayed for m minutes at bus number i, and it is assumed that the waiting time after completing bus number i is 2 minutes, and the delay time transferred to bus number j is n minutes. The travel time required in the process is Δ ij ,like Figure 7 shown.

[0158] It can be deduced that the departure time of train number i can be expressed as The departure time of the next train j to be executed continuously is Then, the relevant equation can be established. The present invention previously mentioned that the delay time m, n is discretized. In order to calculate the relevant probability, it is necessary to artificially divide it into a range, such as equation (50).

[0159]

[0160]

[0161] In the formula, m represents the delay time of train number i, n represents the delay time of train number j, Δ ij represents the actual travel time of the bus from bus number i to bus number j, F i Represents the cumulative time distribution function of train number i.

[0162] Considering that there are many combinations of travel time and interval time, the probabilities of travel time and waiting time should have many situations, such as Figure 5 As shown, the derivation of each case will not be explained one by one, but can be summarized as piecewise functions (51) and (52), through the cumulative time distribution function F of train number i i , we can get the delay time transfer probability and waiting time occupancy probability

[0163] By analyzing the impact of delays between trains, we can deduce the propagation mechanism of delay impacts on the task chain, such as Figure 7 As shown. In the previous text, the present invention has used the Markov train chain to characterize the delay effect caused by random travel time from the perspective of the vehicle. Similarly, random travel time may also interfere with the bus driver performing driving operations, which is collectively referred to as the delay effect of the task chain. Unlike the vehicle delay that occurs in a fixed train chain, the driver's delay effect may be transferred. Specifically, after the original train is delayed, the driver's shift change action may transfer the delay to another train.

[0164] Figure 7 The example shows a situation where a shift point is delayed and transferred. Assume that there is an affected train chain consisting of a partial task chain a where driver d1 is located and a partial task chain b where driver d2 is located. Bus v2 was originally driven by driver d2 and was not affected by the delay. Since a shift change is required after the end of bus i, driver d1 will replace driver d2 to complete the subsequent trips of bus v2. However, task chain a is affected by the random travel time and is delayed before the shift change point, that is, bus j-1 is delayed m minutes before departure. Due to the existence of shift changes, Figure 7There is a deviation between the theoretical changeover point and the feasible changeover point. Bus v2 must wait for the delayed driver d1 to arrive, and fails to depart at the original time in the schedule. The impact of this delay will also be transmitted along the subsequent train chain of bus v2.

[0165] The calculation of the impact of delay needs to be considered from both the vehicle and the driver perspectives. The delay time distribution of a particular train corresponds to the maximum expected delay value obtained by the train chain and the task chain, as shown in formula (53). Figure 4 In , compared with the train chain and the task chain, the driver causes a greater delay expectation, so the probability distribution of train j should choose the probability distribution delivered by the task chain.

[0166]

[0167] Algorithm 1 Delayed propagation chain selection

[0168]

[0169] S5. Based on the delay propagation chain selection algorithm and the scheduling chain generation algorithm, the overall model of regional collaborative scheduling of drivers and electric buses is established and output to complete the scheduling process.

[0170] The establishment of the overall regional coordinated dispatch model of drivers and electric buses includes the following steps:

[0171] Define the cost parameters of the overall model of regional collaborative dispatch of drivers and electric buses; define the constraints of the overall model of regional collaborative dispatch of drivers and electric buses; and establish the overall model of regional collaborative dispatch of drivers and electric buses based on the cost parameters and constraints.

[0172] Among them, the cost parameters include bus vehicle costs, driver salary costs and penalty costs; the constraints include vehicle scheduling constraints, electric bus endurance constraints, driver scheduling constraints and delay and waiting time constraints.

[0173] It should be noted that the electricity cost during charging is not the focus of this problem, so it is not included in the cost item. However, for other charging schemes except slow charging, additional equipment is required to support them, so it needs to be considered in the cost item. Constraint (54) indicates that the total cost required for vehicle scheduling and charging is the lowest; is the vehicle dispatch cost, Indicates the additional cost of using a fast charging station, Represents the additional cost of using batteries in the battery swap solution.

[0174] Formula (55) represents the driver's salary cost, which is positively correlated with the number of drivers used in the dispatch scheme, where C s A fixed daily salary for the company's drivers.

[0175] Formula (56) can be used to obtain the delay penalty and waiting penalty with the help of the expectation formula, which are included in the objective function as the generalized cost CP, where C d is the unit delay time penalty factor, C w is the unit waiting time penalty factor.

