Railway vehicle sanding control system and method

Abstract

The present application belongs to the field of vehicle sanding control technology, and relates to a rail vehicle sanding control system and a control method thereof. The present application dynamically determines whether to output a sanding signal and an ice surface distribution characteristic label based on clustering analysis of the front track section ice surface image and the track cross slope angle; dynamically sets the sanding spray angle by fusing the vehicle speed, the track cross slope angle and the ice surface characteristics, sets the target sanding flow according to the air source pressure and determines the sanding duration; monitors the actual sanding flow and the target flow deviation in real time, and triggers the compensation action when the deviation exceeds the limit; monitors the nozzle pressure difference change rate during and after the sanding execution, and performs the sanding operation when the change rate deviates from the normal trend. The present application effectively solves the problems of inaccurate control, low sanding efficiency and nozzle blockage of the existing sanding system on the icy track, and achieves the effects of improving the sanding accuracy and automation level, enhancing the friction between the wheels and the track, ensuring the driving safety and reducing the maintenance cost.

Description

Technical Field

[0001] This invention belongs to the field of vehicle sand spreading and cleaning control technology, and relates to a control system and control method for sand spreading and cleaning of rail vehicles. Background Technology

[0002] During the operation of rail vehicles, the sand spreading system is a key device to improve the friction coefficient between the wheels and the rails and ensure braking safety. Especially in low temperature and freezing weather, the ice layer formed on the rail surface will significantly reduce the adhesion coefficient between the wheels and the rails, causing the wheels to slip, spin, or even slide when braking, which seriously threatens the safety of train operation and control accuracy.

[0003] Given that the timing, location, and amount of sand applied directly affect its adhesion-enhancing effect and driving safety, precise control of the sand application process is necessary. In existing technologies, the control methods for sand spreaders are typically adjusted based on the condition of the track surface and vehicle operating parameters. For example, Chinese invention patent publication number CN120135227A discloses a sand spreader and its control method, which determines whether there is water accumulation or freezing on the track surface by acquiring track surface image information and vehicle operating speed, and adjusts the blowing speed of the sand spreader according to the determination result. This method recognizes that the amount of sand applied needs to be increased to improve adhesion under freezing conditions and attempts to control the sand flow rate by adjusting the air blowing head parameters.

[0004] However, the existing technologies mentioned above still exhibit crude and outdated control strategies when dealing with the core safety issue of wheel slippage on ice. Specifically, this manifests in the following ways: (1) The sand-spreading decision does not fully consider the dynamic characteristics of the track. The judgment is based solely on planar images and vehicle speed, without considering key track parameters such as the track cross slope angle. On ice surfaces such as slopes or curves, the stress state of the wheels is more complex. The sand-spreading decision of existing methods is disconnected from the track direction and cannot adjust the timing and location of sand spreading in advance according to the track curvature and slope. This results in a severe lag in sand spreading when entering ice curves or slopes, or the sand particles failing to cover the actual contact area between the wheel and rail, thus greatly reducing the effectiveness of sand spreading.

[0005] (2) However, at the execution level, because its sand spreading flow control is in open-loop mode, it lacks a real-time monitoring and compensation mechanism. Fluctuations in air source pressure or pipeline losses may cause the actual sand spreading amount to deviate from the set value. When braking on an ice surface where sand particles are urgently needed to increase adhesion, insufficient sand spreading amount will directly increase the risk of wheel slippage.

[0006] Furthermore, to ensure the reliability of the sand-spreading system, regular sand cleaning is essential. This aims to remove accumulated impurities or clumps of sand from the nozzles and keep the pipeline unobstructed. Existing sand cleaning technologies primarily address the common issue of sand particles sticking together due to moisture, often employing periodic or reactive methods, lacking a real-time health monitoring and proactive sand cleaning mechanism based on the nozzle's operating status. Once the nozzle becomes partially clogged due to impurities or freezing, the sand-spreading capacity drops sharply, potentially causing anti-slip protection to fail on icy surfaces. Summary of the Invention

[0007] In view of this, in order to solve the problems mentioned in the background technology, a sand spreading and cleaning control system for rail vehicles and its control method are proposed.

[0008] The first aspect of the present invention proposes a sand spreading and clearing control system for rail vehicles, comprising the following modules: a state prediction module: based on real-time acquired ice surface images and track cross slope angles of the preceding track section, performing cluster analysis on the ice surface images, and combining the cluster analysis results with the track cross slope angles to determine whether to output a sand spreading signal and ice surface distribution feature labels for the ice area.

