Train operation control method and device, electronic equipment and storage medium

Abstract

Embodiments of the present application disclose a train operation control method and device, electronic equipment and storage medium, belonging to the technical field of train control. The method comprises: when the crosswind speed is equal to or greater than a first preset threshold, determining a plurality of candidate guide vane configuration parameters according to the crosswind speed and a first running speed of the train; generating a plurality of groups of candidate parameters according to the crosswind speed, the first running speed and the plurality of candidate guide vane configuration parameters, inputting each group of candidate parameters into a pre-trained guide vane adjustment model to obtain a predicted aerodynamic load coefficient corresponding to each group of candidate parameters; in the predicted aerodynamic load coefficient corresponding to each group of candidate parameters, determining the candidate guide vane configuration parameter corresponding to the smallest predicted aerodynamic load coefficient as a target guide vane configuration parameter; and adjusting the running state of the train according to the target guide vane configuration parameter. The present application can timely control the train operation in the crosswind environment, thereby improving the train operation safety.

Description

Technical Field

[0001] This application relates to the field of train control technology, specifically to a train operation control method, device, electronic equipment, and storage medium. Background Technology

[0002] Due to the unique body structure of trains and their reliance on wheel-rail constraints for close-to-the-ground operation, trains have a significantly larger lateral wind-exposed area than other modes of transportation. Therefore, crosswinds can cause a sharp drop in lateral stability, leading to violent swaying of the train body. In extreme wind conditions, this poses a serious safety hazard, potentially causing derailment, overturning, and even significant casualties. Related technologies primarily mitigate the impact of strong crosswinds by improving the train's aerodynamic shape and adding fixed wind-resistant structures along the railway line.

[0003] However, the fixed aerodynamic shape of the train or the flow field mitigation measures along the line cannot flexibly adapt to changes in train speed and crosswind along the line, making it difficult to make timely adjustments in complex and ever-changing real crosswind environments, thereby reducing the train's operating safety. Summary of the Invention

[0004] This application provides a train operation control method, device, electronic equipment, and storage medium, which can timely adjust the train operation in crosswind conditions, thereby improving the train's driving safety.

[0005] To achieve the above objectives, one embodiment of this application provides a train operation control method, including:

[0006] Obtain the crosswind speed of the train's operating environment during operation;

[0007] When the crosswind speed is equal to or greater than the first preset threshold, multiple candidate guide vane configuration parameters are determined based on the crosswind speed and the first running speed of the train.

[0008] Multiple sets of candidate parameters are generated based on crosswind speed, first operating speed and multiple candidate guide vane configuration parameters. Each set of candidate parameters is input into a pre-trained guide vane adjustment model to obtain the predicted aerodynamic load coefficient corresponding to each set of candidate parameters. The pre-trained guide vane adjustment model is trained based on sample crosswind speed, sample operating speed, sample guide vane configuration parameters and label aerodynamic load coefficient.

[0009] Among the predicted aerodynamic load coefficients corresponding to each group of candidate parameters, the candidate guide vane configuration parameter corresponding to the smallest predicted aerodynamic load coefficient is determined as the target guide vane configuration parameter.

[0010] The train's operating status is adjusted based on the target air deflector configuration parameters.

[0011] In some embodiments, after obtaining the crosswind speed of the operating environment in which the train is operating, the method further includes:

[0012] When the crosswind speed is greater than the second preset threshold and less than the first preset threshold, the real-time aerodynamic load coefficient of the train and the configuration parameters of the train's first guide vane are obtained.

[0013] The updated guide vane configuration parameters are obtained by updating the first guide vane configuration parameters based on the real-time aerodynamic load coefficient;

[0014] The train's operating status is adjusted based on the updated guide vane configuration parameters.

[0015] In some embodiments, updating the configuration parameters of the first guide vane based on the real-time aerodynamic load coefficient to obtain the updated guide vane configuration parameters includes:

[0016] Obtain multiple historical aerodynamic load coefficients of the train within the target time period;

[0017] Based on multiple historical aerodynamic load coefficients, the fluctuation value of the aerodynamic load coefficient of the train within the target time period is determined;

[0018] The initial deflector deployment angle of the train is determined based on the configuration parameters of the first deflector, and the real-time control energy efficiency coefficient is determined based on the ratio between the real-time aerodynamic load coefficient and the initial deflector deployment angle.

[0019] The deployment angle of the train's updated guide vanes is determined based on the ratio between the aerodynamic load coefficient fluctuation value and the real-time control energy efficiency coefficient.

[0020] The configuration parameters of the first guide vane are updated based on the updated guide vane deployment angle to obtain the updated guide vane configuration parameters.

[0021] In some embodiments, before obtaining the crosswind speed of the operating environment in which the train is operating, the method further includes:

[0022] Acquire meteorological data of the crosswind area and the train's second operating speed;

[0023] The predicted crosswind speed in the crosswind area is determined based on meteorological data, and the configuration parameters of the train's second guide vane are determined based on the predicted crosswind speed and the second operating speed.

[0024] The train's operating status is adjusted according to the configuration parameters of the second guide vane.

[0025] In some embodiments, multiple candidate air deflector configuration parameters are determined based on the crosswind speed and the train's first operating speed, including:

[0026] Multiple initial guide vane configuration parameters were determined based on the crosswind speed and the train's initial operating speed.

[0027] Determine the parameter level to which each initial deflector configuration parameter belongs, and determine the degree of parameter crowding between each initial deflector configuration parameter and other initial deflector configuration parameters at the same parameter level;

[0028] Multiple initial guide vane configuration parameters are filtered based on parameter level and parameter crowding level to obtain the filtered initial guide vane configuration parameters;

[0029] Obtaining genetic manipulation factors;

[0030] The initial guide vane configuration parameters after screening are updated based on the genetic operation factors to obtain the updated initial guide vane configuration parameters;

[0031] Return to the steps of determining the parameter level to which each initial guide vane configuration parameter belongs, until a preset iteration stop condition is reached, resulting in multiple candidate guide vane configuration parameters.

[0032] In some embodiments, multiple sets of candidate parameters are generated based on crosswind speed, a first operating speed, and multiple candidate guide vane configuration parameters, including:

[0033] Multiple sets of initial candidate parameters are generated based on crosswind speed, first operating speed, and multiple candidate guide vane configuration parameters;

[0034] The mean and standard deviation of the parameters are determined based on multiple sets of initial candidate parameters;

[0035] For each set of initial candidate parameters, the parameter difference is determined based on the difference between the initial candidate parameter and the parameter mean.

[0036] For each set of initial candidate parameters, the normalized candidate parameters are determined based on the ratio between the parameter difference and the parameter standard deviation, thus obtaining multiple sets of candidate parameters.

[0037] In some embodiments, before inputting each set of candidate parameters into the pre-trained airfoil adjustment model, the method further includes:

[0038] Obtain parameters from multiple sets of samples;

[0039] Each set of sample parameters is input into the initial guide vane adjustment model to obtain the sample aerodynamic load coefficient corresponding to each set of sample parameters;

[0040] The total loss value is determined based on the aerodynamic load coefficient of each sample and the aerodynamic load coefficient of the label corresponding to the sample aerodynamic load coefficient;

[0041] The guide vane adjustment model is iteratively trained based on the total loss value until the model loss value is less than the preset loss threshold, thus obtaining the trained guide vane adjustment model.

[0042] In some embodiments, obtaining multiple sets of sample parameters includes:

[0043] Obtain the range of crosswind speed, train operating speed, guide vane height, number of guide vanes deployed, and guide vane deployment angle;

[0044] The ranges of crosswind speed, train speed, guide vane height, number of guide vanes deployed, and guide vane deployment angle were divided into intervals to obtain multiple parameter intervals.

[0045] Sample each parameter interval to obtain the sample parameters corresponding to each parameter interval;

[0046] Multiple sets of sample parameters are determined based on the sample parameters corresponding to each parameter range.

[0047] In some embodiments, the total loss value is determined based on the aerodynamic load coefficient of each sample and the tag aerodynamic load coefficient corresponding to the sample aerodynamic load coefficient, including:

[0048] Determine the sample lateral force coefficient and sample lateral rolling moment coefficient corresponding to the aerodynamic load coefficient of each sample;

[0049] The first loss value is determined based on the lateral force coefficient of each sample and the aerodynamic load coefficient of the first tag corresponding to the lateral force coefficient of each sample;

[0050] The second loss value is determined based on the side rolling moment coefficient of each sample and the aerodynamic load coefficient of the second label corresponding to the side rolling moment coefficient of each sample;

[0051] The total loss value is determined based on the first loss value and the second loss value.

[0052] To achieve the above objectives, one embodiment of this application provides a train operation control device, comprising:

[0053] The acquisition module is used to acquire the crosswind speed of the train's operating environment during operation;

[0054] The candidate guide vane configuration parameter determination module is used to determine multiple candidate guide vane configuration parameters based on the crosswind speed and the first running speed of the train when the crosswind speed is equal to or greater than a first preset threshold.

[0055] The predictive aerodynamic load coefficient determination module is used to generate multiple sets of candidate parameters based on crosswind speed, first operating speed and multiple candidate guide vane configuration parameters. Each set of candidate parameters is input into the pre-trained guide vane adjustment model to obtain the predicted aerodynamic load coefficient corresponding to each set of candidate parameters. The pre-trained guide vane adjustment model is trained based on sample crosswind speed, sample operating speed, sample guide vane configuration parameters and labeled aerodynamic load coefficient.

[0056] The target airfoil configuration parameter selection module is used to determine the candidate airfoil configuration parameter corresponding to the smallest predicted aerodynamic load coefficient among the predicted aerodynamic load coefficients corresponding to each group of candidate parameters as the target airfoil configuration parameter.

[0057] The train operation status adjustment module is used to adjust the train's operation status according to the target guide vane configuration parameters.

[0058] To achieve the above objectives, one aspect of this application provides a computer-readable storage medium storing multiple instructions adapted for loading by a processor to execute the steps in the train operation control method provided in this application.

[0059] To achieve the above objectives, one aspect of this application provides a computer device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program, it implements the steps in the train operation control method provided in this application.

[0060] To achieve the above objectives, one aspect of this application provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement the steps in the train operation control method provided in this application.

