A dynamic marshalling method applied to virtual marshalling and a train control system

Abstract

The application provides a dynamic marshalling method applied to virtual marshalling, and the method comprises the following steps: in response to a command of dynamic marshalling, obtaining a marshalling section and enabling adjacent trains to establish train-to-train communication to obtain front train and rear train information in the adjacent trains; based on the obtained front train and rear train information, a first optimization target meeting a virtual marshalling constraint condition is constructed with the target of accurate marshalling and low energy consumption; a dynamic programming method is adopted to calculate an optimal control strategy in a virtual marshalling process and an offline speed curve corresponding to the optimal control strategy based on the first optimization target; the rear train is controlled to run according to the offline speed curve based on the optimal control strategy; the running condition of the rear train is monitored in real time during the running of the rear train; when the actual running speed curve of the rear train deviates from the offline speed curve, a sequential quadratic programming method is adopted to calculate and update the optimal control strategy and the speed curve in a subsequent stage, and the rear train is controlled based on the updated optimal control strategy.

Description

Technical Field

[0001] This invention relates to the field of rail transit, specifically to the field of train interval control in rail transit systems, and more specifically to a virtual formation method for train control systems. Background Technology

[0002] In the field of urban rail transit control, train operation is controlled through a train operation control system. Train interval control is one of the fundamental functions of this system and a necessary measure to ensure the orderly and safe operation of trains. Among existing technologies, the spatial interval method is a typical and widely used interval control method. Under this method, a strict distance is maintained between the following and preceding trains, effectively preventing head-on collisions and rear-end collisions. Furthermore, while ensuring operational safety, shortening the train interval can effectively improve the transport capacity of rail transit. Therefore, to improve operational efficiency, many subway lines have completed the conversion from fixed block system to moving block system, and many mainline railways have also completed the conversion from fixed block system to quasi-moving block system.

[0003] However, whether it is a fixed block, moving block, or quasi-moving block, the existence of absolute braking distance tracking interval restrictions (i.e., the train interval should not be less than the braking distance of the following train) has led to a bottleneck in the way to shorten the train interval and improve transportation efficiency, and there is an urgent need for a more effective way to improve transportation efficiency. Summary of the Invention

[0004] To address one of the aforementioned technical deficiencies, this application provides a novel dynamic train formation method and train control system applicable to virtual train formation.

[0005] According to a first aspect of the present invention, a dynamic train formation method for virtual train formation is provided. The method includes: responding to a dynamic train formation command, acquiring a train formation section and establishing train-to-train communication between adjacent trains to obtain information about the preceding and following trains in the adjacent trains; based on the acquired preceding and following train information, constructing a first optimization objective that satisfies virtual train formation constraints with the goals of accurate train formation and low energy consumption; using a dynamic programming method to calculate the optimal control strategy and the corresponding offline speed curve during the virtual train formation process based on the first optimization objective; the following train controls itself to run according to the offline speed curve based on the optimal control strategy, and monitors its operation in real time during its operation; when the actual operating speed curve of the following train deviates from the offline speed curve, using a sequential quadratic programming method to calculate and update the optimal control strategy and speed curve for subsequent stages; the following train controls itself to run according to the updated speed curve based on the updated optimal control strategy until the dynamic train formation is completed. By combining dynamic programming and sequential quadratic programming, the transition from a moving block state to a virtual train formation state is achieved while ensuring computational efficiency.

[0006] Preferably, when using the sequential quadratic programming method to calculate and update the optimal control strategy and speed curve for subsequent stages, the following steps are performed: The current positions of the preceding and following trains are acquired in real time, and the train's current formation decision stage is determined based on this; a second optimization objective satisfying the virtual formation constraints is constructed with low energy consumption as the goal, and the sequential quadratic programming method is used to calculate the optimal control strategy and the corresponding speed curve for subsequent formation decision stages based on the second optimization objective. By applying the sequential quadratic programming method to the offline speed curve of dynamic programming, computational efficiency can be ensured while effectively addressing train rolling control under external interference conditions.

[0007] Preferably, the method further includes: after the first detection that the actual operating speed curve of the following vehicle deviates from the offline speed curve, in each subsequent grouping decision stage, using a sequential quadratic programming method to continuously calculate the optimal control strategy and the speed curve corresponding to the optimal control strategy for the following vehicle in each subsequent grouping decision stage.

[0008] Preferably, the method further includes: monitoring the operation of the preceding vehicle in real time based on its speed curve; when the preceding vehicle brakes suddenly, the preceding vehicle sends a warning message to the following vehicle via vehicle-to-vehicle communication to induce the following vehicle to take emergency braking; updating the dynamic formation request when the preceding vehicle deviates from its speed curve; and ending the dynamic formation when formation conditions are no longer met. This effectively addresses the emergency braking situation of the preceding vehicle during dynamic formation, preventing collisions.

[0009] Preferably, the method further includes: when the train-to-train communication between the preceding and following trains is lost, the following train operates according to the offline speed curve; after re-establishing train-to-train communication with the preceding train, the method monitors whether the actual operating speed curve of the following train deviates from the offline speed curve; if the actual operating speed curve of the following train deviates from the offline speed curve, a sequential quadratic programming method is used to calculate and update the optimal control strategy and speed curve for the subsequent stages; the following train is controlled to operate according to the updated speed curve based on the updated optimal control strategy until dynamic formation is completed. This ensures that dynamic formation can be effectively achieved even when train communication is unavailable.

[0010] According to a second aspect of the invention, a train control system is provided, the system being configured to perform dynamic formation as described in the first aspect of the invention, and to closely track the train based on the speed at the end of the formation after the dynamic formation is completed.

[0011] Preferably, the system includes: a ground control module for issuing control commands to the train to control its operation, and issuing dynamic formation commands to the train when train interval control is required; multiple formation control modules, each dynamic formation module being installed on a train, wherein, upon receiving a dynamic formation command from the ground control module, the dynamic formation module on the following train performs dynamic formation according to the method described in the first aspect of the present invention to obtain the optimal control strategy and the speed curve corresponding to the optimal control strategy during the dynamic formation process; and multiple on-board control modules, each on-board control module being installed on a train, for controlling the following train to run according to the speed curve based on the optimal control strategy obtained by the dynamic formation module on its train until the virtual formation is completed.