[0009] Sand spreading decision module: When a sand spreading signal is received, the module dynamically sets the spray angle of the sand spreading nozzle based on the current vehicle speed and ice surface distribution feature tags, integrates the track cross slope angle, sets the target sand spreading flow rate according to the air source pressure, and determines the sand spreading duration and sand spreading start time.

[0010] Execution verification module: Sand spreading is performed at the spraying angle at the start of sand spreading. During the sand spreading process, the actual sand spreading flow rate is compared with the target sand spreading flow rate in real time. If the deviation exceeds the tolerance range dynamically determined based on the flow rate calibration error, a compensation action is triggered.

[0011] Sand cleaning management module: During and after sand spreading, the nozzle pressure difference change rate is monitored in real time. If the change rate deviates from the normal trend, a sand cleaning operation is performed.

[0012] The second aspect of the present invention proposes a method for controlling sand spreading and clearing on rail vehicles, comprising the following steps: based on real-time acquired images of the ice surface of the track section ahead and the track cross slope angle, performing cluster analysis on the ice surface images, and combining the cluster analysis results with the track cross slope angle to determine whether to output a sand spreading signal and ice surface distribution feature labels for the ice area.

[0013] When a sand-spreading signal is received, the spraying angle of the sand-spreading nozzles is dynamically set based on the current vehicle speed and ice surface distribution feature tags, and the track cross slope angle is integrated.

[0014] Set the target sand spraying flow rate based on the air source pressure, and determine the sand spraying duration and sand spraying start time.

[0015] Sand spreading is performed at the spraying angle at the start of sand spreading. During the sand spreading process, the actual sand spreading flow rate is compared with the target sand spreading flow rate in real time. If the deviation exceeds the tolerance range dynamically determined based on the flow rate calibration error, a compensation action is triggered.

[0016] During and after sand spreading, the nozzle pressure differential change rate is monitored in real time. If the change rate deviates from the normal trend, a sand cleaning operation is performed.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention uses a sand-spreading decision mechanism that combines track cross slope angle and ice surface image clustering analysis. When determining whether to spread sand, the system not only identifies the existence of ice areas, but also combines the track cross slope and the coverage characteristics of the ice surface in the wheel-rail contact area to achieve intelligent identification of slippage risk under complex line conditions. This solves the problem that the timing of sand spreading and the wheel-rail contact area do not match due to the neglect of track geometry in the prior art, and effectively improves the pertinence and timeliness of sand spreading in sections such as curves and slopes.

[0018] (2) This invention collects the actual flow rate in real time during the sand spreading process and dynamically sets the target flow rate and tolerance range based on the air source pressure. Once the deviation exceeds the limit, the opening of the air source valve is dynamically adjusted to implement flow compensation. This solves the problem of insufficient or excessive actual sand spreading caused by the lack of closed-loop feedback in traditional control. It ensures that the sand supply is stable and reliable during the critical braking stage, thereby effectively suppressing wheel slippage and improving braking response accuracy and operational safety.

[0019] (3) This invention constructs a real-time monitoring mechanism based on the nozzle pressure difference change rate and compares the similarity of the pressure difference trend under historical normal working conditions. Once an abnormality is identified, an active sand cleaning program of airflow pulse and negative pressure suction is triggered. This solves the problem that traditional sand cleaning methods rely on post-processing or only target the wet state of the sand and cannot cope with the risk of nozzle blockage. It ensures that the sand spreading system can continue to operate reliably under harsh working conditions such as freezing and avoids the failure of anti-slip function due to nozzle performance deterioration. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram showing the connection of each module in a sand spreading and cleaning control system for rail vehicles according to the present invention.

[0022] Figure 2 This is a flowchart of the state prediction module in this invention.

[0023] Figure 3 This is a flowchart illustrating the steps of a sand-spreading and sand-cleaning control method for rail vehicles according to the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Example 1

[0026] Please see Figure 1 As shown, the present invention provides a sand spreading and cleaning control system for rail vehicles, comprising: a state prediction module, a sand spreading decision module, an execution verification module, and a sand cleaning management module. The connection relationship between the modules is as follows: the state prediction module and the sand spreading decision module are connected, the sand spreading decision module and the execution verification module are connected, and the execution verification module and the sand cleaning management module are connected.

[0027] State prediction module: Based on the real-time acquired ice surface image and track cross slope angle of the track section ahead, cluster analysis is performed on the ice surface image. The cluster analysis results are combined with the track cross slope angle to determine whether to output a sand-spreading signal and ice surface distribution feature labels of the ice area.