[0061] The train operation control method, device, electronic equipment, and storage medium proposed in this application acquire the crosswind speed of the operating environment in which the train is located during operation; when the crosswind speed is equal to or greater than a first preset threshold, multiple candidate guide vane configuration parameters are determined based on the crosswind speed and the first operating speed of the train; multiple sets of candidate parameters are generated based on the crosswind speed, the first operating speed, and the multiple candidate guide vane configuration parameters, and each set of candidate parameters is input into a pre-trained guide vane adjustment model to obtain the predicted aerodynamic load coefficient corresponding to each set of candidate parameters. The pre-trained guide vane adjustment model is trained based on sample crosswind speed, sample operating speed, sample guide vane configuration parameters, and labeled aerodynamic load coefficients; among the predicted aerodynamic load coefficients corresponding to each set of candidate parameters, the candidate guide vane configuration parameter corresponding to the smallest predicted aerodynamic load coefficient is determined as the target guide vane configuration parameter; and the train's operating state is adjusted according to the target guide vane configuration parameter.

[0062] This application embodiment, by real-time monitoring of train operation status and introducing a multivariable pre-trained guide vane adjustment model, rapidly predicts aerodynamic load coefficients under multiple candidate guide vane configuration parameters. This avoids the high computational cost problem of traditional computational fluid dynamics, which involves excessively long solutions, and realizes the transformation from offline high-cost calculation to online rapid query and optimization. Furthermore, by selecting the target guide vane configuration parameters corresponding to the minimum aerodynamic load coefficient to dynamically adjust the train operation status, it can accurately and effectively suppress the formation and development of leeward turbulent vortices, reduce the pressure difference between the two sides of the train, and improve train operation safety. The pre-trained guide vane adjustment model in this application embodiment can quickly determine the target guide vane configuration parameters, thereby enabling timely train operation control in crosswind environments. Moreover, the adjusted train operation status not only significantly reduces the lateral force and roll moment experienced during travel but also significantly improves the train's operational stability and environmental adaptability in crosswind environments, effectively avoiding potential hazards such as vehicle body swaying, derailment, or overturning under extreme wind conditions. Thus, this application embodiment achieves real-time and efficient protection of high-speed train operation safety.

[0063] Other features and advantages of this application will be set forth in the following description and will be apparent in part from the description or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the description, claims and drawings. Attached Figure Description

[0064] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0065] Figure 1 This is a schematic diagram of the system framework corresponding to the train operation control method provided in the embodiments of this application;

[0066] Figure 2 This is a schematic flowchart of the train operation control method provided in the embodiments of this application;

[0067] Figure 3 This is a comparative schematic diagram showing the influence of the guide vane on the lateral force coefficient of the train provided in the embodiments of this application;

[0068] Figure 4 This is a comparative schematic diagram showing the influence of the guide vane on the train's side roll moment coefficient provided in the embodiments of this application;

[0069] Figure 5This is a schematic diagram of the hidden layer data processing of the guide vane adjustment model provided in the embodiments of this application;

[0070] Figure 6 This is a schematic diagram comparing the effects of different guide vane heights on the train's lateral force coefficient, provided in an embodiment of this application.

[0071] Figure 7 This is a schematic diagram comparing the effects of different guide vane heights on the train's side roll moment coefficient, provided in an embodiment of this application.

[0072] Figure 8 This is a schematic diagram comparing the effects of different guide vane deployment heights on train aerodynamic loads, provided in an embodiment of this application.

[0073] Figure 9 This is a schematic diagram comparing the influence of the same guide vane deployment height on the train's lateral force coefficient, provided in an embodiment of this application.

[0074] Figure 10 This is a schematic diagram comparing the effects of the same guide vane deployment height on the train's side roll moment coefficient, provided in the embodiments of this application.

[0075] Figure 11 This is a schematic diagram of the module structure of the train operation control device provided in the embodiments of this application;

[0076] Figure 12 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0077] To enable those skilled in the art to better understand the solutions of this application, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0078] It should be noted that in all specific embodiments of this application, when it is necessary to obtain the train's operating speed, permission or consent from the relevant personnel managing the train is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when this application embodiment needs to obtain sensitive personal information of relevant personnel, separate permission or consent from the relevant personnel is obtained through pop-up windows or redirection to a confirmation page. Only after obtaining the separate permission or consent of the relevant personnel is the train's operating speed, which is necessary for the normal operation of this application embodiment, obtained. Other data obtained in this application embodiment are all authorized and legal data, and will not be elaborated upon here.

[0079] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, programmable consumer computer devices, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0080] The technical problems existing in the related technologies are as follows:

[0081] Due to the unique body structure of trains and their reliance on wheel-rail constraints for close-to-the-ground operation, trains have a significantly larger lateral wind-exposed area than other modes of transportation. Therefore, crosswinds can cause a sharp drop in lateral stability, leading to violent swaying of the train body. In extreme wind conditions, this poses a serious safety hazard, potentially causing derailment, overturning, and even significant casualties. Related technologies primarily mitigate the impact of strong crosswinds by improving the train's aerodynamic shape and adding fixed wind-resistant structures along the railway line.

[0082] However, the fixed aerodynamic shape of the train or the flow field mitigation measures along the line cannot flexibly adapt to changes in train speed and crosswind along the line, making it difficult to make timely adjustments in complex and ever-changing real crosswind environments, thereby reducing the train's operating safety.

[0083] The train operation control method, device, electronic equipment, and storage medium proposed in this application acquire the crosswind speed of the operating environment in which the train is located during operation; when the crosswind speed is equal to or greater than a first preset threshold, multiple candidate guide vane configuration parameters are determined based on the crosswind speed and the first operating speed of the train; multiple sets of candidate parameters are generated based on the crosswind speed, the first operating speed, and the multiple candidate guide vane configuration parameters, and each set of candidate parameters is input into a pre-trained guide vane adjustment model to obtain the predicted aerodynamic load coefficient corresponding to each set of candidate parameters. The pre-trained guide vane adjustment model is trained based on sample crosswind speed, sample operating speed, sample guide vane configuration parameters, and labeled aerodynamic load coefficients; among the predicted aerodynamic load coefficients corresponding to each set of candidate parameters, the candidate guide vane configuration parameter corresponding to the smallest predicted aerodynamic load coefficient is determined as the target guide vane configuration parameter; and the train's operating state is adjusted according to the target guide vane configuration parameter.

[0084] This application embodiment, by real-time monitoring of train operation status and introducing a multivariable pre-trained guide vane adjustment model, rapidly predicts aerodynamic load coefficients under multiple candidate guide vane configuration parameters. This avoids the high computational cost problem of traditional computational fluid dynamics, which involves excessively long solutions, and realizes the transformation from offline high-cost calculation to online rapid query and optimization. Furthermore, by selecting the target guide vane configuration parameters corresponding to the minimum aerodynamic load coefficient to dynamically adjust the train operation status, it can accurately and effectively suppress the formation and development of leeward turbulent vortices, reduce the pressure difference between the two sides of the train, and improve train operation safety. The pre-trained guide vane adjustment model in this application embodiment can quickly determine the target guide vane configuration parameters, thereby enabling timely train operation control in crosswind environments. Moreover, the adjusted train operation status not only significantly reduces the lateral force and roll moment experienced during travel but also significantly improves the train's operational stability and environmental adaptability in crosswind environments, effectively avoiding potential hazards such as vehicle body swaying, derailment, or overturning under extreme wind conditions. Thus, this application embodiment achieves real-time and efficient protection of high-speed train operation safety.

[0085] For example, a high-speed train traveling at high speed on a mountainous or coastal railway line with complex terrain will frequently pass through long tunnels, viaducts, or natural mountain passes, creating a typical scenario of variable crosswinds. The crosswind speed and direction in these specific sections often change drastically in an instant due to rapid changes in terrain and sudden changes in weather conditions.

[0086] If traditional static protection solutions are used, relying solely on the streamlined front design of the train at the factory, or on fixed windbreaks and barriers built along the railway line, the inherent limitation of these physical protective structures—that they cannot deform or move in real time—often makes it difficult to achieve precise safety coverage across the entire railway line and at all times. If a train suddenly exits a tunnel at a high speed of 350 km / h and enters a strong wind valley (for example, when the crosswind speed suddenly increases to 50 m / s), the pre-set fixed aerodynamic shape and barriers along the line can only be statically optimized for certain specific average train speeds or wind speeds, and cannot generate a dynamic flow field mitigation effect based on the current instantaneous extreme wind speed and train operating status.

[0087] Understandably, traditional methods cannot control train operation in response to changing crosswind conditions. Insufficiently timely adjustments cause strong turbulent vortices to accumulate and dissipate rapidly on the leeward side of the train. The train body is instantly subjected to enormous lateral forces and rolling moments far exceeding expectations. At this point, the train control system or driver has to passively take emergency braking or drastic speed reduction measures to barely maintain lateral stability. This not only disrupts the train's original smooth operating rhythm and causes delays and running delays in key sections, but also makes it very easy to break through the physical constraint critical point between the wheels and rails under extreme and sudden wind conditions, seriously threatening the overall operational safety of the train system and the safety of passengers' lives.

[0088] The specific details regarding the train operation control method, device, electronic equipment, and storage medium provided in the embodiments of this application will be described in detail below.

[0089] Please see Figure 1 , Figure 1 This is a schematic diagram of the system framework corresponding to the train operation control method provided in this application embodiment. The train operation control method provided in this application embodiment can be applied to this system framework.

[0090] It includes terminal 140, Internet 130, gateway 120, server 110, etc.

[0091] Terminal 140 or server 110 can be a device that performs train operation control methods. For example, terminal 140 can be installed on the train, and server 110 can be installed in a third-party control center, and train operation control can be achieved by interacting with terminal 140 installed on the train.

[0092] Terminal 140 includes, but is not limited to, mobile phones, tablets, computers, and intelligent computing centers. Terminal 140 can be a single device or a collection of multiple devices. For example, multiple computers can be interconnected via a local area network, sharing a single monitor to work collaboratively, thus forming a terminal 140. Terminal 140 can communicate with the Internet 130 via wired or wireless means to exchange data.