[0012] Preferably, each train formation control module includes: an information acquisition module, used to acquire information about its own train and the train ahead of it when receiving a dynamic formation command. The information about its own train includes the speed and position of the train behind it, and the information about the train ahead of it includes the speed curve, position, and speed of the train ahead of it; a dynamic planning module, used to construct a first optimization objective that satisfies the virtual formation constraints based on the information about its own train and the train ahead of it acquired by the information acquisition module, with the goal of accurate formation and low energy consumption, and to calculate the optimal control strategy and the corresponding offline speed curve during the dynamic formation process using a dynamic planning method based on the optimization objective, and to load the offline speed curve into the corresponding on-board control module; a monitoring module, used to monitor the operation of its own train in real time, and to send a rolling optimization command to the sequence planning module when the actual speed curve of the train deviates from the offline speed curve; and a sequence planning module, used to calculate and update the optimal control strategy and speed curve of its own train in the subsequent stage of virtual formation using a sequence quadratic programming method based on the rolling optimization command, and to load the updated speed curve into the corresponding on-board control module.

[0013] Compared with existing technologies, the advantages of this invention are as follows: This invention achieves the transition of trains from moving block mode to virtual formation mode through a dynamic formation control scheme combining dynamic programming and sequential quadratic programming. By establishing a target model for dynamic formation control of the train, and based on the dynamic programming algorithm, a formation curve calculation method is proposed, and the speed curve is calculated for the train to be formed through offline optimization. The dynamic programming algorithm provides the train with an ideal speed curve, guiding the formation process under interference-free conditions, achieving the goals of high formation accuracy and low energy consumption. Based on the offline speed curve, combined with the sequential quadratic programming algorithm, an online sequential quadratic programming rolling optimization control method is proposed. This method can effectively provide online control for the train under external interference conditions while ensuring computational efficiency, correcting train speed deviations caused by external interference. The solution of this invention not only improves formation efficiency but also greatly improves train operation efficiency. Attached Figure Description

[0014] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0015] Figure 1 This is a schematic diagram illustrating the change in the distance between vehicles during dynamic formation according to an embodiment of the present invention; wherein, Figure 1 (a) shows a diagram illustrating the distance between the leading and trailing vehicles at the start of the train formation. Figure 1 (b) shows a diagram illustrating the changes in distance and speed between the following vehicle and the vehicle in front during a car chase. Figure 1 (c) shows the distance between the rear and front vehicles and the speed at the end of the formation;

[0016] Figure 2 This is a schematic diagram of velocity discretization according to an embodiment of the present invention;

[0017] Figure 3 This is a schematic diagram of sub-interval indices according to an embodiment of the present invention;

[0018] Figure 4 This is a schematic diagram of a rolling optimization control train using a sequential quadratic programming method according to an embodiment of the present invention;

[0019] Figure 5 This is a schematic diagram before and after dynamic grouping with operational deviations according to an embodiment of the present invention. Detailed Implementation

[0020] To make the technical solutions and advantages of the embodiments of this application clearer, the exemplary embodiments of this application will be described in further detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not an exhaustive list of all embodiments. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other.

[0021] As described in the background section, under existing train interval control methods, the approach of improving transportation efficiency by shortening train intervals has reached a bottleneck. To break free from the constraints of existing block systems, researchers have proposed a virtual train formation scheme. In this scheme, each train maintains a safe distance from the preceding train, a distance far lower than the braking distance required for a complete stop. The driver applies the brakes based on the brake lights of the preceding train. Under virtual formation, trains use wireless communication to establish connections with adjacent trains within the section, exchanging information such as speed and position. Combined with reference signals from ground equipment and train-to-train communication data, the trains in the formation operate in close tracking, significantly reducing train intervals. This interval is not constrained by absolute braking distances, further improving transportation efficiency.

[0022] The implementation of a virtual train formation scheme involves two essential steps: dynamic formation and close tracking. Dynamic formation refers to the process by which trains transition from a moving block system to a close formation during operation. Close tracking is the process by which trains closely follow each other at the speeds maintained after the dynamic formation is completed.

[0023] It should be noted that dynamic train formation control is a multi-objective control process. On the one hand, it requires that at the end of the dynamic formation, the speeds of the trains before and after are the same and the spacing is the ideal formation interval. On the other hand, it is necessary to effectively cope with external interference and reduce the traction energy consumption of the train.

[0024] To achieve efficient dynamic formation in virtual train operation, this invention proposes a dynamic formation control method that combines offline speed curve planning and online optimization control. First, an optimization problem is constructed with the objectives of accurate formation and low energy consumption. This problem is solved using a dynamic programming algorithm to obtain the offline optimal speed curve. Then, under actual operating conditions, an optimization problem is continuously established based on the real-time intervals. Using the offline speed results as the initial solution, a sequential quadratic programming algorithm is used to solve for the optimal control strategy online, thereby ensuring high computational efficiency.

[0025] To better understand this invention, it is described in detail below with reference to the accompanying drawings and embodiments. Specifically, for ease of understanding, this invention first introduces the target model, then describes the process of obtaining offline velocity curves based on dynamic programming, next describes the process of solving the optimal control strategy online based on sequential quadratic programming, and finally describes the control principles under abnormal dynamic grouping conditions.

[0026] 1. Target Model

[0027] The optimization objectives of this invention are accurate grouping and low energy consumption. The accurate grouping model and the energy consumption model are described below.

[0028] 1.1 Precise Grouping Model

[0029] Precise formation is the primary goal of dynamic formation control. Precise formation means that after dynamic formation is completed, the distance between the train and the preceding train is the ideal formation interval, and the two trains have the same speed.

[0030] Under virtual train formation control, the following train (represented by f in this embodiment) and the preceding train (represented by p in this embodiment) share the same running section ahead, and the two trains being close in distance and having the same speed are the conditions for starting dynamic train formation. With the cooperation of ground equipment, the preceding and following trains complete car-to-car communication connection and begin the dynamic train formation process.