[0176]

[0177] The overall regional dispatch model of electric buses with cooperative drivers considering the delay propagation effect in multiple charging modes is established as follows:

[0178] mmin CV+CR+CP (57)

[0179] st

[0180] (3)-(7)

[0181] (13)-(19)

[0182] (20)-(43) (51), (52), (53)

[0184]

[0185] The model objective function is shown in constraint (57), which mainly consists of three parts: bus cost, driver cost, and penalty cost. Equations (3)-(7) are vehicle scheduling equations; Equations (13)-(19) are electric bus endurance equations; Equations (20)-(43) are driver scheduling equations; Equations (58) and (59) indicate that there will be no delay when the bus departs from the depot, and Equations (51), (52), and (53) are the derivations of delay time and waiting time. Equations (60) and (61) indicate that the probability is a continuous decision variable. The chain generation algorithm for the regional scheduling problem of buses and drivers considering delay propagation is as follows:

[0186]

[0187]

[0188] In summary, with the help of the above technical solutions of the present invention, the method for coordinated dispatching of public transportation vehicles and drivers provided by the present invention can not only comprehensively consider the coordinated dispatching problems of vehicles and drivers, but also flexibly respond to the random travel time caused by changes in road conditions, thereby ensuring the efficiency and punctuality of public transportation operations. Therefore, public transportation operators can more flexibly arrange the operation plans of vehicles and drivers, minimize charging costs, improve operational efficiency, improve service quality, and meet the travel needs of passengers. The present invention can more accurately simulate and predict the dispatching of vehicles and drivers by constructing a time expansion network and a multi-layer time graph, and set corresponding costs for different types of connection arcs in the network, so that the dispatching decision can be optimized based on cost-effectiveness. At the same time, by considering multiple charging modes and driver responsibility chains, the dispatching method can flexibly respond to various emergencies, so that delays can be effectively predicted and managed, and their negative impact on the entire network can be reduced. Through accurate dispatching models and algorithms, the reliability and time accuracy of public transportation services can be guaranteed, thereby improving passenger satisfaction and overall service quality.

[0189] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for coordinated dispatching of public transport vehicles and drivers, characterized in that: The method for coordinated dispatching of public transport vehicles and drivers comprises the following steps: Establish a coordinated dispatching network between drivers and electric buses in multiple parking lots, and establish a time expansion diagram for the coordinated dispatching network corresponding to each parking lot; Establish and classify the connection arc sets in the time expansion graph, and set cost functions for different types of connection arcs in the collaborative scheduling network; Use the collaborative dispatching network to connect and limit the coverage of each bus under multiple parking lots; establish an electric bus energy consumption model and a driver work model based on the responsibility chain; Construct a Markov train chain to obtain the impact of road conditions on train travel time; construct a delay transfer model based on the delay accumulation effect and delay propagation effect on the train chain, and design a delay propagation chain selection algorithm; Based on the delay propagation chain selection algorithm and the scheduling chain generation algorithm, the overall model of regional collaborative scheduling between drivers and electric buses is established and output to complete the scheduling process.

2. A method for coordinated dispatching of public transport vehicles and drivers according to claim 1, characterized in that: The establishment of a coordinated dispatching network of drivers and electric buses in multiple parking lots includes: Based on the regional dispatch characteristics of drivers and electric buses, a corresponding collaborative dispatch network layer of drivers and electric buses is established for each depot, and its expression is: N k =o k ∪s k ∪T k In the formula, G k Represents the collaborative dispatching network layer between drivers and electric buses; 0 k represents the departure station in the current layer network k; N k represents the set of train numbers and parking lot nodes in the current layer network k; s k represents the departure station in the current layer network k; T k represents the set of tasks that all electric buses and drivers belonging to the current layer network k should complete; A k Represents a valid connection formed between two nodes in the connection arc set.

3. A method for coordinated dispatching of public transport vehicles and drivers according to claim 2, characterized in that: The establishing and classifying of connection arc sets in the time extension graph and setting cost functions for different types of connection arcs in the collaborative scheduling network include the following steps: Define the key parameters of each train in multiple yards to ensure that the time and location logic between trains meet the constraints; The connection arcs are classified according to the key parameters of each train and the execution activities of the connection arc set in the time expansion diagram, and a cost function is set for each type of connection arc.