[0093] Server 110 refers to a computer system that can provide certain services to terminal 140. Compared to ordinary terminal 140, server 110 has higher requirements in terms of stability, security, and performance. Server 110 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0094] Gateway 120, also known as an internetwork connector or protocol converter, is a computer system or device that acts as a translator, enabling network interconnection at the transport layer. It bridges the gap between two systems using different communication protocols, data formats, languages, or even completely different architectures. Gateways can also provide filtering and security functions. Messages sent from terminal 140 to server 110 are forwarded to the corresponding server 110 via gateway 120. Messages sent from server 110 to terminal 140 are also forwarded to the corresponding terminal 140 via gateway 120.

[0095] The embodiments of this application can be applied to various scenarios, such as traditional trunk line high-speed passenger EMUs, intercity EMUs and fast passenger trains, high-speed freight EMUs and high-speed maglev trains, etc. This is only an example and does not mean that the embodiments of this application limit the scenarios in which the train operation control method is applied.

[0096] Next, we will describe it from the perspective of train operation control devices, such as... Figure 2 As shown, Figure 2 This is a schematic flowchart of the train operation control method provided in the embodiments of this application. The train operation control method is applied to a train operation control device. Figure 2 The method may include, but is not limited to, the following steps 210 to 250. When the train operation control device executes the train operation control method, the specific process is as follows. It should be noted first that this embodiment... Figure 2 The order of steps 210 to 250 is not specifically limited. The order of steps can be adjusted or some steps can be reduced or added according to actual needs.

[0097] Step 210: Obtain the crosswind speed of the operating environment in which the train is located during operation;

[0098] Step 220: When the crosswind speed is equal to or greater than the first preset threshold, multiple candidate guide vane configuration parameters are determined based on the crosswind speed and the first running speed of the train.

[0099] Step 230: Generate multiple sets of candidate parameters based on crosswind speed, first operating speed and multiple candidate guide vane configuration parameters, and input each set of candidate parameters into the pre-trained guide vane adjustment model to obtain the predicted aerodynamic load coefficient corresponding to each set of candidate parameters. The pre-trained guide vane adjustment model is trained based on sample crosswind speed, sample operating speed, sample guide vane configuration parameters and label aerodynamic load coefficient.

[0100] Step 240: Among the predicted aerodynamic load coefficients corresponding to each group of candidate parameters, the candidate guide vane configuration parameter corresponding to the smallest predicted aerodynamic load coefficient is determined as the target guide vane configuration parameter.

[0101] Step 250: Adjust the train's operating status according to the target guide vane configuration parameters.

[0102] Steps 210 to 250 are described in detail below.

[0103] In step 210, the crosswind speed of the operating environment in which the train is located during operation is obtained.

[0104] Crosswind speed refers to the component of environmental wind speed acting vertically or obliquely on the side of the train body. Crosswind speed is a major external environmental factor that induces lateral instability of the train and generates aerodynamic safety hazards. Crosswind speed can be collected in real time and transmitted to the train operation control system using fixed ultrasonic or propeller-type anemometers and wind vanes deployed along key sections of the railway line (such as viaducts and windy areas); alternatively, onboard wind speed sensors can be pre-installed on the train body, and the crosswind speed can be calculated by using an aerodynamic model in the train control and management system, combined with the train's speed, attitude, and acceleration data, to obtain the relative wind speed to the train; or, the crosswind speed of the operating environment during train operation can be determined by combining the crosswind speed measured at fixed points and the crosswind speed monitored in real time on the train body. Of course, the embodiments of this application do not limit the method of obtaining crosswind speed, and the specific method can be adjusted according to the actual situation.

[0105] The train has multiple deflectors installed at different heights on both sides, each of which can deploy at a certain angle within a predefined range to resist crosswinds.

[0106] In some embodiments, before obtaining the crosswind speed of the operating environment in which the train is operating, the method further includes:

[0107] (1.1) Obtain meteorological data of the crosswind area and the second operating speed of the train;

[0108] (1.2) Determine the predicted crosswind speed in the crosswind area based on meteorological data, and determine the configuration parameters of the train's second guide vane based on the predicted crosswind speed and the second operating speed;

[0109] (1.3) Adjust the train's operating status according to the configuration parameters of the second guide vane.

[0110] In some embodiments, the train body is equipped with multiple guide vanes of varying heights and a certain number of vanes. A guide vane is a rotatable, adjustable aerodynamic component installed on the leeward side of the head car and intermediate carriages (typically arranged symmetrically on both sides of the train body). When the train faces a crosswind environment, the train can determine the guide vane configuration parameters based on the real-time operating speed and crosswind speed. These parameters include the vane's activation height, the number of activated vanes, and the deployment angle (yaw angle) of each vane. The deployed guide vanes effectively interfere with and suppress the formation and development of turbulent vortices on the leeward side of the train, thereby reducing the air pressure difference between the two sides of the train and improving the lateral stability and driving safety of high-speed trains in complex crosswind environments.

[0111] In some embodiments, the train typically does not deploy the deflector when there is no crosswind. If the train deploys the deflector after entering the crosswind area, it will take some time for the deflector to expand from its initial closed state to the target state. When the train is already in the crosswind area, the delayed deflector adjustment will prevent the train from obtaining effective aerodynamic compensation in time when entering the wind zone. During this response delay period, strong turbulent vortices can easily cause the train body to sway violently laterally, which will seriously reduce the train's operational stability in crosswind environments. In extreme and sudden wind conditions, there is a serious safety hazard of derailment and overturning of the train.

[0112] Based on this, in this embodiment of the application, before obtaining the crosswind speed of the operating environment in which the train is operating, the configuration parameters of the guide vanes are determined in advance to make preliminary adjustments to the train's operating status:

[0113] In some embodiments, meteorological data of the crosswind area and the second operating speed of the train in motion are first acquired. The meteorological data refers to a set of environmental parameters collected in real time by meteorological monitoring stations (such as ultrasonic anemometers and wind vanes) deployed along the crosswind area. This data includes not only instantaneous wind speed and direction, but also gust coefficients, air density, temperature, and humidity. After filtering and spatiotemporal interpolation, this data is used to construct a high-precision local wind field model of the area to quantify the aerodynamic loads exerted on the train by the external environment. Aerodynamic loads refer to the distributed force system formed on the surface of the train body due to the aerodynamic effects generated by the relative motion between the train body and the surrounding air during operation. Macroscopically, these loads mainly manifest as drag, lift, lateral force acting on the train's center of gravity, as well as roll, pitch, and yaw moments that cause the train to overturn or yaw. The magnitude of these loads directly depends on air density, the square of the relative wind speed, the train's frontal area, and the aerodynamic shape characteristics of the train body. They are core physical quantities for evaluating train operational stability, structural strength, and anti-overturning safety. The second operating speed refers to the instantaneous actual speed of the train when it enters or is in a specific crosswind area, which is measured and fed back in real time by the on-board speed measurement system (such as Doppler radar, speed sensor or traction control system) that is pre-set on the train. The operating speed reflects the train's motion state relative to the ground at the current moment.

[0114] Furthermore, the device determines the predicted crosswind speed in the crosswind area based on meteorological data, and then determines the configuration parameters of the train's second deflector based on the predicted crosswind speed and the second operating speed. The predicted crosswind speed refers to a value obtained by reasonably predicting the environmental wind speed in the wind zone ahead based on current monitoring data. By combining the predicted wind speed with the train's own second operating speed, the desired configuration parameters for the second deflector are determined before the train enters the crosswind area ahead. It should be noted that the term "second" here is only used to distinguish it from the other operating speed and deflector configuration parameters described in this paper, and has no other meaning.

[0115] Furthermore, the train control memory can pre-store a mapping database containing "wind speed-vehicle speed-guide vane status" parameter mapping items. Given the predicted crosswind speed in the crosswind area and the train's second operating speed, the current optimal adjustment state of the train's guide vanes can be quickly determined by looking up the table. Typically, this adjustment is performed in the pre-crosswind area before the train enters the crosswind area to avoid premature adjustment of the guide vanes affecting the normal operation of the train.

[0116] In some embodiments, the on-board control unit sends control commands to the drive module (such as a high-precision servo motor) to rapidly and smoothly deploy the deflector on the leeward side of the vehicle body to the optimal posture specified by the second deflector configuration parameters before the train actually enters the crosswind area ahead. The second deflector configuration parameters include the deflector setting height, the number of deflectors deployed, and the deflector deployment angle.

[0117] For example, the train control module obtains the predicted crosswind speed through meteorological monitoring stations along the line and train-to-ground communication. and the current operating speed of the train Next, in the mapping database, by The corresponding optimal guide vane state is retrieved. Furthermore, if the actual operating condition is not on a discrete grid point, interpolation is performed using the optimal solutions from multiple neighboring operating points to obtain the second guide vane configuration parameters under continuous operating conditions. ,in, Indicates the height of the air deflector. Indicates the number of deflector wings deployed, This indicates the deployment angle of each deflector; next, the control module sends control commands to the drive module to rapidly deploy the corresponding number of deflectors at the appropriate height before the train enters the crosswind area. .

[0118] It should be noted that single or multiple guide vanes can be installed at different heights. The train can deploy only one height of guide vane or a combination of multiple heights of guide vane. If the latter, the configuration parameters of the second guide vane are as follows: Typically, guide vanes deployed at the same altitude have the same deployment angle.

[0119] Understandably, this embodiment of the application obtains meteorological data of the road ahead and the current train speed, and pre-calculates the optimal deflector configuration parameters to adjust the train's aerodynamic shape in advance. This effectively overcomes the problems of sudden aerodynamic load changes and control lag caused by relying on real-time feedback when the train encounters sudden crosswinds. Thus, this embodiment of the application can provide appropriate aerodynamic compensation the instant the train enters a strong wind zone, effectively suppressing the impact of turbulent vortices, greatly reducing the transient destructive force of crosswinds on the train body, and reducing the sudden load changes when the train enters the crosswind zone. This ensures a smooth transition and ultimate driving safety for high-speed trains in complex and variable wind environments.

[0120] In some embodiments, after obtaining the crosswind speed of the operating environment in which the train is operating, the method further includes:

[0121] (2.1) When the crosswind speed is greater than the second preset threshold and less than the first preset threshold, obtain the real-time aerodynamic load coefficient of the train and the configuration parameters of the first guide vane of the train;

[0122] (2.2) The updated guide vane configuration parameters are obtained by updating the configuration parameters of the first guide vane based on the real-time aerodynamic load coefficient;

[0123] (2.3) Adjust the train's operating status according to the updated guide vane configuration parameters.