[0031] According to one example of the invention, such as Figure 1 As shown in (a), at the start of the train formation, the train is in a moving block state, and the following train f maintains an absolute braking distance interval with the preceding train p, where d inter v represents the initial distance between the two vehicles. p and v f The speeds of the front and rear vehicles are s, respectively. start This represents the position of the car ahead at the start time.

[0032] During the car chase, such as Figure 1 As shown in (b), the following vehicle f needs to increase its speed to reduce the distance between the two vehicles, and then decelerate after a period of time to coordinate with the speed of the preceding vehicle p.

[0033] End of formation time, such as Figure 1 As shown in (c), the speeds of the rear vehicle f and the front vehicle p are the same, and the distance between the two vehicles is d′. inter The interval is much smaller than the absolute braking distance, where s end This indicates the position of the car preceding the vehicle at the end of the time.

[0034] During the entire formation process, the distance traveled by the following car is l. f The distance traveled by the preceding vehicle is l p =s end -s startTherefore, the distance between the two cars at the end time is d′. inter =d inter -(l f -l p The ideal grouping interval is defined as d. vcm At the end of the train formation, the preceding and following trains need to maintain an ideal formation distance and the same speed to achieve precise formation. According to one embodiment of the invention, formation accuracy is measured using formation error, which includes speed error and distance error. Speed ​​error is the speed difference between the preceding and following trains at the end of the formation, denoted by Δv. error The distance error is the difference between the actual train interval and the ideal train interval at the end time, as shown in the following formula:

[0035] l error =d′ inter -d vcm (Equation 1)

[0036] Among them l error For grouping distance error, d′ inter This refers to the actual train interval at the end of the journey. The closer the train formation error is to zero, the higher the formation accuracy. d′ inter The closer the train interval is to the ideal, the better. During the train formation process, the goal should be to minimize the formation error. To facilitate the description of the optimization problem, this invention introduces the concept of coordination distance, which is the ideal reduction in the interval between preceding and following trains during dynamic formation, as shown in the following formula:

[0037] l cd =d inter -d vcm (Equation 2)

[0038] The goal of establishing precise grouping is:

[0039] Δv error →0 (Equation 3)

[0040] |l f -l p |→l cd (Equation 4)

[0041] At the end of the train formation, the closer the difference in running distance between the preceding and following trains is to the coordinated distance, the closer the goal of precise train formation is to being achieved. Considering the various numerical measurement errors that may exist during actual train operation, a maximum permissible error Δl is set for the train formation target. error_max Therefore, equation (4) can be replaced with:

[0042] l cd -Δl error_max ≤|l f -l p |≤lcd +Δl error_max (Equation 5)

[0043] 1.2 Energy Consumption Model

[0044] Train energy consumption includes many aspects, such as power supply to carriage equipment and train traction. For ease of understanding, this embodiment of the invention does not analyze the process of electrical energy transfer, but only considers the mechanical energy consumed by the train, i.e., only the energy consumption caused by train traction. The train traction system is modeled after a subway system. The power supply comes from the power grid, and after power is supplied, it undergoes DC-AC conversion. The corresponding AC traction motor converts electrical energy into mechanical energy to traction the train.

[0045] Based on the law of conservation of energy, establish the energy consumption equation:

[0046]

[0047] P(t)=u(t)v(t) (Equation 7)

[0048] Where e is the total traction energy consumption of the train, which is expressed as the power integral over time; T is the total formation time; P is the instantaneous output power of the train at time t; u(t) is the traction force of the train at time t; and v is the speed of the train at time t. In this embodiment of the invention, energy loss in the train transmission system is ignored, and it is assumed that the train cannot recover braking energy, so the traction energy consumption is zero when the train actively brakes.

[0049] The goal of low energy consumption is to minimize total traction energy consumption during dynamic formation. Based on the above description, the low energy consumption target model is established as follows:

[0050]

[0051] 2. Dynamic Programming

[0052] 2.1 Introduction to Dynamic Programming

[0053] Dynamic programming is an effective method for solving multi-stage decision optimization problems, primarily used for optimizing dynamic processes divided into stages by time or space. Although dynamic programming is computationally expensive and cannot be directly used for online computation, it has advantages such as accurate model description, high-quality optimization results, and ease of handling complex objective functions.

[0054] To facilitate understanding, we will first provide an overview of the commonly used terminology and basic principles of dynamic programming.

[0055] Stages: Stages are the natural divisions of the entire process, and can be divided according to time or spatial order. In this embodiment of the invention, k represents the stage sequence number variable, and the number of stages is n.

[0056] State: The objective situation at the starting point of each stage is the state.

[0057] Decision: When the state of a stage is determined, different choices can be made to transition the target system to the next state, and the choice made at this stage is the decision.

[0058] Policy: A sequence composed of decisions is called a policy. The process from the k-th stage to the last stage is the subsequent sub-process, and the policy adopted by the subsequent sub-process is the subsequent sub-policy, denoted as p k,n .

[0059] Optimal policy: A policy that achieves the optimal effect within the allowable range is called the optimal policy.

[0060] State transition equation: At the state s in the k-th stage k , making a decision u k , the state at the k + 1-th stage is s k+1 The relationship between the two states before and after satisfies s k+1 = T(s k , u k ), and this relationship is called the state transition equation.

[0061] Index function: The index function refers to the index that measures the quality of a policy in the dynamic programming method. It is a quantitative function defined on the whole process and all subsequent sub-processes. Using V k,n (s k , p k,n ) to represent the value of the index function of the subsequent sub-process when using the policy p k at the state s in the k-th stage k,n . When k = 1, V k,n (s k , p k,n ) represents the value of the index function of the whole process.

[0062] Based on the above concepts, the optimality theorem of dynamic programming is introduced: For the initial state s1, the policy is the optimal policy if and only if, for any k (1 < k < n), the index function satisfies the following relationship:

[0063]

[0064] where p 1,n = (p 1,k-1 , p k,n ), indicating that the decision of the whole process can be divided into a combination of two decisions with the k-th stage as the demarcation point. is the state at the k-th stage determined under the initial state s1 and the sub-policy p 1,k-1 .