[0028] In one embodiment of the present invention, considering that track icing and track cross slope angle are two core factors affecting wheel-rail adhesion and vehicle lateral stability: track surface icing forms a lubricating film, significantly reducing the longitudinal adhesion coefficient between the wheel and rail, easily causing wheel spin or skidding; while the track cross slope angle, under icing conditions, introduces a gravitational component that causes vehicle sideslip, coupling with the already reduced adhesion, drastically exacerbating the risk of lateral instability. Therefore, cluster analysis is used to accurately segment the icy area and the background, and the cross slope angle parameter is integrated to comprehensively determine the necessity of sand application, thereby achieving targeted sand application that combines anti-skid and anti-side-skid measures, moving beyond simple anti-skid measures.

[0029] Preferably, in one embodiment of the present invention, the state prediction module further includes defining the forward track segment: based on the vehicle braking performance, a basic look-ahead distance is pre-calibrated, and the vehicle's current position is moved forward by the basic look-ahead distance along the track travel direction as the starting position of the forward track segment.

[0030] The aforementioned basic look-ahead distance is based on the maximum safe operating speed of this type of rail vehicle under low-adhesion ice conditions before system deployment. Maximum deceleration and system response delay time The basic look-ahead distance is calculated using the following formula: ,in The distance the vehicle travels during the system response delay time. To improve speed With maximum deceleration The distance required to brake to a stop.

[0031] Based on the starting position, a predetermined fixed length is extended along the direction of track movement to form the track section ahead. The fixed length is determined according to the maximum effective recognition range of the on-board sensor.

[0032] Further, please refer to Figure 2 As shown, the specific content of the state prediction module is as follows: by performing cluster analysis on the ice surface image in the track section ahead, the ice area and the background area are segmented.

[0033] Specifically, the steps include: first, preprocessing the acquired ice surface image, including grayscale conversion, histogram equalization, and Gaussian filtering for noise reduction; based on the brightness difference between the ice surface and the track background, using image segmentation algorithms such as threshold segmentation or clustering segmentation to segment the image into bright areas and background areas.

[0034] The segmented highlighted areas were identified as ice areas. Finally, morphological closing operations were used to process the identified binary ice area image to eliminate small holes and isolated noise points, resulting in a continuous and complete ice area contour for subsequent analysis.

[0035] Based on the segmentation results, the starting position, longitudinal continuous length, and lateral coverage width of the ice zone are calculated and used as the feature labels of the ice surface distribution.

[0036] Specifically, the steps include: taking the starting point of the preceding track section as the origin of the coordinate system, establishing a longitudinal X-axis along the track extension direction, and establishing a transverse Y-axis perpendicular to the track direction.

[0037] Transform the pixel coordinates of the segmented binary ice area image to this orbital coordinate system to obtain the actual coordinate positions of each pixel in the ice area.

[0038] In the orbital coordinate system, search for the coordinates of the first pixel to appear in the ice zone along the positive X-axis. The vertical coordinate value of this point is determined as the starting point of the ice zone. When multiple discrete ice zones exist, the starting point closest to the vehicle is taken as the valid starting point.

[0039] Starting from a defined starting position, the continuous ice coverage is statistically analyzed along the positive X-axis. The difference in vertical distance between the farthest point of consecutively occurring ice pixels and the starting point is taken as the vertical continuity length of the ice zone. If there are discontinuities, the length of the longest consecutive segment is taken as this feature value.

[0040] Within the longitudinal range of the ice zone, multiple sampling sections are selected at fixed intervals along the X-axis. For each section, the maximum coverage span of the ice zone pixels in the Y-axis direction is calculated. The average of the maximum spans across all sections is taken as the lateral coverage width of the ice zone.

[0041] It should be understood that the starting position of the ice zone indicates the specific starting position of sand spreading, the longitudinal continuous length indicates the continuous coverage scale of the ice zone along the track direction, reflecting the duration of longitudinal risk, and the lateral coverage width indicates the coverage range of the ice zone on the track cross section, indicating the distribution of risk in the lateral direction.

[0042] The sand-spreading signal and ice surface distribution feature label are output when the following conditions are met simultaneously; otherwise, no sand-spreading signal is output.

[0043] (a) The longitudinal continuous length of the ice zone shall not be less than the contact length of a single axle wheelset of the train.

[0044] (b) The absolute value of the track cross slope angle is greater than the set threshold, and the lateral coverage width of the ice zone on the uphill side in the direction of vehicle travel is not less than the safety critical ratio of the standard track width.