[0124] The first and second preset thresholds together define the fluctuation range of crosswind speed. This range indicates that the measured wind speed data fluctuates within the allowable error range and has not reached the point where a strong correction mode requiring significant adjustment is needed (if it exceeds the first preset threshold, a strong correction mode is entered). At this point, the system enters a fine-tuning steady-state mode. In this mode, the device collects aerodynamic pressure difference data in real time through state monitoring modules such as pressure sensors installed on the vehicle surface. It then calculates the real-time aerodynamic load coefficient reflecting the current stress state of the train and updates the current configuration parameters of the first guide vane based on the real-time aerodynamic load coefficient, obtaining the fine-tuned updated guide vane configuration parameters.

[0125] The following sections describe how embodiments of this application update the configuration parameters of the first guide vane based on real-time aerodynamic load coefficients to obtain the updated guide vane configuration parameters:

[0126] (2.2.1) Obtain multiple historical aerodynamic load coefficients of the train within the target time period;

[0127] (2.2.2) Based on multiple historical aerodynamic load coefficients, determine the fluctuation value of the aerodynamic load coefficient of the train within the target time period;

[0128] (2.2.3) Determine the initial deflector deployment angle of the train based on the configuration parameters of the first deflector, and determine the real-time control energy efficiency coefficient based on the ratio between the real-time aerodynamic load coefficient and the initial deflector deployment angle;

[0129] (2.2.4) The deployment angle of the updated guide vanes of the train is determined based on the ratio between the aerodynamic load coefficient fluctuation value and the real-time control energy efficiency coefficient;

[0130] (2.2.5) Update the configuration parameters of the first guide vane according to the updated guide vane deployment angle to obtain the updated guide vane configuration parameters.

[0131] In some embodiments, the real-time aerodynamic load coefficient includes the lateral force coefficient and the roll moment coefficient. That is, the updated guide vane configuration parameters are obtained by updating the first guide vane configuration parameters based on the lateral force coefficient and the roll moment coefficient.

[0132] The target time period refers to a recent time window preceding the current moment of the train, such as a continuous sampling period within the past few seconds. Historical aerodynamic load coefficients include the lateral force coefficient and roll moment coefficient continuously recorded by the device within the target time period. Both the lateral force coefficient and roll moment coefficient are dimensionless aerodynamic parameters. The lateral force coefficient quantifies the magnitude of the lateral aerodynamic thrust experienced by the entire train body when traveling in crosswinds. A larger lateral force coefficient indicates a more pronounced tendency for the train to be pushed laterally off the track, and a higher risk of lateral instability and derailment. The roll moment coefficient characterizes the torsional moment caused by aerodynamic forces that cause the train to roll or overturn around its longitudinal axis under the combined influence of external crosswinds and complex airflow around the train body. A larger roll moment coefficient indicates a greater risk of the train tilting, swaying violently, or even overturning under the influence of factors such as turbulent vortices on the leeward side.

[0133] Among them, the aerodynamic load coefficient fluctuation value is a physical index used to quantify the amplitude of aerodynamic load change over time, characterizing the high-frequency oscillation characteristics of the aerodynamic forces on the train in the time dimension. For example... Figure 3 and Figure 4 As shown, Figure 3 This is a schematic diagram comparing the influence of the guide vane on the lateral force coefficient of the train according to the embodiments of this application. Figure 4 This is a comparative schematic diagram showing the effect of the guide vane on the train's roll moment coefficient, as provided in the embodiments of this application. Compared to trains without guide vanes, the train's guide vanes can significantly reduce the high-frequency fluctuation components of aerodynamic loads. Figure 3 and Figure 4 As can be seen, after 0.3 seconds of adjustment, both the lateral force coefficient and the roll moment coefficient have decreased significantly. Since the train is constantly moving, the lateral force coefficient and the roll moment coefficient are constantly fluctuating. The aerodynamic load coefficient fluctuation value refers to the lateral force coefficient fluctuation value or the roll moment coefficient fluctuation value.

[0134] Furthermore, the initial deflector deployment angle refers to the reference rotation angle of the train deflector before the start of the fine-tuning steady-state cycle. This angle information is included in the first deflector configuration parameters currently executed by the control module.

[0135] Furthermore, the deployment angle of the train's updated guide vanes is determined using the following formula:

[0136] .in, This represents the fluctuation value of the aerodynamic load coefficient. This is the real-time aerodynamic load factor. The initial guide vane deployment angle, To control the energy efficiency coefficient in real time, To update the deflector deployment angle. If Increasing θ will Get bigger, if Decreasing θ will If it becomes smaller A large absolute value indicates that the load will change significantly with a slight adjustment of the guide vane angle (adjustment should be done with more care).

[0137] Alternatively, the deployment angle of the train's updated guide vanes can be determined using the following formula:

[0138] .in, This represents the fluctuation value of the aerodynamic load coefficient. This is the real-time aerodynamic load factor. The initial guide vane deployment angle, To control the energy efficiency coefficient in real time. If Increasing θ will Get bigger, if Decreasing θ will become smaller; if A large absolute value indicates that the load will change significantly with a slight adjustment of the guide vane angle (adjustment should be done with more care).

[0139] Furthermore, the train control module replaces or supplements the original reference rotation angle with the newly calculated unfolding angle value containing minute angle changes, forming updated guide vane configuration parameters containing the latest physical state adjustment instructions. The updated guide vane configuration parameters are then sent to the underlying drive mechanism as target control variables, guiding the physical guide vane to complete precise angle correction in a very short time.

[0140] It is understood that the embodiments of this application eliminate the huge computational overhead of calling complex simulation or global optimization algorithms, and can quickly derive the fine-tuning angle required to achieve aerodynamic damping. In this way, the train guide vane can complete high-frequency, small-amplitude precise compensation actions in millisecond-level response time, effectively offsetting the high-frequency lateral and roll sway of the carriage caused by crosswinds.

[0141] Next, we will continue with step (2.3):

[0142] In some embodiments, the train's operating state is adjusted based on updated deflector configuration parameters. The train's onboard control module translates the updated deflector configuration parameters into specific mechanical action commands and sends them to the deflector drive module, controlling the leeward deflector to make continuous, minute adjustments with the updated deflector configuration parameters as the target. Thus, by continuously changing the deflector's deployment angle within a small range according to the high-frequency fine-tuning commands, the local flow field around the train is subjected to dynamically changing aerodynamic intervention, directly affecting the train's wind-driven operating state, counteracting the high-frequency lateral sway of the vehicle body caused by transient factors such as gusts, thereby improving train operating safety.

[0143] In step 220, when the crosswind speed is equal to or greater than the first preset threshold, multiple candidate guide vane configuration parameters are determined based on the crosswind speed and the first running speed of the train.

[0144] The first preset threshold is a pre-set safety wind speed critical value. When the ambient crosswind exceeds this critical value, it indicates that the aerodynamic lateral impact faced by the train has reached a level that may threaten the safety of train operation, thereby triggering the strong correction mode of the guide vane. The first operating speed is the current actual speed of the train. The candidate guide vane configuration parameters refer to the guide vane states that can be selected by the device. The candidate guide vane configuration parameters include the candidate installation height of the guide vane, the combination of candidate quantities, and the candidate rotatable deployment angle, etc.

[0145] In some embodiments, multiple candidate air deflector configuration parameters are determined based on the crosswind speed and the train's first operating speed, including:

[0146] (3.1) Multiple initial guide vane configuration parameters were determined based on the crosswind speed and the train's first operating speed;

[0147] (3.2) Determine the parameter level to which each initial guide vane configuration parameter belongs, and determine the degree of parameter crowding between each initial guide vane configuration parameter and other initial guide vane configuration parameters at the same parameter level;

[0148] (3.3) The initial guide vane configuration parameters are filtered according to the parameter level and parameter crowding degree to obtain the filtered initial guide vane configuration parameters;

[0149] (3.4) Obtaining genetic manipulation factors;

[0150] (3.5) Update the initial guide vane configuration parameters after screening according to the genetic operation factors to obtain the updated initial guide vane configuration parameters;

[0151] (3.6) Return to the step of determining the parameter level to which each initial guide vane configuration parameter belongs, until the preset iteration stop condition is reached, and obtain multiple candidate guide vane configuration parameters.

[0152] In some embodiments, after obtaining the crosswind speed and the first operating speed, the device randomly or according to certain rules generates an initial population of a certain size within the allowable design variable range. The design variable range includes, for example, the allowable range of guide vane setting height, the allowable range of guide vane deployment number, and the allowable range of guide vane deployment angle. Each initial guide vane configuration parameter represents a possible combination of guide vane setting angle, deployment number, and deployment angle, thus providing a basic solution space for subsequent iterative search of the optimal solution for the comprehensive aerodynamic load.

[0153] In this context, parameter level refers to the non-dominated frontier level of each configuration parameter after rapid non-dominated sorting in a multi-objective optimization evaluation system. A higher level indicates that the configuration scheme is more advantageous in reducing aerodynamic loads. Parameter crowding is used to measure the density of the distribution of each configuration parameter in the target solution space within the same parameter level (i.e., the same non-dominated front), so as to guide the solution set to be evenly distributed in subsequent screening, avoid the optimization results from falling into local clustering, and thus maintain the diversity of candidate schemes.

[0154] Furthermore, the device prioritizes selecting configuration parameters with higher parameter levels. When configuration parameters with the same parameter level compete for resources, the device tends to retain individuals with lower parameter crowding (i.e., sparser distribution and larger crowding distance). In this way, through a dual screening strategy of ranking by merit and diversity of distribution, the device eliminates inferior solutions and retains the high-performing and representative solution set to obtain the initial guide vane configuration parameters after screening, which serve as the parent samples for the next generation.

[0155] Furthermore, the mathematical rules and parameter sets for guiding heuristic algorithms on how to perform population propagation and mutation to explore new solution spaces are established through genetic operating factors. These genetic operating factors typically include selection operators, crossover operators, and mutation operators. Genetic operating factors are used to shuffle and recombine selected high-quality solutions to obtain more possible candidate solutions. Specifically, based on the genetic operating factors, parent samples are subjected to selection, crossover, and / or random mutation operations to obtain a batch of offspring individuals containing novel combinations of guide vane features. This allows for the determination of updated initial guide vane configuration parameters (including parent and newly generated offspring individuals), thereby greatly expanding the device's ability to find better aerodynamic performance schemes in unknown configuration spaces.