[0065] prove:

[0066] Proof of necessity: Let This is the optimal strategy, which, based on the separability of the index function, means that for stage k:

[0067]

[0068]

[0069] in Corresponding to the preceding subsequence p 1,k-1 ,by There exists an allowed set of sub-policies for the starting point, denoted as . In set P 1,n Finding the optimal solution on (s1) is equivalent to first finding the optimal solution with a certain sub-policy p. 1,k-1 ∈P 1,k-1 (s1) corresponds to the sub-strategy combination set Find the optimal solution, then find these suboptimal solutions, and finally combine them with P. 1,k-1 The optimal solution on (s1). Therefore, the above formula can be written as:

[0070]

[0071] Sufficiency proof: Let strategy Make the theorem true, and prove it. This is the optimal strategy.

[0072] Let p 1,n =(P 1,k-1 ,P k,n )∈P 1,n If (s1) is any allowed policy in the minimization problem, then:

[0073]

[0074] Based on the above, the right-hand side of the formula represents the index value under the optimal strategy. Thus, the theorem is proved. Based on the above theorem and the state transition equation, the optimal strategy can be obtained by using the forward reasoning method.

[0075] 2.2 Offline speed curve planning based on dynamic programming

[0076] As mentioned earlier, dynamic programming has significant advantages in solving multi-objective optimization problems, but it is generally applicable to offline computation. Dynamic train formation in virtual train formation is precisely a low-energy, high-precision multi-objective optimization problem. Therefore, dynamic programming algorithms can be used to plan the speed curve offline for the dynamic train formation process, and then the offline speed curve can be stored on the train for direct access. This allows for the rapid acquisition of the optimal offline speed curve for the multi-objective optimization problem.

[0077] According to an embodiment of the present invention, the virtual train formation offline speed curve planning process based on dynamic programming can be summarized as follows: First, the dynamic formation process is divided into stages in the time domain, and the problem of solving the dynamic formation speed curve is transformed into an optimization strategy problem; then, in order to reduce the state space, the train running speed is discretized, and the state transition equation for connecting the speed values ​​of each sub-section is calculated in combination with the dynamic model; next, the objective function and constraints are established according to the control objective model; finally, the optimality theorem of dynamic programming is used to solve the problem, obtain the optimal control strategy, and then generate the optimal speed curve.

[0078] The following detailed explanation, using specific examples, details the process of obtaining offline speed curves using dynamic programming:

[0079] (1) Grouping time allocation

[0080] In dynamic programming, the process to be optimized is generally divided into stages according to time or spatial order. According to one embodiment of the present invention, the dynamic grouping optimization process is divided using a time sequence. Although the goal of offline optimization is to obtain the optimal speed-distance curve, in the time dimension, the distance and speed relationship between the preceding and following vehicles at the same moment can be easily calculated. Constraining the distance and speed difference can ensure operational safety; therefore, the present invention chooses the principle of time division. Grouping time is the total duration of the grouping process. Assuming the speed-distance curve of the preceding vehicle in the grouping area is known, and the preceding vehicle can travel according to this curve, the grouping time can be obtained from this speed curve. According to one embodiment of the present invention, when using dynamic programming for dynamic grouping, the grouping area is first divided into n equal-length sub-intervals. The speed of the preceding vehicle at the beginning and end of each sub-interval is obtained. The running time Δt of the preceding vehicle in each sub-interval is calculated using the speed-time integral relationship and combined with historical operating data. k (k=1,2,…,n), according to the time length of each sub-interval Δt k The rules divide the formation time; the start time of each sub-section is defined as the driving force / braking force decision stage, and the operation sequence composed of the control forces in each decision stage is the control strategy of dynamic train formation (based on the previous sub-section division, each sub-section corresponds to a decision stage (also called the formation stage)); the train speed in the decision stage is defined as the "state" in the dynamic programming method.

[0081] After dividing the train formation process into time domains, the exact location of the train cannot be determined during the decision-making stage. This may make it difficult for the train to obtain information about the track, such as the gradient, radius of curvature, and speed limit. However, given that the train formation process is relatively short, generally lasting within tens of seconds, it is assumed that the gradient, curvature, and speed limit of the entire formation section are constant and known.

[0082] (2) Velocity discretization

[0083] To improve solution efficiency and reduce the state search space, such as Figure 2 As shown in the figure, the speed at each decision-making stage is discretized in this embodiment of the invention. The speed at each decision-making stage is subject to minimum and maximum limits, where the minimum limit speed is the speed of the preceding train, and the maximum limit speed is the maximum limit speed of the line. Using the speed of the preceding train as the minimum speed limit ensures that the following train is always faster than the preceding train, thus ensuring that the train interval is minimized at the end of the train formation and avoiding rear-end collisions due to excessively small train intervals during the formation process. (Figure t) start and t end These represent the start and end times of the grouping process, respectively, and the grouping time is the difference between the two. Δt k v represents the time length of each sub-interval. p v is the speed of the vehicle in front. limit For line speed limit, v k This represents the speed of the following train in the k-th sub-interval. The dots in the diagram represent discretized speed values; hollow dots represent speeds outside the permissible range, while solid dots represent permissible speeds. The discretized permissible speed represents the train's permissible state during the decision-making phase. Δv represents the discretization precision of the speed. The initial speed of the train in phase k is expressed as: v k =iΔv, i∈N. v k+1 This is the final speed of the train in stage k, and also the starting speed in stage k+1.

[0084] (3) Establishment of state transition relationships at different speeds

[0085] In order to accurately calculate the control force and running time of a single sub-section, the dynamic differential equation of the train is first established.

[0086] Depend on

[0087]

[0088] We can obtain:

[0089]

[0090]

[0091] The train state transition equation expresses the relationship between the starting speed, ending speed, driving force / braking force, and travel distance within a single sub-section. Assuming the driving force / braking force of the train remains constant within a single sub-section, to more accurately describe the motion state, when the ending speed and starting speed are not the same, the speed change within the sub-section is divided into l equal parts, resulting in: h = (v... k+1 -v k According to Equation 13, combined with the composite trapezoidal rule, the state transition satisfies the following relationship:

[0092]

[0093] Equation 15 represents the change in driving force / braking force from v under different driving force / braking force u. k Transfer to v k+1 The time consumed is Δt′ k F represents the net force acting on the train.