4. A method for coordinated dispatching of public transport vehicles and drivers according to claim 3, characterized in that: The types of connection arcs include circulation arc, exit arc, entry arc, service arc, waiting arc and empty arc; The loop arc has no actual scheduling significance and is used to maintain the flow balance in the coordinated scheduling network. The starting point of the loop arc is the receiving yard and the end point is the departure yard. The exit arc indicates that the electric bus starts to perform service from the depot. The starting point of the exit arc is the departure depot where the electric bus is currently located, and the end point is any departure station. The entry arc indicates that the electric bus returns to the depot to end the service. The starting point of the entry arc is any bus collection station, and the end point is the bus collection depot where the current electric bus is located. The service arc means that the electric bus continuously executes two trips on the timetable without waiting or empty running in between; The waiting arc indicates that the electric bus is waiting for the next bus trip. The starting point of the waiting arc is the stop of the previous bus trip, and the end point is the departure point of the next bus trip. The empty running arc represents the empty running activity of the electric bus between different stations. The starting point of the empty running arc is the arrival station of the previous bus, and the end point is the departure station of the next bus.

5. A method for coordinated dispatching of public transport vehicles and drivers according to claim 4, characterized in that: The establishment of the electric bus energy consumption model comprises the following steps: Collect the initial battery status of each electric bus before starting service, and dynamically adjust the energy consumption forecast and charging requirements of each electric bus on the route based on the initial battery status; Obtain the battery energy status, set the power status limit and set the upper and lower limits of the battery energy status to ensure that the power status of each electric bus remains within the optimal range during operation; The charging time of the electric bus under each battery charging mode is calculated according to the required supplementary power of the electric bus, and an electric bus energy consumption model is established to constrain the battery charging and discharging state. At the same time, the energy consumption of each electric bus is predicted through the electric bus energy consumption model.

6. A method for coordinated dispatching of public transport vehicles and drivers according to claim 5, characterized in that: The establishment of a driver work model based on the chain of responsibility includes the following steps: Build a driver work model that maps the driver's work behavior; connect various dispatch nodes based on the preset responsibility chain to clarify the division of responsibilities between the driver and the electric bus; Use the driver work model to dynamically control the driver's work behavior and manage and enforce the constraints on the driver's work behavior; Among them, the driver's work behavior includes the driver's working hours, the driver's continuous driving time, and the driver's rest and dining behavior.

7. A method for coordinated dispatching of public transport vehicles and drivers according to claim 6, characterized in that: The method of constructing a Markov train chain to obtain the influence of road conditions on the train travel time; constructing a delay transfer model based on the delay accumulation effect and delay propagation effect on the train chain, and designing a delay propagation chain selection algorithm includes the following steps: The travel time between trains is regarded as a random variable, and the time interval between trains is used to estimate the probability distribution of the random variable to construct a Markov train chain. The Markov train chain is used to calculate the delay transmission probability from the current train to the next train, and the impact of road conditions on the train travel time is analyzed based on the delay transmission probability; The delay expectation calculation technology is combined with the cumulative effect of delays on the train chain to obtain the probability of waiting time under different delay times, and a delay transfer model is constructed based on the delay propagation effect and occurrence probability. The delay transfer model is used to determine the actual probability distribution of each train number, and the probability distribution of the train number is calculated by integrating it with the impact results of the train number's travel time. The maximum value of the probability distribution is selected as the basis for designing the delay propagation chain selection algorithm.

8. A method for coordinated dispatching of public transport vehicles and drivers according to claim 7, characterized in that: The establishment of the overall regional coordinated dispatching model of drivers and electric buses includes the following steps: Define the cost parameters of the overall model of regional cooperative dispatch of drivers and electric buses; Define the constraints of the overall model of regional coordinated dispatch of drivers and electric buses; Based on cost parameters and constraints, a general model of regional collaborative dispatching of drivers and electric buses is established.

9. A method for coordinated dispatching of public transport vehicles and drivers according to claim 8, characterized in that: The cost parameters include bus costs, driver salary costs and penalty costs; The constraints include vehicle dispatch constraints, electric bus range constraints, driver dispatch constraints, and delay and waiting time constraints.

10. A method for coordinated dispatching of public transport vehicles and drivers according to claim 9, characterized in that: The cost function of each type of connection arc is expressed as: In the formula, represents the vehicle dispatching cost of the connecting arc (i, j) in parking lot k; C f is the fixed cost per vehicle; C m Represents the unit vehicle driving cost; C w represents the unit vehicle waiting cost; Indicates the departure time. Indicates the time of car collection; is the stop for train number i To the departure station of train number j of idle time; θ(i, j) represents the waiting time of the connection arc (i, j).