[0156] Furthermore, the device will perform non-dominated sorting and congestion distance calculations again for the new initial guide vane configuration parameters, and repeat this iterative calculation cycle. The preset iteration stopping condition can be that the number of iterations reaches a pre-set maximum limit, or that the multi-objective optimal solution (Pareto front) has converged and stabilized; the specific condition can be set according to the actual situation. When the preset iteration stopping condition is met, the algorithm terminates, and the final Pareto solution set output at this time consists of multiple candidate guide vane configuration parameters. These parameters represent the guide vane setting height, number of deployed guide vanes, and guide vane deployment angle that can be adopted under the current specific vehicle speed and wind speed. These multiple candidate guide vane configuration parameters are subsequently input into a pre-trained guide vane adjustment model to obtain the adjustment result corresponding to each candidate guide vane configuration parameter.

[0157] Understandably, by eliminating the inefficiency and limitations of manual trial and error or simple exhaustive search, the embodiments of this application can quickly and stably select a series of the most scientific and reasonable optimal guide vane adjustment schemes for trains under different crosswind and speed conditions, so as to maintain the diversity distribution of candidate schemes and effectively avoid local convergence of solutions.

[0158] In step 230, multiple sets of candidate parameters are generated based on the crosswind speed, the first operating speed, and multiple candidate guide vane configuration parameters. Each set of candidate parameters is input into the pre-trained guide vane adjustment model to obtain the predicted aerodynamic load coefficient corresponding to each set of candidate parameters. The pre-trained guide vane adjustment model is trained based on the sample crosswind speed, sample operating speed, sample guide vane configuration parameters, and label aerodynamic load coefficient.

[0159] In some embodiments, crosswind speed and the train's current first operating speed Each candidate guide vane configuration parameter generated according to steps (3.1) to (3.6) is respectively compared with the configuration parameters of each candidate guide vane. Perform one-to-one matching to generate multiple sets of candidate parameters. Among them, the crosswind speed in each group of candidate parameters and first running speed same.

[0160] Furthermore, each set of candidate parameters The input is fed into a pre-trained deflector adjustment model to obtain the predicted aerodynamic load coefficients corresponding to each set of candidate parameters. The predicted aerodynamic load coefficients refer to the quantitative indicators characterizing the strength of the train's exposure to wind pressure, which are rapidly derived and calculated by the pre-trained deflector adjustment model under the given candidate parameters. These indicators include the predicted train lateral force coefficient and roll moment coefficient.

[0161] In some embodiments, multiple sets of candidate parameters are generated based on crosswind speed, a first operating speed, and multiple candidate guide vane configuration parameters, including:

[0162] (4.1) Generate multiple sets of initial candidate parameters based on crosswind speed, first operating speed and multiple candidate guide vane configuration parameters;

[0163] (4.2) Determine the mean and standard deviation of the parameters based on multiple sets of initial candidate parameters;

[0164] (4.3) For each set of initial candidate parameters, determine the parameter difference based on the difference between the initial candidate parameter and the parameter mean;

[0165] (4.4) For each set of initial candidate parameters, the normalized candidate parameters are determined based on the ratio between the parameter difference and the parameter standard deviation, so as to obtain multiple sets of candidate parameters.

[0166] In some embodiments, since the magnitudes of the input parameters (crosswind speed, operating speed, guide vane setting height, number of guide vanes deployed, and guide vane deployment angle) differ too much, it is necessary to normalize the obtained initial candidate parameters.

[0167] First, the parameter mean and standard deviation are determined based on multiple initial candidate parameters. The parameter mean refers to the arithmetic mean of parameters in the same dimension within the sample set, reflecting the central tendency of the data distribution in that dimension. Specifically, the parameter mean includes the mean crosswind speed, mean operating speed, mean guide vane setup height, mean number of guide vanes deployed, and mean guide vane deployment angle. The parameter standard deviation is a statistical indicator measuring the dispersion of these parameters in the same dimension from the parameter mean. It is used to eliminate calculation biases caused by different dimensions. The parameter standard deviation includes the standard deviations of crosswind speed, operating speed, guide vane setup height, number of guide vanes deployed, and guide vane deployment angle.

[0168] Furthermore, for each set of initial candidate parameters, the corresponding candidate parameters are determined using the following formula:

[0169] .in, x k The first of the initial candidate parameters k One parameter; μ k This is the mean value of the parameter corresponding to that parameter. For example, if x k For the train's running speed, then μ k This represents the average running speed. s kThis is the standard deviation of the parameter. The difference between parameters; These are the normalized parameters.

[0170] Furthermore, by integrating the normalized results of each parameter in the initial candidate parameters, the corresponding candidate parameters are obtained, and thus multiple sets of candidate parameters are obtained.

[0171] Understandably, due to the significant differences in numerical magnitude and physical dimensions between physical parameters such as train speed, crosswind speed, deflector height, and deployment angle, directly inputting the raw data into the pre-trained deflector adjustment model would lead to severe weight skew and feature masking effects during model processing. This application's embodiment transforms multidimensional heterogeneous data into the same standard scale range by sequentially performing mean decentering and standard deviation scaling on multiple sets of initial candidate parameters, effectively eliminating the negative impact of magnitude differences. This significantly improves the accuracy of the pre-trained model in predicting aerodynamic load coefficients and the convergence speed of algorithm optimization, and further ensures the efficiency and stability of online simulation, providing more solid and reliable data support for the agile and safe control of high-speed trains in crosswind environments.

[0172] In some embodiments, before inputting each set of candidate parameters into the pre-trained airfoil adjustment model, the method further includes:

[0173] (5.1) Obtain parameters from multiple sets of samples;

[0174] (5.2) Input each set of sample parameters into the initial guide vane adjustment model to obtain the sample aerodynamic load coefficients corresponding to each set of sample parameters;

[0175] (5.3) Determine the total loss value based on the aerodynamic load coefficient of each sample and the aerodynamic load coefficient of the label corresponding to the sample aerodynamic load coefficient;

[0176] (5.4) The guide vane adjustment model is iteratively trained based on the total loss value until the model loss value is less than the preset loss threshold, and the trained guide vane adjustment model is obtained.

[0177] In some embodiments, to obtain a pre-trained deflector adjustment model, it is necessary to first acquire multiple sets of sample parameters to train the deflector adjustment model. Each set of sample parameters includes sample crosswind speed, sample running speed, and sample deflector configuration parameters. Then, each set of sample parameters... Inputting these parameters into the initial guide vane adjustment model yields the sample aerodynamic load coefficients corresponding to each set of sample parameters. ,in, The lateral force coefficient of the sample. The sample side rolling moment coefficient is defined similarly to the predicted aerodynamic load coefficient, and will not be repeated here. It should be noted that the embodiments of this application use the sample aerodynamic load coefficient, which includes the sample side force coefficient and the sample side rolling moment coefficient, for illustration. In actual application, only one of them can be selected or other coefficients that meet the definition can be added according to the specific situation.

[0178] The guide vane adjustment model can be a radial basis function (RBF) neural network or other neural networks. Taking the RBF neural network model as an example, for instance... Figure 5 As shown, Figure 5 This is a schematic diagram of the hidden layer data processing of the guide vane adjustment model provided in this application embodiment. The RBF hidden layer consists of M radial base nodes (10 in the example in the figure). As shown in the following formula, the j-th RBF node performs the following data processing on the input data:

[0179] Where M is the number of RBF nodes; j is the j-th RBF node; c j The center of the j-th node is σ; σ is the width of the RBF. This is the output of the j-th RBF node.

[0180] Furthermore, as shown in the following formula, the outputs of all RBF nodes are combined into a vector:

[0181] .in, This is the RBF output vector.

[0182] Furthermore, the output layer uses a linear combination to obtain two predicted values:

[0183] Where b is the bias term of the output layer; W is the weight matrix of the output layer.

[0184] Furthermore, the aerodynamic load coefficient corresponding to the aerodynamic load coefficient of each sample group is obtained, and the total loss value is calculated based on the aerodynamic load coefficient of each sample group and the corresponding label aerodynamic load coefficient. The label aerodynamic load coefficient refers to the actual lateral force and actual roll moment data directly obtained through high-fidelity but time-consuming computational fluid dynamics (CFD) numerical simulation under physical conditions where the corresponding sample parameters are exactly the same.

[0185] Furthermore, the device uses an optimizer to continuously fine-tune and update the connection weight matrix and bias terms between each neuron node in the network based on the gradient descent direction of the total loss value. Then, through continuous iterative training to correct the error, the model output continuously approaches the real physical boundary of the CFD simulation. This training process continues until the calculated total loss value decreases and stabilizes within a very small preset allowable error range, that is, less than the preset loss threshold. At this point, the guide vane adjustment model training is complete.

[0186] It is understood that the embodiments of this application introduce sample parameters covering a wide range of operating conditions and use high-precision CFD simulation results as labeled data for supervised learning to train a neural network model that can accurately characterize the complex aerodynamic characteristics of high-speed trains. The trained airfoil adjustment model avoids the bottlenecks of traditional CFD simulation in multivariate joint optimization, such as excessively high computational cost, long time consumption, and inability to be called online. It realizes the transformation from offline high-cost calculation to online fast query and optimization. This allows the train to accurately predict aerodynamic loads under any combination of vehicle speed and crosswind with minimal computing power and millisecond-level response time during actual inference, thereby quickly determining the most suitable airfoil configuration parameters.

[0187] In some embodiments, obtaining multiple sets of sample parameters includes:

[0188] (5.1.1) Obtain the range of crosswind speed, the range of train operating speed, the range of guide vane height, the range of guide vane deployment quantity, and the range of guide vane deployment angle;

[0189] (5.1.2) The range of crosswind speed, the range of train speed, the range of guide vane height, the range of guide vane deployment quantity, and the range of guide vane deployment angle are divided into intervals to obtain multiple parameter intervals;

[0190] (5.1.3) Sample each parameter interval to obtain the sample parameters corresponding to each parameter interval;

[0191] (5.1.4) Determine multiple sets of sample parameters based on the sample parameters corresponding to each parameter interval.

[0192] In some embodiments, the ranges of crosswind speed, train operating speed, deflector height, number of deflectors deployed, and deflector deployment angle are obtained:

[0193] (1) Crosswind speed range: Defines the crosswind speed The range of values ​​can be determined based on meteorological statistics along the route. For example, the range of crosswind speed can be defined as 10 m / s to 40 m / s.