[0094] The initial and final velocities of the sub-interval determine the magnitude of the control force to be applied during the decision-making phase. This is based on the actual sub-interval time length Δt. k He Shi 15 obtained v k to v k+1 Control force u under state transition k :

[0095] u(v k ,v k+1 )={u|min{|Δt′ k (u,v k ,v k+1 )-Δt k |},u∈} (Equation 16)

[0096] Then based on the control force u k Using Equation 14 and the composite trapezoidal rule, the train's running distance Δs in the sub-section is obtained. k :

[0097]

[0098] Thus, the present invention links the initial velocity, final velocity, and running distance of a single sub-interval through equations 15, 16, and 17.

[0099] (4) Establishment of energy consumption-distance index function

[0100] Based on the target model of dynamic formation, index functions describing energy consumption and operating distance are established. Energy consumption corresponds to the low-energy consumption target, and operating distance corresponds to the precise formation target. The specific index functions are as follows:

[0101]

[0102]

[0103] g k This represents the stage indicator for the k-th sub-interval. This refers to the relationship coefficient. In this embodiment of the invention, under the index function designed in Equation 18, the relationship between total energy consumption and total operating distance can be adjusted by changing the magnitude of the relationship coefficient. According to an example of the invention, such as... Figure 3 As shown, calculating the index values ​​for all sub-intervals using the above formula yields a schematic diagram of the index values ​​for each sub-interval. Circles represent train speeds at each decision stage, and lines connecting circles represent train transitions at different speeds. Solid lines indicate that speed transitions can occur under train traction / braking force constraints, while dashed lines indicate that speed transitions are not possible. The values ​​on the solid lines represent the index value g under different speed transitions at each decision stage. k .

[0104] (5) Establishment of dynamic grouping optimization problem

[0105] Based on the dynamic grouping optimization objective, an optimization problem is established.

[0106]

[0107] Among them, the constraints are:

[0108] u k = u(v k ,v k+1 (Equation 21)

[0109]

[0110] u k ∈U (Equation 23)

[0111]

[0112]

[0113] Where n represents the number of subintervals to be divided into during the dynamic programming calculation, k represents the subinterval number, and g k v represents the stage indicator for the k-th sub-interval. k Let u represent the speed of the following vehicle in the k-th sub-interval, and Equation 20 represents the optimization objective as obtaining the global minimum of the energy consumption-distance index; k Equation 21 represents the transfer control force of the rear vehicle from the speed of the kth sub-interval to the speed of the (k+1)th sub-interval. v represents the speed of the vehicle ahead in the k-th sub-interval. limutEquation 22 represents the speed limit of the line; U represents the allowable range of control force; Equation 23 represents the control force constraint; v n+1 This indicates the speed of the following train when the dynamic formation is completed. Equation 24 represents the speed of the preceding train when the dynamic formation is completed. Equation 24 is the speed terminal constraint, which means that the speeds of the preceding and following trains should remain the same at the end of the formation.

[0114] After solving the above optimization problem using dynamic programming, the energy consumption and total travel distance of the following train in the formation process can be obtained. According to the precise formation model described above, the travel distance of the following train must satisfy the following relationship.

[0115]

[0116] Equation 25 can verify whether the optimization result satisfies the precise formation objective. At the end of the formation, if the train state satisfies the constraints of Equation 25, it is called precise formation. In optimization algorithms, the objective function is generally a single expression. For problems with two objectives, this paper adopts a method of iteratively verifying by modifying the relation coefficients.

[0117] (6) Solving by dynamic programming

[0118] The optimization problem established in the above embodiment is solved based on the optimality theorem of dynamic programming. The core of the optimality theorem solution method is to transform the original problem into a series of single problems. This invention uses the forward dynamic programming solution method, and referring to Equation 9, the recurrence equation is established as follows:

[0119] f k (v k+1 )=min{f k-1 (v k )+g k (v,v k+1 )},v∈V (Equation 26)

[0120] Where f k (v k+1 ) represents the velocity v at the end of the k-th subinterval. k+1 At that time, the optimal value from the first subinterval (decision stage, the subinterval corresponds to a grouping stage, that is, a decision stage in dynamic programming. For ease of description, the decision stage, subinterval, and grouping stage will be used interchangeably hereafter, in order to better understand the present invention) to the kth subinterval (decision stage). k-1 (v k ) represents the velocity v at the end of the (k-1)th subinterval. k When, the optimal value from the first decision stage to the (k-1)th decision stage. k (v,v k+1 ) represents the velocity at the end of the k-th subinterval, which is v. k+1The index value at time, where v represents the starting velocity of the k-th sub-interval, and V represents the constraint range of the velocity.

[0121] The calculation begins by solving for the optimal index f under different endpoint velocities in each interval, starting from the first stage and proceeding sequentially. k (v k+1 ), until the final state v k+1 The corresponding indicator f n Then, starting from the endpoint state, a reverse search is performed using the backtracking table to obtain the control strategy and the speed values ​​for each stage. The total running distance is then calculated using the speed values ​​for each stage to verify whether the precise grouping objective described in Equation 25 is satisfied. If not, the relationship coefficients are modified. The process of recalculating the optimal result using the dynamic programming algorithm continues until the objective is met. The optimal speed curve obtained through offline dynamic programming is represented as follows: This speed is the ideal speed for dynamic formation. Based on this curve, the ideal control strategy can be calculated, that is, the optimal control force of the following vehicle at each decision stage.

[0122] If the train runs at the speed mentioned above, it can successfully complete the dynamic formation task.