[0045] The method for obtaining the threshold value of the track cross slope angle is as follows: based on the vehicle's operational safety data on the cross slope track, a minimum cross slope angle value is determined as the threshold value through actual vehicle testing or simulation analysis.

[0046] It's important to explain that the contact between the train wheelset and the track is spatially continuous. If the length of the ice zone is less than the contact length of a single axle wheelset, it means that at any given moment, at least part of the wheelset remains in contact with the ice-free track, the adhesion coefficient is not completely lost, and the vehicle still maintains a certain degree of controllability. Conversely, there will be a moment when the entire wheelset is completely on the ice surface, causing a sharp drop in the adhesion coefficient, thus significantly increasing the risk of wheel spin or skidding.

[0047] Therefore, condition (a) ensures that the identified ice zone is safe in the direction of train travel, excluding sporadic ice layers that are too short-lived to affect the traction / braking stability of the train.

[0048] Condition (b) first filters out track sections that have significant tilt and thus generate a large lateral gravity component by ensuring that the absolute value of the track cross slope angle is greater than a set threshold. The effect of a small cross slope angle is negligible.

[0049] The safety threshold ratio, where the lateral coverage width of the ice zone on the uphill side in the vehicle's direction of travel is no less than the standard track width, is set because when a vehicle traverses a curve with a cross slope, its center of gravity shifts towards the outer side of the track (downhill side), while the inner side (uphill side) provides crucial support to prevent rollover and skidding. A safety threshold ratio of, for example, 50%, means that when more than half of the effective contact width of the uphill rail is covered by ice, the lateral support it provides is compromised, and the risk level escalates from caution to requiring immediate action.

[0050] The simultaneous fulfillment of both conditions reflects a shift from single anti-slip to coordinated control of anti-slip and anti-side slip.

[0051] Sand spreading decision module: When a sand spreading signal is received, the module dynamically sets the spraying angle of the sand spreading nozzle based on the current vehicle speed and ice surface distribution feature tags, and sets the target sand spreading flow rate based on the track cross slope angle, thereby determining the sand spreading duration.

[0052] In one embodiment of the present invention, considering the influence of different vehicle speeds, ice zone morphology and track conditions on the sand spreading effect, the spraying angle and flow rate of the sand spreading nozzle are dynamically adjusted by integrating multi-dimensional parameters to achieve optimal coverage of the sand spreading material on the ice surface.

[0053] Specifically, this step first obtains the current vehicle speed, the longitudinal continuous length and lateral coverage width of the ice zone, and the track cross slope angle; then, these parameters are fused together, and an angle compensation value is generated through a pre-calibrated compensation mapping relationship of the system; the pre-set basic spraying angle value is superimposed with the angle compensation value to obtain the final spraying angle.

[0054] The current vehicle speed, the longitudinal continuous length and lateral coverage width of the ice zone, and the track cross slope angle are all normalized to eliminate the influence of dimensions. This is existing technology and will not be elaborated further.

[0055] The method for constructing the compensation mapping relationship is as follows: A sample dataset containing the optimal spraying angles corresponding to different vehicle speeds, longitudinal lengths of the ice zone, lateral widths of the ice zone, and track cross slope angles is constructed in an experimental platform or real-vehicle test. The optimal spraying angle for each working condition is determined by evaluating the actual coverage ratio of the sand on the ice zone under different spraying angles and selecting the angle with the largest coverage ratio. Based on this dataset, using the four parameters as input and the optimal spraying angle as output, a data fitting method such as multiple regression is used to establish a mapping function, thereby obtaining the compensation mapping relationship.

[0056] It should be understood that the obtained spraying angle takes into account the dynamic effects of vehicle speed and ice zone characteristics. Due to the lag in the landing point of the sand caused by inertia when driving at high speed, the angle automatically includes a forward compensation component to ensure that the sand can accurately cover the corresponding point of the ice zone. Compensation is made according to the lateral coverage width of the ice zone so that the sand spraying range matches the actual distribution of the ice zone.

[0057] When the track is at a cross slope angle, the spraying angle is dynamically adjusted to point towards the uphill side in the direction of vehicle travel, allowing sand to cover the working surface of the rail on the uphill side. This actively restores the lateral adhesion coefficient between the wheel and rail on that side, providing the vehicle with the lateral support force necessary to resist sideslip.

[0058] Furthermore, the method for obtaining the sand-spreading duration includes: firstly, obtaining the benchmark sand-spreading flow rate corresponding to the calibrated pressure. Measure the real-time gas source pressure and calculate the ratio r between the real-time gas source pressure and the calibrated pressure.