[0194] (2) Train operating speed range: defines the train operating speed. The range of values ​​is determined based on the permissible speed and safety margin of the line. For example, the operating speed range of the train can be defined as 160km / h to 350km / h.

[0195] (3) Guide vane height range: Defines the range of values ​​for the guide vane installation height H, set as a continuous height variable with the rail surface as the reference, for example, in H min To H max Internal changes;

[0196] (4) Range of the number of guide vanes deployed: The range of values ​​for the number of guide vanes deployed N is defined. Considering discrete working conditions such as single-wing, double-wing and multi-wing cooperation, integer codes (such as N=1, 2, 3...) are used to represent the number of guide vanes or their combination forms;

[0197] (5) Deployment angle range of the guide vane: defines the deployment angle of the guide vane. The range of values, for example, defining the airfoil deployment angle in a continuous range from 0° to 90°, to capture the optimal deployment angle.

[0198] Furthermore, a five-dimensional parameter space is constructed based on the ranges of crosswind speed, train operating speed, guide vane height, number of guide vanes deployed, and guide vane deployment angle. Then, the optimal Latin hypercube sampling (OLHS) method is used to select sample parameters within this five-dimensional parameter space. Specifically:

[0199] (1) Divide each variable into n parameter intervals (e.g., n=40~60) within its corresponding value range, and randomly select a sampling position in each interval;

[0200] (2) By using the Latin hypercube principle, we can ensure that each variable appears only once in each interval in all samples, thereby achieving uniform coverage in all dimensions;

[0201] (3) For continuous variables, the initial Latin hypercube samples are optimized using the "maximum-minimum distance" or "center distance" criteria to obtain the OLHS sampling scheme, so that the distance between different sample points in the high-dimensional space is as large as possible, avoiding sample clustering and improving the ability to capture nonlinear changes.

[0202] (4) For discrete variables, the nearest integer rounding strategy is adopted, that is, sampling is first performed in the [0,1] normalized space, and then mapped to the corresponding integer number of pieces to ensure that the samples are evenly distributed among different combinations of the number of guide vanes.

[0203] (5) Multiple sets of sample parameters were obtained using the OLHS method described above. , .

[0204] Understandably, this application's embodiments, by accurately acquiring a multidimensional feature range covering environmental wind conditions and physical actuator states, and based on this, introduce spatially uniform interval partitioning and optimal Latin hypercube sampling methods, effectively avoid the defects of traditional blind random sampling, such as high-dimensional clustering of samples, uneven coverage distribution, or generation of meaningless operating conditions. This design can ensure that each variable achieves maximum uniform coverage across all samples while controlling the sampling scale, thereby accurately capturing the high-order nonlinear coupling relationship between "vehicle speed – wind speed – airfoil configuration," and outputting a high-quality dataset with extremely high generalization representativeness for training the aerodynamic adjustment surrogate model.

[0205] In some embodiments, the total loss value is determined based on the aerodynamic load coefficient of each sample and the tag aerodynamic load coefficient corresponding to the sample aerodynamic load coefficient, including:

[0206] (5.3.1) Determine the sample lateral force coefficient and sample side rolling moment coefficient corresponding to the aerodynamic load coefficient of each sample;

[0207] (5.3.2) Determine the first loss value based on the lateral force coefficient of each sample and the aerodynamic load coefficient of the first tag corresponding to the lateral force coefficient of each sample;

[0208] (5.3.3) Determine the second loss value based on the side rolling moment coefficient of each sample and the second label aerodynamic load coefficient corresponding to the side rolling moment coefficient of each sample;

[0209] (5.3.4) Determine the total loss value based on the first loss value and the second loss value.

[0210] In some embodiments, the sample lateral force coefficient and sample roll moment coefficient corresponding to each sample aerodynamic load coefficient are then used to determine the total loss value through weighted mean square error (MSE), specifically expressed as the following formula:

[0211] Where N is the number of training samples; α and β These are predefined weighting coefficients; This represents the lateral force coefficient of the i-th sample. This represents the aerodynamic load coefficient of the first label corresponding to the lateral force coefficient of the i-th sample. This represents the first loss value; This represents the side rolling moment coefficient of the i-th sample. This represents the aerodynamic load coefficient of the second label corresponding to the side rolling moment coefficient of the i-th sample. This represents the second loss value.

[0212] It is understood that the embodiments of this application accurately decompose the macroscopic aerodynamic load prediction results into two independent dimensions: the lateral force coefficient and the roll moment coefficient. These are then compared with the corresponding high-precision label true values, and finally fused into a total loss value through weight allocation. This effectively overcomes the problem of weight skew or local overfitting that may occur in the prediction of complex physical fields by a single error index. This allows the guide vane adjustment model to balance the prediction accuracy of both lateral force and roll moment during iterative training. It also ensures that the overall optimization direction of the model is more in line with the real aerodynamic operation law of trains, thereby greatly improving the high-fidelity prediction capability and reliability of the final pre-trained surrogate model under complex crosswind conditions.

[0213] In step 240, among the predicted aerodynamic load coefficients corresponding to each group of candidate parameters, the candidate guide vane configuration parameter corresponding to the smallest predicted aerodynamic load coefficient is determined as the target guide vane configuration parameter.

[0214] In some embodiments, the candidate guide vane configuration parameter corresponding to the smallest predicted aerodynamic load coefficient is selected from multiple predicted aerodynamic load coefficients and determined as the target guide vane configuration parameter. In complex aerodynamic scenarios, the smallest predicted aerodynamic load coefficient means that the lateral thrust and overturning moment experienced by the train under this set of guide vane configurations are reduced to the minimum, the overall aerodynamic performance is optimized, and the train's wind resistance attitude is most stable.

[0215] In step 250, the train's operating status is adjusted according to the target guide vane configuration parameters.

[0216] In some embodiments, the device drives the adjustable guide vane to rotate or deploy rapidly to the target guide vane configuration parameters, which will deploy the corresponding number and angle of guide vanes at the corresponding installation height, so as to actively and physically intervene in the local aerodynamic shape of the train in operation, thereby effectively suppressing the evolution and shedding of strong turbulent vortices on the leeward side, and changing the wind-driven operation state of the train from the flow field level.

[0217] For example, the multiple sets of candidate parameters cover eight guide vane deployment heights: not deployed, position A, position B, position C, position D, position E, position F, and position G. The deployment angle gradually increases from position A to position G. The multiple sets of candidate parameters cover deployment angles from 0° to 90°. This application embodiment uses a single guide vane deployed to only one height as an example to illustrate the effect. Figure 6 and Figure 7 As shown, Figure 6 This is a schematic diagram comparing the effects of different guide vane heights on the train's lateral force coefficient, provided in an embodiment of this application. Figure 7 This is a schematic diagram comparing the effects of different guide vane heights on the train's roll moment coefficient, provided in the embodiments of this application. When the guide vane at position D is deployed, the train's lateral force coefficient and roll moment coefficient are minimized.

[0218] Furthermore, such as Figure 8 As shown, Figure 8 This is a schematic diagram comparing the effects of different guide vane deployment heights on the aerodynamic loads of a train, provided in an embodiment of this application. In the original operating condition without guide vanes, a rapidly developing local main vortex V1 and a longer main vortex V2 extending beyond the train's leeward side are generated. When the guide vane is positioned at a higher location (A), its impact on the structure and velocity of the main vortex V2 is relatively small. However, when it is positioned at lower locations (F and G), the main vortex V2 moves smoothly after being generated at the train's front, and the influence of the guide vane on it is also weakened. As the guide vane installation height is gradually adjusted to locations C, D, and E, a new main vortex V3 is induced in the flow field. This main vortex V3 can directly cut into and influence the development path of the main vortex V2, causing significant energy dissipation in V2. In particular, when the guide vane is located at position D, the structure and velocity of the main vortex V2 are reduced to the maximum extent. This intuitively explains from the aerodynamic flow field mechanism why the negative pressure on the leeward side surface of the train is significantly reduced at this specific height, thereby minimizing the lateral force and aerodynamic overturning risk experienced by the train. Among them, the main vortex, as a large-scale ordered coherent structure in the flow field, carries most of the fluid's kinetic energy. The main vortex transfers energy step by step to smaller turbulent vortices until it is dissipated as heat energy.

[0219] Furthermore, such as Figure 9 and Figure 10 As shown, Figure 9 This is a schematic diagram comparing the influence of the same guide vane deployment height on the train's lateral force coefficient, provided in an embodiment of this application. Figure 10 This is a comparative schematic diagram showing the effect of the same guide vane deployment height on the train roll moment coefficient, provided in an embodiment of this application. The guide vane at deployment position D significantly reduces the lateral force coefficient and roll moment coefficient compared to the case where the guide vane is not deployed. The original operating condition refers to the case where the guide vane is not deployed. To facilitate a better understanding of this application, another complete embodiment is shown below:

[0220] (1) Hardware deployment: Adjustable air deflectors are installed on the leeward side of the head car and intermediate cars (symmetrically arranged on both sides of the car body). The drive module adopts a high-precision servo motor, and the status monitoring module includes an ultrasonic anemometer and wind direction instrument installed at the head of the car body and an array of pressure sensors on the surface of the car body.

[0221] (2) Dynamic control: For example, if a train is traveling at 350 km / h on a railway line and is about to pass through a mountain pass with strong winds, then the following steps (A) to (D) will be executed:

[0222] (A) Feedforward prediction: When the distance to the wind zone is 2km, the vehicle system receives a warning signal from a meteorological station along the route, predicting that the crosswind speed ahead is 25m / s. The control unit immediately queries the predefined database and determines that the optimal deflector deployment angle under this condition is 42°.

[0223] (B) Pre-action: 5 seconds before the train enters the wind zone, the drive module is activated to smoothly deploy the leeward guide vane to 42° and complete the preparatory posture.

[0224] (C) Feedback Correction: When the train enters the wind zone, the crosswind speed measured by the front sensor is 28 m / s (slightly higher than the predicted value). The control unit determines that the deviation exceeds the threshold and immediately activates the strong correction mode. Based on the approximate model, it calculates in real time and instructs the guide vane angle to be corrected to 46°.