[0123] 2.3 Sequence Quadratic Programming Rolling Optimization

[0124] If the following train runs according to the offline speed curve obtained by the above dynamic programming, the dynamic formation task can be successfully completed without deviation. However, in actual operation, due to uncertainties such as external interference, the actual running curve of the train may deviate from the ideal speed curve. After a large deviation, continuing to control the train according to the offline strategy will not meet the goal of accurate formation, and may even lead to the danger of rear-end collision. Therefore, online real-time control of the train is required to correct the deviation. However, in the above dynamic programming algorithm, the train motion model is accurate and the algorithm solution time is long, while online control has high real-time requirements. Therefore, further optimization is needed when using dynamic programming algorithm to implement online real-time control. According to an embodiment of the present invention, the present invention uses a sequential quadratic programming rolling optimization method to perform rolling optimization control on the following train. The sequential quadratic programming method has the characteristics of good convergence and high computational efficiency. Therefore, the rolling optimization process of the present invention can ensure rapid optimization of the speed curve and meet the real-time control requirements.

[0125] During dynamic train formation, when the deviation between the actual train speed curve and the ideal speed curve exceeds a certain threshold, it is determined to be in a deviation state. The sequential quadratic programming rolling optimization control method of the present invention is then used to control the train in real time after the deviation occurs.

[0126] According to one embodiment of the present invention, such as Figure 4As shown, the basic principle of rolling optimization control is to perform rolling optimization at each decision stage. The steps can be simply summarized as follows: First, at the moment when a deviation is detected, determine the current formation stage of the train and calculate the remaining formation time; then, acquire data such as the speed and position of the preceding train through vehicle-to-vehicle communication, calculate the positional interval and coordination distance between the two trains at the current stage, re-establish the dynamic formation optimization problem, and use a sequential quadratic programming algorithm to solve the optimal control strategy for the train; finally, propose a rolling optimization strategy to implement the above optimization control process at each decision stage. Each step is explained in detail below.

[0127] (1) Train status data acquisition

[0128] The rolling optimization problem requires data such as the current stage number of the train, the positions of the preceding and following cars, and the distance from the preceding car to the end of the train section. The stage number is obtained from the onboard train control module, and the positions of the preceding car and the distance from the preceding car to the end of the train section at the k-th decision time are also considered. The preceding vehicle sends information to the following vehicle via vehicle-to-vehicle communication, and the following vehicle's position is obtained based on its onboard positioning module. The coordination distance for the current stage is calculated based on this data. Taking stage k as an example, the coordination distance l... cd,k =d inter,k -d vcm , where l cd,k and d inter,k Let d represent the coordination distance and the interval between two vehicles in the k-th decision stage of the formation process, respectively. vcm The ideal grouping interval.

[0129] (2) Establishment of the optimization problem

[0130] The goal of online control is to correct speed deviations caused by external disturbances, ensuring that the distance between trains approaches the ideal interval at the end of the train formation, and minimizing energy consumption. Therefore, the optimization problem can be described as:

[0131]

[0132] Constraints:

[0133]

[0134] u k = u(v k ,v k+1 (Equation 29)

[0135]

[0136] u k ∈U (Equation 31)

[0137]

[0138]

[0139]

[0140] In equation 27, f′ n It is the energy consumption indicator of the grouping process, e k It is the energy consumption value of a single sub-interval, where x represents the sub-interval number where the train is currently in the decision-making stage, i.e., the actual running speed of the following train deviates from the offline speed curve. This indicates the change in the vehicle's position after the k-th sub-interval. l represents the distance traveled by the vehicle before the k-th sub-interval. cd,k Δl represents the ideal reduction in the distance between the vehicle before and after it in the k-th sub-interval. error_max Equation 28 represents the maximum allowable distance error during virtual formation. It is the running distance constraint of the following vehicle. By constraining the difference in the running distance between the two vehicles within the allowable range of coordination distance error, the deviation is corrected and precise formation is achieved. Equation 29 is the state transition constraint. Equation 30 is the speed constraint. Equation 31 is the control force constraint. Equation 32 is the speed terminal constraint. This is the initial solution for solving the optimization problem. This represents the speed of the k-th sub-interval in the offline speed curve calculated by dynamic programming, where k0 represents the sub-interval number where the actual speed curve of the following vehicle deviates from the offline speed curve for the first time. Let represent the velocity of the (k-1)th sub-interval in the velocity curve calculated using the sequential quadratic programming method, and Equation 33 represents the initial solution constraint. In the stage k0 where the deviation is first detected, the offline optimal solution obtained by dynamic programming is used as the initial solution for the optimization calculation process. In subsequent optimization calculations, the optimal solution obtained in the previous stage is used as the initial solution to improve the solution speed.

[0141] Another important factor limiting the efficiency of solving optimization problems is the overly precise description of the train formation process model. According to an embodiment of the present invention, to further improve the solution efficiency, the energy consumption equation in the state transition constraint equation and the objective function is simplified: the train operation state of a single sub-section is simplified to uniformly variable speed, and state transition constraint 29 is changed to equation 34; the energy consumption calculation process ignores the change in resistance within the sub-section, and the energy consumption e within a single section is obtained according to the law of conservation of energy. k As shown in Equation 35:

[0142]

[0143]

[0144] in,

[0145]

[0146] E k (v)=(F b (v)+F a +F R )×Δs k (Equation 37)

[0147] In equation 35, e(v) k ) and e(v k+1 E represents the train's kinetic energy at the start and end times of the sub-interval. k This refers to the energy consumed to overcome basic and additional resistance. Consider a special case where the energy consumption calculated according to Equation 35 during train braking may be negative. In this case, to maintain consistency with the dynamic programming method, the energy consumption value is set to zero.

[0148] (3) Optimization problem solving

[0149] Based on the optimization problem established above, the grouping control problem is transformed into a nonlinear optimization problem with a quadratic objective function. Sequential quadratic programming is one of the most effective methods for solving nonlinear optimization problems with quadratic objective functions. Compared with other optimization algorithms, sequential quadratic programming has advantages such as good convergence and fast convergence speed.