[0059] Next, the target sand-spreading flow rate is calculated. In an ideal situation, the fluid flow rate and pressure have a linear relationship. However, in actual gas-solid two-phase flow systems, the relationship between flow rate and pressure is usually nonlinear. Therefore, a flow-pressure exponent n is introduced to fit the power-law characteristic of flow rate changing with pressure in the actual system. This value can be obtained by measuring the actual sand-spreading flow rate under different gas source pressures and performing nonlinear regression fitting based on these pressure-flow data pairs using a power function. The value of n depends on the inherent flow resistance characteristics of the system. The smaller the resistance, the closer the value of n is to 1; the larger the resistance, the closer the value of n is to 2.

[0060] The formula for calculating the target sand spreading flow rate is: .

[0061] Then, the theoretical sand-spreading duration is calculated without considering the effect of pressure. The longitudinal continuous length of the ice zone is divided by the current vehicle speed, and the quotient is taken as the theoretical sand-spreading duration. This time represents the ideal time required to fully cover the ice zone at the baseline flow rate.

[0062] It's important to understand that if r < 1, it indicates insufficient gas source pressure, because... Therefore, since the target sand-spreading flow rate is less than the baseline sand-spreading flow rate, it means that the amount of sand spread onto the track per unit time has decreased. Therefore, a compensation strategy of extending the sand-spreading time is adopted. The initial sand-spreading duration is obtained by applying pressure compensation to the theoretical sand-spreading duration. The formula is: By extending the sand spreading time, the sand volume shortfall caused by reduced flow rate is compensated for throughout the sand spreading process when the pressure is insufficient.

[0063] If r=1, it means that the working condition is ideal and no pressure compensation is required.

[0064] If r>1, it indicates that the air source pressure is excessive, causing the target sand spreading flow rate to be greater than the benchmark sand spreading flow rate. In order to ensure that the total amount of sand required to fully cover the ice area is accurately controllable, the formula is used to shorten the initial sand spreading duration to compensate for the increase in flow rate caused by the increase in pressure, thereby avoiding sand waste.

[0065] Considering the decrease in adhesion coefficient caused by wheel-rail contact geometry changes when a train passes through a curved track, the amount of sand applied in the curved area is dynamically increased to provide anti-slip protection. Specifically, the radius of curvature of each point on the track section ahead is obtained based on the track topographic map. The smaller the radius of curvature, the greater the curvature of the track at that point.

[0066] The track section is divided into multiple continuous unit segments. For each unit segment, the quotient of the unit segment length and the corresponding radius of curvature is calculated. This value represents the comprehensive measure of the curvature of the track segment relative to the straight line and its length. The quotients of all unit segments are added together to obtain the cumulative curvature value k, which is a dimensionless number representing the total curvature of the entire track in the ice zone.

[0067] A larger k value indicates more and sharper curves ahead. In such curve sections, the wheelset needs to generate lateral creep to provide guidance force when the vehicle passes through, which consumes some of the longitudinal adhesion capacity available for traction and braking; at the same time, the wheel-rail contact point shifts towards the wheel flange root, resulting in increased contact stress. All of these factors together lead to a decrease in wheelset adhesion capacity.

[0068] Therefore, the sand-spreading duration needs to be obtained by curvature compensation of the initial sand-spreading duration. , where α is the curvature compensation coefficient, αk constitutes the compensation term, the value of which represents the percentage of the additional sand to be spread relative to the base sand spread amount, and 1+αk indicates that the final total sand spread amount should be a multiple of the base sand spread amount. Multiplying this multiple by the initial sand spread duration converts the total compensation relationship into a time compensation relationship, thus obtaining the sand spread duration after curvature compensation.

[0069] When the trajectory is an ideal straight line, k=0, then no compensation is needed.

[0070] When there is a curve in the track, k>0, the duration of sand application needs to be extended proportionally by introducing a curvature compensation term αk to compensate for the decrease in adhesion coefficient caused by the curve.

[0071] The method for determining the curvature compensation coefficient α includes the following steps: In a real vehicle test, the vehicle's operating data in the curve section is collected, including vehicle speed and traction / braking system output parameters.

[0072] The longitudinal force between the wheel and rail is calculated based on the output parameters, and the actual adhesion coefficient is estimated in real time using the principles of vehicle dynamics. The estimation is based on dynamic parameters including vehicle mass, running acceleration and rotational inertia, and is achieved through simplified dynamic relationships such as a single-mass multi-rigid-body model.