[0225] (D) Steady-state fine-tuning: When the train is traveling in a windy area, the wind speed is stable at around 28 m / s. The system enters the fine-tuning steady-state mode, and the guide vanes are finely adjusted at a high frequency of ±1.5° on a 46° reference to counteract the high-frequency lateral sway of the train body caused by gusts.

[0226] In this example, compared to trains without guide vanes, the train's lateral force coefficient is reduced by approximately 17.85%, and the lateral roll moment coefficient is reduced by approximately 22.60%, significantly improving the safety margin of the train.

[0227] It is understood that the embodiments of this application, by acquiring vehicle speed and wind speed data in real time and using a pre-trained guide vane adjustment model for multi-parameter joint deduction and rapid optimization, can accurately lock the optimal dynamic configuration scheme of the guide vanes within millisecond-level latency. This not only enables timely suppression of aerodynamic loads on high-speed trains under crosswind conditions, but also significantly improves the lateral stability and driving safety of the train.

[0228] like Figure 11 As shown, Figure 11 This is a schematic diagram of the module structure of the train operation control device provided in this application embodiment. The train operation control device 300 may include the following modules 310 to 350:

[0229] The acquisition module 310 is used to acquire the crosswind speed of the operating environment in which the train is located during operation;

[0230] The candidate guide vane configuration parameter determination module 320 is used to determine multiple candidate guide vane configuration parameters based on the crosswind speed and the first running speed of the train when the crosswind speed is equal to or greater than the first preset threshold.

[0231] The predictive aerodynamic load coefficient determination module 330 is used to generate multiple sets of candidate parameters based on crosswind speed, first operating speed and multiple candidate guide vane configuration parameters, and input each set of candidate parameters into the pre-trained guide vane adjustment model to obtain the predicted aerodynamic load coefficient corresponding to each set of candidate parameters. The pre-trained guide vane adjustment model is trained based on sample crosswind speed, sample operating speed, sample guide vane configuration parameters and labeled aerodynamic load coefficient.

[0232] The target guide vane configuration parameter selection module 340 is used to determine the candidate guide vane configuration parameter corresponding to the smallest predicted aerodynamic load coefficient among the predicted aerodynamic load coefficients corresponding to each group of candidate parameters as the target guide vane configuration parameter.

[0233] The train operation status adjustment module 350 is used to adjust the train operation status according to the target guide vane configuration parameters.

[0234] In some embodiments, the acquisition module 310 is used for:

[0235] When the crosswind speed is greater than the second preset threshold and less than the first preset threshold, the real-time aerodynamic load coefficient of the train and the configuration parameters of the train's first guide vane are obtained.

[0236] The updated guide vane configuration parameters are obtained by updating the first guide vane configuration parameters based on the real-time aerodynamic load coefficient;

[0237] The train's operating status is adjusted based on the updated guide vane configuration parameters.

[0238] In some embodiments, the acquisition module 310 is further configured to:

[0239] Obtain multiple historical aerodynamic load coefficients of the train within the target time period;

[0240] Based on multiple historical aerodynamic load coefficients, the fluctuation value of the aerodynamic load coefficient of the train within the target time period is determined;

[0241] The initial deflector deployment angle of the train is determined based on the configuration parameters of the first deflector, and the real-time control energy efficiency coefficient is determined based on the ratio between the real-time aerodynamic load coefficient and the initial deflector deployment angle.

[0242] The deployment angle of the train's updated guide vanes is determined based on the ratio between the aerodynamic load coefficient fluctuation value and the real-time control energy efficiency coefficient.

[0243] The configuration parameters of the first guide vane are updated based on the updated guide vane deployment angle to obtain the updated guide vane configuration parameters.

[0244] In some embodiments, the acquisition module 310 is further configured to:

[0245] Acquire meteorological data of the crosswind area and the train's second operating speed;

[0246] The predicted crosswind speed in the crosswind area is determined based on meteorological data, and the configuration parameters of the train's second guide vane are determined based on the predicted crosswind speed and the second operating speed.

[0247] The train's operating status is adjusted according to the configuration parameters of the second guide vane.

[0248] In some embodiments, the candidate guide vane configuration parameter determination module 320 is used for:

[0249] Multiple initial guide vane configuration parameters were determined based on the crosswind speed and the train's initial operating speed.

[0250] Determine the parameter level to which each initial deflector configuration parameter belongs, and determine the degree of parameter crowding between each initial deflector configuration parameter and other initial deflector configuration parameters at the same parameter level;

[0251] Multiple initial guide vane configuration parameters are filtered based on parameter level and parameter crowding level to obtain the filtered initial guide vane configuration parameters;

[0252] Obtaining genetic manipulation factors;

[0253] The initial guide vane configuration parameters after screening are updated based on the genetic operation factors to obtain the updated initial guide vane configuration parameters;

[0254] Return to the steps of determining the parameter level to which each initial guide vane configuration parameter belongs, until a preset iteration stop condition is reached, resulting in multiple candidate guide vane configuration parameters.

[0255] In some embodiments, the predictive aerodynamic load coefficient determination module 330 is used for:

[0256] Multiple sets of initial candidate parameters are generated based on crosswind speed, first operating speed, and multiple candidate guide vane configuration parameters;

[0257] The mean and standard deviation of the parameters are determined based on multiple sets of initial candidate parameters;

[0258] For each set of initial candidate parameters, the parameter difference is determined based on the difference between the initial candidate parameter and the parameter mean.

[0259] For each set of initial candidate parameters, the normalized candidate parameters are determined based on the ratio between the parameter difference and the parameter standard deviation, thus obtaining multiple sets of candidate parameters.

[0260] In some embodiments, the predictive aerodynamic load coefficient determination module 330 is further configured to:

[0261] Obtain parameters from multiple sets of samples;

[0262] Each set of sample parameters is input into the initial guide vane adjustment model to obtain the sample aerodynamic load coefficient corresponding to each set of sample parameters;

[0263] The total loss value is determined based on the aerodynamic load coefficient of each sample and the aerodynamic load coefficient of the label corresponding to the sample aerodynamic load coefficient;

[0264] The guide vane adjustment model is iteratively trained based on the total loss value until the model loss value is less than the preset loss threshold, thus obtaining the trained guide vane adjustment model.

[0265] In some embodiments, the predictive aerodynamic load coefficient determination module 330 is further configured to:

[0266] Obtain the range of crosswind speed, train operating speed, guide vane height, number of guide vanes deployed, and guide vane deployment angle;

[0267] The ranges of crosswind speed, train speed, guide vane height, number of guide vanes deployed, and guide vane deployment angle were divided into intervals to obtain multiple parameter intervals.

[0268] Sample each parameter interval to obtain the sample parameters corresponding to each parameter interval;

[0269] Multiple sets of sample parameters are determined based on the sample parameters corresponding to each parameter range.

[0270] In some embodiments, the predictive aerodynamic load coefficient determination module 330 is further configured to:

[0271] Determine the sample lateral force coefficient and sample lateral rolling moment coefficient corresponding to the aerodynamic load coefficient of each sample;

[0272] The first loss value is determined based on the lateral force coefficient of each sample and the aerodynamic load coefficient of the first tag corresponding to the lateral force coefficient of each sample;

[0273] The second loss value is determined based on the side rolling moment coefficient of each sample and the aerodynamic load coefficient of the second label corresponding to the side rolling moment coefficient of each sample;

[0274] The total loss value is determined based on the first loss value and the second loss value.

[0275] The train operation control method, device, electronic equipment, and storage medium proposed in this application acquire the crosswind speed of the operating environment in which the train is located during operation; when the crosswind speed is equal to or greater than a first preset threshold, multiple candidate guide vane configuration parameters are determined based on the crosswind speed and the first operating speed of the train; multiple sets of candidate parameters are generated based on the crosswind speed, the first operating speed, and the multiple candidate guide vane configuration parameters, and each set of candidate parameters is input into a pre-trained guide vane adjustment model to obtain the predicted aerodynamic load coefficient corresponding to each set of candidate parameters. The pre-trained guide vane adjustment model is trained based on sample crosswind speed, sample operating speed, sample guide vane configuration parameters, and labeled aerodynamic load coefficients; among the predicted aerodynamic load coefficients corresponding to each set of candidate parameters, the candidate guide vane configuration parameter corresponding to the smallest predicted aerodynamic load coefficient is determined as the target guide vane configuration parameter; and the train's operating state is adjusted according to the target guide vane configuration parameter.

[0276] This application embodiment, by real-time monitoring of train operation status and introducing a multivariable pre-trained guide vane adjustment model, rapidly predicts aerodynamic load coefficients under multiple candidate guide vane configuration parameters. This avoids the high computational cost problem of traditional computational fluid dynamics, which involves excessively long solutions, and realizes the transformation from offline high-cost calculation to online rapid query and optimization. Furthermore, by selecting the target guide vane configuration parameters corresponding to the minimum aerodynamic load coefficient to dynamically adjust the train operation status, it can accurately and effectively suppress the formation and development of leeward turbulent vortices, reduce the pressure difference between the two sides of the train, and improve train operation safety. The pre-trained guide vane adjustment model in this application embodiment can quickly determine the target guide vane configuration parameters, thereby enabling timely train operation control in crosswind environments. Moreover, the adjusted train operation status not only significantly reduces the lateral force and roll moment experienced during travel but also significantly improves the train's operational stability and environmental adaptability in crosswind environments, effectively avoiding potential hazards such as vehicle body swaying, derailment, or overturning under extreme wind conditions. Thus, this application embodiment achieves real-time and efficient protection of high-speed train operation safety.

[0277] like Figure 12 As shown, Figure 12 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. The electronic device includes:

[0278] The processor 401 can be implemented using a general-purpose central processing unit (CPU), microprocessor, application specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application.

[0279] The memory 402 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 402 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 402 and is called and executed by the processor 401 using the image analysis method of the embodiments of this application.

[0280] Input / output interface 403 is used to implement information input and output;

[0281] The communication interface 404 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0282] Bus 405 transmits information between various components of the device (e.g., processor 401, memory 402, input / output interface 403, and communication interface 404);

[0283] The processor 401, memory 402, input / output interface 403 and communication interface 404 are connected to each other within the device via bus 405.

[0284] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described image analysis method.