[0150] According to one embodiment of the present invention, the present invention utilizes the MATLAB nonlinear tool fmincon function to perform the calculation of the optimization problem using a sequential quadratic programming algorithm. The fmincon function is as follows:

[0151] [v,fval]=fmincon(fun,v0,A,b,Aeq,beq,lb,ub,nonlcon,options)

[0152] This function is an effective tool for solving nonlinear optimization problems in MATLAB. [v, fval] represents the function's return value, where v represents the optimal speed curve at the end of the iteration; fval represents the value of the objective function, i.e., the energy consumption value of the train formation process. fun is the objective function, i.e., Equation 27; v0 is the initial value of the decision vector, i.e., the initial value of the speed curve; A, b, Aeq, beq define linear constraints, i.e., constraints 32 and 33; lb and ub are the lower and upper limits of the decision, i.e., the train speed limit constraint 32; nonlcon is the defined nonlinear constraint, i.e., constraint 28. options defines the optimization parameters, specifically set as: options = optimoptions('fmincon', 'MaxIter', 100, 'Algorithm', 'sqp').

[0153] To ensure the efficiency of iterative calculations, the maximum number of iterations for the algorithm is set to 100.

[0154] (4) Train rolling optimization control

[0155] The speed curve and optimal control strategy are obtained by calculating the sequential quadratic programming algorithm. This invention adopts the predictive control rolling optimization principle, which applies the first control effect of the control strategy to the train. When entering the next decision stage, the train data is collected again and the rolling optimization control is completed based on this.

[0156] This invention utilizes online rolling optimization control based on dynamic programming and sequential quadratic programming. Under normal conditions, the train operates according to the offline dynamic formation curve. When the onboard module detects a deviation, it switches to the rolling optimization control algorithm.

[0157] As can be seen from the description of the above embodiments, the present invention adopts a combination of dynamic programming and sequential quadratic programming. Dynamic programming can ensure the rapid acquisition of the optimal offline speed curve, while sequential quadratic programming can ensure the rapid online rolling optimization of the speed curve. Thus, the purpose of transitioning the train from moving block state to virtual formation state is achieved while ensuring computational efficiency.

[0158] It should be noted that uncontrollable abnormal situations may occur during dynamic train formation, endangering train safety. This invention proposes different handling measures for these uncontrollable abnormal situations: Regarding the situation where the preceding train brakes suddenly, if the preceding train detects a dangerous point on the route ahead and performs emergency braking, while the following train is still in the process of dynamic formation, to prevent a rear-end collision, the preceding train should immediately send a warning message via vehicle-to-vehicle communication. Upon receiving the message, the following train should also take emergency braking measures. Regarding the situation where the preceding train does not follow the prescribed curve, since the calculation of formation time is based on the premise that the preceding train follows the prescribed speed and curve, if the preceding train does not follow the predetermined curve, the formation time will be calculated incorrectly, and the running distance of the preceding train during the formation process will also be incorrect. Figure 5 As shown, s end To determine the ideal position of the train leading at the end of the formation, s′ endThe actual position of the preceding vehicle at the end of the formation time is crucial; any deviation between the two will affect the accuracy of the formation. To ensure smooth formation in such scenarios, this invention configures the onboard equipment of the preceding vehicle to monitor its operating speed. When a deviation is detected between the actual curve and the predetermined curve, a request to update the formation information is immediately sent to the ground. The ground equipment then re-plans the formation area based on the positions of the preceding and following vehicles. If, after calculation, the current conditions are deemed unsuitable for formation, the ground sends a formation cancellation command to the following vehicle and provides driving permission guidance. If the ground equipment successfully plans the formation area, it sends the newly defined formation information to the preceding vehicle. Upon receiving the information, the preceding vehicle sends its speed curve to the following vehicle via vehicle-to-vehicle communication. The following vehicle then recalculates the formation time and begins online rolling optimization control. In the event of a train-to-train communication failure, since the establishment of train-to-train communication is a prerequisite for dynamic train formation control, the preceding train sends position information to the following train for online control. If train-to-train communication fails, the following train cannot obtain accurate train intervals and therefore cannot calculate control strategies to control the train. The preceding train should immediately report its operating status to the ground equipment. The ground equipment commands the following train to switch to offline speed curve control mode and attempt to reconnect. The reconnection attempt time can be set to 2 seconds (or other times depending on actual control requirements; this invention does not limit this). Even if external interference causes the following train's speed to deviate from the offline curve, the speed deviation within 2 seconds can still be corrected by the online control method. If the reconnection is successful, the following train immediately switches to online rolling optimization control mode. If the reconnection fails, the following train immediately applies emergency braking and waits for instructions from the ground equipment.

[0159] The dynamic grouping control scheme combining dynamic programming and sequential quadratic programming in this embodiment of the invention can effectively achieve virtual grouping and close tracking.

[0160] Based on the dynamic formation scheme of the above embodiments, this invention proposes a train control system based thereon. The system includes: a ground control module, used to issue control commands to the train to control its operation, and to issue dynamic formation commands to the train when train interval control is required; multiple formation control modules, each dynamic formation module being installed on a train, wherein, upon receiving a dynamic formation command from the ground control module, the dynamic formation module on the following train performs dynamic formation according to the dynamic formation method of this invention to obtain the optimal control strategy and the speed curve corresponding to the optimal control strategy during the dynamic formation process; and multiple on-board control modules, each on-board control module being installed on a train, used to control the following train to run according to the speed curve based on the optimal control strategy obtained by the dynamic formation module on its train until the virtual formation is completed. Each train formation control module includes: an information acquisition module, used to acquire information about its own train and the train ahead of it when receiving a dynamic formation command. The information about the own train includes the speed and position of the train behind it, and the information about the train ahead of it includes the speed curve, position, and speed of the train ahead of it; a dynamic planning module, used to construct a first optimization objective that satisfies the virtual formation constraints based on the information about its own train and the train ahead of it acquired by the information acquisition module, with the goal of accurate formation and low energy consumption. It then uses a dynamic planning method to calculate the optimal control strategy and the corresponding offline speed curve during the dynamic formation process based on the optimization objective, and loads the offline speed curve into the corresponding on-board control module; a monitoring module, used to monitor the operation of its own train in real time, and send a rolling optimization command to the sequence planning module when the actual speed curve of the train deviates from the offline speed curve; and a sequence planning module, used to calculate and update the optimal control strategy and speed curve of its own train in the subsequent stages of virtual formation based on the rolling optimization command and using a sequence quadratic programming method, and loads the updated speed curve into the corresponding on-board control module.