[0073] Synchronously monitor wheel angular acceleration and vehicle linear acceleration to identify the critical state of wheel-rail relationship when it is in a state of adhesion utilization. The critical state is defined as the instantaneous state in which the wheel is about to spin or slide but has not yet started to continuously spin or slide.

[0074] Using the critical state as the optimization target, multiple sets of repeated tests are conducted by adjusting the α value to obtain the optimal α value sequence under different radii of curvature.

[0075] Regression analysis is performed on the optimal α value sequence to determine the curvature compensation coefficient α applicable to the system.

[0076] Furthermore, to ensure that the sand spreading action can completely cover the entire ice area, the sand spreading needs to be started in advance before the vehicle arrives at the starting point of the ice area. Therefore, the method for obtaining the sand spreading start time is as follows: obtain the first distance between the starting point of the ice area and the current position of the vehicle, and take the distance corresponding to the duration of the sand spreading when the vehicle travels at the current speed as the second distance.

[0077] To ensure that the end of the sand-covered section formed by the sand-spreading operation is precisely aligned with the end of the ice zone, the trigger point of the sand-spreading action must be located before the start of the ice zone. Therefore, the difference between the first distance and the second distance is used as the lead distance between the sand-spreading trigger point and the start of the ice zone. This distance is the vehicle travel distance corresponding to the duration of the sand-spreading operation itself.

[0078] Divide this lead distance by the current vehicle speed to obtain the time lead, thereby determining the time when the system needs to act in advance. The sum of the vehicle's current time and the time lead is taken as the sand spreading start time.

[0079] It should be noted that by adding a lead time to the current moment, the system can compensate for the delay throughout the entire process, from the issuance of control commands to the action of the actuators and the spraying of sand onto the track. This ensures that the sand is precisely covering the track surface the instant the vehicle wheels actually arrive at the starting point of the icy area, thus solving the problem of sand spraying lag.

[0080] If the sand-spreading signal fails before the sand-spreading start time, the sand-spreading operation will be cancelled.

[0081] Execution verification module: Sand spreading is performed at the spraying angle at the start of sand spreading. During the sand spreading process, the actual sand spreading flow rate is compared with the target sand spreading flow rate in real time. If the deviation exceeds the tolerance range dynamically determined based on the flow rate calibration error, a compensation action is triggered.

[0082] In one embodiment of the present invention, considering that the sand spreading system may be affected by factors such as air source pressure fluctuations during operation, real-time flow monitoring and dynamic compensation control are used to maintain the stability of the sand spreading flow and ensure the consistency of the sand spreading effect.

[0083] Specifically, this step includes: during the sand spreading process, the actual sand spreading flow rate in the sand spreading pipeline is collected in real time through a flow sensor to obtain the inherent fluctuation range ΔQ of the benchmark sand spreading flow rate under the calibrated pressure. This range characterizes the inherent accuracy of the system under ideal operating conditions. Combined with the ratio r of the current air source pressure to the calibrated pressure, the inherent fluctuation range is scaled to dynamically determine the theoretically permissible fluctuation range of the target sand spreading flow rate. .

[0084] Calculate the absolute value deviation between the actual sand spreading flow rate and the target sand spreading flow rate, compare the absolute value deviation with the theoretical allowable fluctuation range, and determine that the flow rate deviation exceeds the limit when the absolute value deviation is greater than the theoretical allowable fluctuation range.

[0085] When the flow rate deviation is determined to be excessive, the opening of the air source control valve is adjusted according to the direction and magnitude of the deviation. If the actual sand spreading flow rate is less than the target sand spreading flow rate, the valve opening is increased to enhance the air source power and thus increase the sand spreading flow rate; if the actual sand spreading flow rate is greater than the target sand spreading flow rate, the valve opening is decreased to reduce the air source power and thus reduce the sand spreading flow rate.

[0086] After adjusting the valve opening, continue monitoring the actual sand-spreading flow rate until the absolute value deviation within N consecutive sampling periods (e.g., N=3) is less than or equal to the theoretically permissible fluctuation range. Only then is it determined that the flow rate has stabilized and the compensation action is completed. If the number of consecutive compensations exceeds the limit (e.g., 10), the flow rate compensation is determined to have failed, triggering a system alarm and stopping the sand-spreading operation to avoid an infinite loop.

[0087] Sand cleaning management module: During and after sand spreading, the nozzle pressure difference change rate is monitored in real time. If the change rate deviates from the normal trend, a sand cleaning operation is performed.