[0285] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0286] The train operation control method, device, electronic equipment, and storage medium proposed in this application acquire the crosswind speed of the operating environment in which the train is located during operation; when the crosswind speed is equal to or greater than a first preset threshold, multiple candidate guide vane configuration parameters are determined based on the crosswind speed and the first operating speed of the train; multiple sets of candidate parameters are generated based on the crosswind speed, the first operating speed, and the multiple candidate guide vane configuration parameters, and each set of candidate parameters is input into a pre-trained guide vane adjustment model to obtain the predicted aerodynamic load coefficient corresponding to each set of candidate parameters. The pre-trained guide vane adjustment model is trained based on sample crosswind speed, sample operating speed, sample guide vane configuration parameters, and labeled aerodynamic load coefficients; among the predicted aerodynamic load coefficients corresponding to each set of candidate parameters, the candidate guide vane configuration parameter corresponding to the smallest predicted aerodynamic load coefficient is determined as the target guide vane configuration parameter; and the train's operating state is adjusted according to the target guide vane configuration parameter.

[0287] This application embodiment, by real-time monitoring of train operation status and introducing a multivariable pre-trained guide vane adjustment model, rapidly predicts aerodynamic load coefficients under multiple candidate guide vane configuration parameters. This avoids the high computational cost problem of traditional computational fluid dynamics, which involves excessively long solutions, and realizes the transformation from offline high-cost calculation to online rapid query and optimization. Furthermore, by selecting the target guide vane configuration parameters corresponding to the minimum aerodynamic load coefficient to dynamically adjust the train operation status, it can accurately and effectively suppress the formation and development of leeward turbulent vortices, reduce the pressure difference between the two sides of the train, and improve train operation safety. The pre-trained guide vane adjustment model in this application embodiment can quickly determine the target guide vane configuration parameters, thereby enabling timely train operation control in crosswind environments. Moreover, the adjusted train operation status not only significantly reduces the lateral force and roll moment experienced during travel but also significantly improves the train's operational stability and environmental adaptability in crosswind environments, effectively avoiding potential hazards such as vehicle body swaying, derailment, or overturning under extreme wind conditions. Thus, this application embodiment achieves real-time and efficient protection of high-speed train operation safety.

[0288] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0289] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0290] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0291] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0292] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0293] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0294] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0295] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0296] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0297] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0298] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A train operation control method, characterized in that, include: Obtain the crosswind speed of the train's operating environment during operation; When the crosswind speed is greater than the second preset threshold and less than the first preset threshold, the real-time aerodynamic load coefficient of the train and the configuration parameters of the first guide vane of the train are obtained. The updated guide vane configuration parameters are obtained by updating the first guide vane configuration parameters based on the real-time aerodynamic load coefficient; The train's operating status is adjusted according to the updated guide vane configuration parameters; When the crosswind speed is equal to or greater than a first preset threshold, multiple candidate guide vane configuration parameters are determined based on the crosswind speed and the first running speed of the train. Multiple sets of candidate parameters are generated based on the crosswind speed, the first operating speed, and multiple candidate guide vane configuration parameters. Each set of candidate parameters is input into a pre-trained guide vane adjustment model to obtain the predicted aerodynamic load coefficient corresponding to each set of candidate parameters. The pre-trained guide vane adjustment model is trained based on sample crosswind speed, sample operating speed, sample guide vane configuration parameters, and labeled aerodynamic load coefficient. In each group of candidate parameters, the candidate airfoil configuration parameter corresponding to the smallest predicted aerodynamic load coefficient is determined as the target airfoil configuration parameter. The train's operating status is adjusted according to the target air deflector configuration parameters; The process of determining multiple candidate guide vane configuration parameters based on the crosswind speed and the train's first operating speed includes: Multiple initial guide vane configuration parameters were determined based on the crosswind speed and the first operating speed of the train. Determine the parameter level to which each of the initial deflector configuration parameters belongs, and determine the degree of parameter crowding between each of the initial deflector configuration parameters and other initial deflector configuration parameters at the same parameter level; Multiple initial guide vane configuration parameters are filtered based on the parameter level and the parameter crowding degree to obtain the filtered initial guide vane configuration parameters; Obtaining genetic manipulation factors; The initial guide vane configuration parameters after screening are updated according to the genetic operation factors to obtain the updated initial guide vane configuration parameters; Return to the step of determining the parameter level to which each of the initial guide vane configuration parameters belongs, until a preset iteration stop condition is reached, and obtain multiple candidate guide vane configuration parameters.

2. The train operation control method according to claim 1, characterized in that, The step of updating the first guide vane configuration parameters based on the real-time aerodynamic load coefficient to obtain the updated guide vane configuration parameters includes: Obtain multiple historical aerodynamic load coefficients of the train within the target time period; Based on multiple historical aerodynamic load coefficients, the fluctuation value of the aerodynamic load coefficient of the train during the target time period is determined; The initial deflector deployment angle of the train is determined based on the first deflector configuration parameters, and the real-time control energy efficiency coefficient is determined based on the ratio between the real-time aerodynamic load coefficient and the initial deflector deployment angle. The updated guide vane deployment angle of the train is determined based on the ratio between the aerodynamic load coefficient fluctuation value and the real-time control energy efficiency coefficient. The configuration parameters of the first guide vane are updated based on the updated guide vane deployment angle to obtain the updated guide vane configuration parameters.

3. The train operation control method according to claim 1, characterized in that, Before obtaining the crosswind speed of the train's operating environment during operation, the method further includes: Acquire meteorological data of the crosswind area and the train's second operating speed; The predicted crosswind speed in the crosswind area is determined based on the meteorological data, and the configuration parameters of the second guide vane of the train are determined based on the predicted crosswind speed and the second operating speed. The train's operating status is adjusted according to the configuration parameters of the second air deflector.

4. The train operation control method according to claim 1, characterized in that, The generation of multiple sets of candidate parameters based on the crosswind speed, the first operating speed, and multiple candidate guide vane configuration parameters includes: Multiple sets of initial candidate parameters are generated based on the crosswind speed, the first operating speed, and multiple candidate guide vane configuration parameters; The mean and standard deviation of the parameters are determined based on the multiple sets of initial candidate parameters. For each set of initial candidate parameters, the parameter difference is determined based on the difference between the initial candidate parameter and the mean of the parameters; For each set of initial candidate parameters, the normalized candidate parameters are determined based on the ratio between the parameter difference and the parameter standard deviation, so as to obtain multiple sets of candidate parameters.

5. The train operation control method according to claim 1, characterized in that, Before inputting each set of candidate parameters into the pre-trained airfoil adjustment model, the method further includes: Obtain parameters from multiple sets of samples; Each set of sample parameters is input into the initial guide vane adjustment model to obtain the sample aerodynamic load coefficient corresponding to each set of sample parameters; The total loss value is determined based on the aerodynamic load coefficient of each sample and the aerodynamic load coefficient of the tag corresponding to the aerodynamic load coefficient of the sample; The guide vane adjustment model is iteratively trained based on the total loss value until the total loss value is less than a preset loss threshold, thus obtaining the trained guide vane adjustment model.

6. The train operation control method according to claim 5, characterized in that, The acquisition of multiple sets of sample parameters includes: Obtain the range of crosswind speed, train operating speed, guide vane height, number of guide vanes deployed, and guide vane deployment angle; The crosswind speed range, the train operating speed range, the guide vane height range, the guide vane deployment number range, and the guide vane deployment angle range are divided into intervals to obtain multiple parameter intervals; Sample each parameter interval to obtain the sample parameters corresponding to each parameter interval; Multiple sets of sample parameters are determined based on the sample parameters corresponding to each parameter range.

7. The train operation control method according to claim 5, characterized in that, The step of determining the total loss value based on the aerodynamic load coefficient of each sample and the corresponding tag aerodynamic load coefficient includes: Determine the sample lateral force coefficient and sample lateral rolling moment coefficient corresponding to each sample aerodynamic load coefficient; The first loss value is determined based on the lateral force coefficient of each sample and the first tag aerodynamic load coefficient corresponding to the lateral force coefficient of each sample; The second loss value is determined based on the side rolling moment coefficient of each sample and the second tag aerodynamic load coefficient corresponding to each side rolling moment coefficient of the sample; The total loss value is determined based on the first loss value and the second loss value.

8. A train operation control device, characterized in that, include: The acquisition module is used to acquire the crosswind speed of the train's operating environment during operation; When the crosswind speed is greater than the second preset threshold and less than the first preset threshold, the real-time aerodynamic load coefficient of the train and the configuration parameters of the first guide vane of the train are obtained. The updated guide vane configuration parameters are obtained by updating the first guide vane configuration parameters based on the real-time aerodynamic load coefficient; The train's operating status is adjusted according to the updated guide vane configuration parameters; The candidate guide vane configuration parameter determination module is used to determine multiple initial guide vane configuration parameters based on the crosswind speed and the first running speed of the train when the crosswind speed is equal to or greater than a first preset threshold. Determine the parameter level to which each of the initial deflector configuration parameters belongs, and determine the degree of parameter crowding between each of the initial deflector configuration parameters and other initial deflector configuration parameters at the same parameter level; Multiple initial guide vane configuration parameters are filtered based on the parameter level and the parameter crowding degree to obtain the filtered initial guide vane configuration parameters; the genetic operation factor is then obtained. The initial guide vane configuration parameters after screening are updated according to the genetic operation factors to obtain the updated initial guide vane configuration parameters; Return to the step of determining the parameter level to which each of the initial guide vane configuration parameters belongs, until a preset iteration stop condition is reached, and obtain multiple candidate guide vane configuration parameters; The predictive aerodynamic load coefficient determination module is used to generate multiple sets of candidate parameters based on the crosswind speed, the first operating speed and multiple candidate guide vane configuration parameters, and input each set of candidate parameters into a pre-trained guide vane adjustment model to obtain the predicted aerodynamic load coefficient corresponding to each set of candidate parameters. The pre-trained guide vane adjustment model is trained based on sample crosswind speed, sample operating speed, sample guide vane configuration parameters and label aerodynamic load coefficient. The target airfoil configuration parameter selection module is used to determine the candidate airfoil configuration parameter corresponding to the smallest predicted aerodynamic load coefficient among the predicted aerodynamic load coefficients corresponding to each group of candidate parameters as the target airfoil configuration parameter. The train operation status adjustment module is used to adjust the train operation status according to the target guide vane configuration parameters.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the train operation control method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the train operation control method according to any one of claims 1 to 7.

Images