[0161] This invention utilizes a dynamic train formation control scheme combining dynamic programming and sequential quadratic programming to transition trains from a moving block state to a virtual formation state. By establishing a target model for dynamic train formation control and based on the dynamic programming algorithm, a formation curve calculation method is proposed. The speed curve is calculated for the train to be formed through offline optimization. The dynamic programming algorithm provides an ideal speed curve for the train, guiding the formation process under interference-free conditions, achieving high formation accuracy and low energy consumption. Based on the offline speed curve and combined with the sequential quadratic programming algorithm, an online sequential quadratic programming rolling optimization control method is proposed. This method can effectively provide online control for the train under external interference conditions while ensuring computational efficiency, correcting speed deviations caused by external disturbances. This not only improves formation efficiency but also significantly enhances train operation efficiency.

[0162] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0163] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0164] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0165] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0166] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0167] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A dynamic grouping method applied to virtual grouping, characterized in that, The method includes: In response to the command for dynamic train formation, the train formation section is obtained and adjacent trains establish car-to-car communication to obtain information about the preceding and following cars in the adjacent trains; Based on the acquired information of the preceding and following vehicles, a first optimization objective is constructed to meet the virtual formation constraints with the goals of accurate formation and low energy consumption. The dynamic programming method is then used to calculate the optimal control strategy and the corresponding offline speed curve during the virtual formation process based on the first optimization objective. The following vehicle operates according to the offline speed curve based on the optimal control strategy. During its operation, the following vehicle's status is monitored in real time. When the actual speed curve of the following vehicle deviates from the offline speed curve, a sequential quadratic programming method is used to calculate and update the optimal control strategy and speed curve for subsequent stages. The following vehicle then operates according to the updated speed curve based on the updated optimal control strategy until dynamic grouping is completed. The steps for calculating and updating the optimal control strategy and speed curve for subsequent stages using the sequential quadratic programming method are as follows: The system can obtain the current positions of the preceding and following trains in real time and determine the train's formation decision-making stage based on this information. A second optimization objective that satisfies the virtual grouping constraint is constructed with low energy consumption as the goal, and the optimal control strategy and the corresponding velocity curve of the subsequent grouping decision stage are calculated based on the second optimization objective using the sequential quadratic programming method.

2. The method according to claim 1, characterized in that, The method further includes: after the first detection that the actual operating speed curve of the following vehicle deviates from the offline speed curve, in each subsequent grouping decision stage, a sequential quadratic programming method is used to continuously calculate the optimal control strategy and the speed curve corresponding to the optimal control strategy for the following vehicle in each subsequent grouping decision stage.

3. The method according to any one of claims 1-2, characterized in that, The method further includes: Based on the speed curve of the preceding vehicle, the system monitors the operation of the preceding vehicle in real time. When the preceding vehicle brakes suddenly, it sends a warning message to the following vehicle via vehicle-to-vehicle communication to enable the following vehicle to take emergency braking action. When the preceding vehicle deviates from its speed curve, the system updates the dynamic formation request and terminates the dynamic formation when the formation conditions are not met.

4. The method according to claim 3, characterized in that, The method further includes: When vehicle-to-vehicle communication between the preceding and following vehicles is lost, the following vehicle operates according to the offline speed curve. After re-establishing vehicle-to-vehicle communication with the preceding vehicle, the following vehicle monitors whether its actual operating speed curve deviates from the offline speed curve. When the following vehicle's actual operating speed curve deviates from the offline speed curve, the optimal control strategy and speed curve for the subsequent stage are calculated and updated using a sequential quadratic programming method. The following vehicle controls itself to operate according to the updated speed curve based on the updated optimal control strategy until dynamic formation is completed.

5. A train control system, characterized in that, The system is configured to perform dynamic train formation using the method described in any one of claims 1-4, and to closely track trains based on the speed at the end of the formation after the dynamic formation is completed.

6. The system according to claim 5, characterized in that, The system includes: The ground control module is used to issue control commands to the train to control its operation, and to issue dynamic formation commands to the train when train interval control is required. Multiple formation control modules are provided, each dynamic formation module is set on a train, wherein when a dynamic formation command is received from the ground control module, the dynamic formation module on the following train performs dynamic formation according to the method described in any one of claims 1-4 to obtain the optimal control strategy and the speed curve corresponding to the optimal control strategy during the dynamic formation process. Multiple on-board control modules are installed on a train. Each on-board control module is used to control the following trains to run according to the speed curve based on the optimal control strategy obtained from the dynamic formation module on its train until the virtual formation is completed.

7. The system according to claim 6, characterized in that, Each grouping control module includes: The information acquisition module is used to acquire information about the train it is in and the train ahead of it when it receives a dynamic formation command. The information about the train it is in includes the speed and position of the train behind it, and the information about the train ahead of it includes the speed curve, position and speed of the train ahead of it. The dynamic programming module is used to construct a first optimization objective that satisfies the virtual formation constraints based on the information of the train and its preceding train obtained by the information acquisition module, with the goal of accurate formation and low energy consumption. The dynamic programming method is used to calculate the optimal control strategy and the corresponding offline speed curve in the dynamic formation process based on the optimization objective, and the offline speed curve is loaded into the corresponding on-board control module. The monitoring module is used to monitor the operation of the train in real time. When the actual operating speed curve of the train deviates from the offline speed curve, it sends a rolling optimization command to the sequence planning module. The sequence planning module is used to calculate and update the optimal control strategy and speed curve of the train in the subsequent stage of virtual formation based on the rolling optimization command and the sequence quadratic programming method, and then load the updated speed curve into the corresponding on-board control module.

8. A computer-readable storage medium, characterized in that, It stores a computer program that can be executed by a processor to implement the steps of the method according to any one of claims 1 to 4.

9. An electronic device, characterized in that, include: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the electronic device to perform the steps of the method as described in any one of claims 1 to 4.

Images