[0088] In one embodiment of the present invention, considering that the sand-spreading material may accumulate at the nozzle and cause blockage, early warning and automatic clearing of nozzle blockage can be achieved by monitoring the differential pressure change rate and trend analysis.

[0089] Specifically, the steps include: after the sand spreading begins, continuously collecting the pressure difference between the inlet and outlet of the sand spreading nozzle at a fixed sampling period; performing differential calculation on the pressure difference for M consecutive sampling periods, for example, M=10, that is, calculating the difference between the pressure difference at the next moment and the pressure difference at the previous moment, to obtain a nozzle pressure difference change sequence composed of M-1 differences.

[0090] The nozzle pressure difference change sequence of the current sand spreading process is compared with the pressure difference change sequence recorded in the database for historical normal sand spreading processes under the same or similar target sand spreading flow conditions. The comparison is achieved by calculating the similarity between the two sequences. Specifically, the pressure difference change sequence is regarded as a feature vector and a vector similarity algorithm such as cosine similarity or Euclidean distance is used for calculation.

[0091] When one of the following conditions is met, the nozzle differential pressure change rate is determined to deviate from the normal trend: (i) the similarity calculation result is less than the preset similarity threshold. For example, the threshold is set to 0.8, which indicates that the current differential pressure change pattern is different from the pattern in the healthy state. The similarity threshold can be determined by calculating the similarity distribution between pairs of historical normal differential pressure change sequences and selecting the low quantile of the distribution.

[0092] (ii) The differential pressure change sequence shows a monotonically increasing trend and the current differential pressure value exceeds the maximum differential pressure value of the historical normal sand spreading process by a multiple. This indicates that the nozzle flow capacity is continuously deteriorating and has exceeded the normal fluctuation range. For example, it can be set to 1.5 times. This value can be determined by statistically analyzing the extreme values ​​of differential pressure during the historical normal sand spreading process and, based on its statistical characteristics, such as taking the average value plus a certain number of standard deviations, or directly taking the high percentile.

[0093] When the nozzle differential pressure change rate is determined to deviate from the normal trend, the current sand spreading operation is immediately interrupted, the sand spreading material supply is turned off, the airflow pulse and negative pressure suction are started, and the nozzle differential pressure change rate is re-detected after the sand is cleaned.

[0094] If the nozzle differential pressure change rate is restored to the normal trend after retesting, the original sand spreading operation will continue. If it still does not return to the normal trend, the sand cleaning time will be extended and the sand cleaning procedure will be repeated. If the sand cleaning procedure is not restored after a maximum of three times, a nozzle blockage alarm signal will be issued to prompt maintenance personnel to perform manual intervention and maintenance.

[0095] Example 2

[0096] See Figure 3 As shown, this invention proposes a sand spreading and clearing control system and its control method for rail vehicles, including the following steps: based on the real-time acquired ice surface image and track cross slope angle of the track section ahead, perform cluster analysis on the ice surface image, and combine the cluster analysis results with the track cross slope angle to determine whether to output a sand spreading signal and the ice surface distribution feature label of the ice area.

[0097] When a sand-spreading signal is received, the spraying angle of the sand-spreading nozzles is dynamically set based on the current vehicle speed and ice surface distribution feature tags, and the track cross slope angle is integrated.

[0098] Set the target sand spraying flow rate based on the air source pressure, and determine the sand spraying duration and sand spraying start time.

[0099] Sand spreading is performed at the spraying angle at the start of sand spreading. During the sand spreading process, the actual sand spreading flow rate is compared with the target sand spreading flow rate in real time. If the deviation exceeds the tolerance range dynamically determined based on the flow rate calibration error, a compensation action is triggered.

[0100] During and after sand spreading, the nozzle pressure differential change rate is monitored in real time. If the change rate deviates from the normal trend, a sand cleaning operation is performed.

[0101] In summary, this invention first uses a state prediction module to intelligently identify ice zones based on ice surface images and track cross slope angles, and then decides whether to issue a sand-spreading signal. Next, a sand-spreading decision module dynamically sets the spraying angle based on vehicle speed, ice surface characteristics, and cross slope angle, and sets the target sand-spreading flow rate based on air source pressure, while accurately calculating the sand-spreading duration and start time. Then, a verification module compares the actual and target flow rates in real time during sand-spreading, triggering closed-loop compensation if the deviation exceeds the limit. Finally, a sand-cleaning management module monitors the nozzle pressure difference trend throughout the process, automatically performing graded sand-cleaning operations when a blockage risk is detected, thereby achieving control and maintenance of the entire sand-spreading process.

[0102] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0103] Those skilled in the art will recognize that the algorithmic steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0104] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0105] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0106] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A rail vehicle sanding and blowing control system, characterized by, include: The state prediction module performs cluster analysis on the ice surface image and track cross slope angle based on the real-time acquired ice surface image of the track section ahead. It then combines the cluster analysis results with the track cross slope angle to determine whether to output a sand-spreading signal and ice surface distribution feature labels for the ice area. The sand-spreading decision module, upon receiving a sand-spreading signal, dynamically sets the spraying angle of the sand-spreading nozzles based on the current vehicle speed and ice surface distribution feature tags, integrates the track cross slope angle, sets the target sand-spreading flow rate according to the air source pressure, and determines the sand-spreading duration and sand-spreading start time. The execution verification module performs sand spreading at the spraying angle at the start of sand spreading. During the sand spreading process, the actual sand spreading flow rate is compared with the target sand spreading flow rate in real time. If the deviation exceeds the tolerance range dynamically determined based on the flow rate calibration error, a compensation action is triggered. The sand cleaning management module monitors the nozzle pressure difference change rate in real time during and after sand spreading. If the change rate deviates from the normal trend, a sand cleaning operation is performed. The contents of the sand removal management module are as follows: After the sand spreading operation begins, the pressure difference between the inlet and outlet of the sand spreading nozzle is continuously collected at a fixed sampling period. The pressure difference values ​​of M consecutive sampling periods are differentially calculated to obtain the nozzle pressure difference change sequence; The similarity between the nozzle differential pressure change sequence of the current sand spreading process and the differential pressure change sequence of the historical normal sand spreading process under the same sand spreading flow conditions is calculated. The nozzle differential pressure change rate is considered to deviate from the normal trend when one of the following conditions is met: (i) The similarity calculation result is less than the preset similarity threshold; (ii) The differential pressure change sequence shows a monotonically increasing trend and the current differential pressure value exceeds the maximum differential pressure value of the historical normal sand spreading process by a multiple; The state prediction module also includes defining the forward track section: pre-calibrating the basic look-ahead distance based on the vehicle's braking performance; The vehicle's current position is moved forward by the basic forward distance along the track's direction of travel, which serves as the starting point of the track segment ahead. Based on the starting point, a predetermined fixed length is extended along the track to form the track segment ahead. The specific content of the state prediction module is as follows: by performing cluster analysis on the ice surface image in the track section ahead, the ice area and the background area are segmented; based on the segmentation results, the starting position, longitudinal continuous length and lateral coverage width of the ice area are calculated and used as ice surface distribution feature labels; when the following conditions are met at the same time, the sand-spreading signal and the ice surface distribution feature label are output, otherwise no sand-spreading signal is output: (a) the longitudinal continuous length of the ice area is not less than the contact length of the single axle wheelset of the train; (b) the absolute value of the track cross slope angle is greater than the set threshold, and the lateral coverage width of the ice area on the uphill side of the vehicle travel direction is not less than the safety critical ratio of the standard track width; The spraying angle of the sand-spraying nozzle is set as follows: by integrating the current vehicle speed, track cross slope angle, and the longitudinal continuous length and lateral coverage width of the ice surface distribution feature label, an angle compensation value is generated through the compensation mapping relationship calibrated by the system; the basic spraying angle value and the angle compensation value are superimposed to obtain the final spraying angle.

2. The sand spreading and cleaning control system for rail vehicles as described in claim 1, characterized in that, The method for obtaining the sand-spreading duration is as follows: Obtain the benchmark sand-spreading flow rate corresponding to the calibrated pressure, measure the real-time air source pressure, and calculate the ratio of the real-time air source pressure to the calibrated pressure; Based on the benchmark sand spreading flow rate and the ratio, combined with the preset flow pressure index, the target sand spreading flow rate is calculated. Divide the longitudinal continuous length of the ice zone by the current vehicle speed, and use the quotient as the theoretical sand-spreading duration. The theoretical sand-spreading duration is corrected based on the ratio of the baseline sand-spreading flow rate to the target sand-spreading flow rate to obtain the initial sand-spreading duration; Based on the topographic map of the track line, obtain the radius of curvature of each point in the track section ahead; The track section is divided into multiple continuous unit segments. For each unit segment, the quotient of the unit segment length and the corresponding radius of curvature is calculated, and the quotients of all unit segments are added together to obtain the cumulative curvature value. The sand-spreading duration is obtained by curvature compensation of the initial sand-spreading duration.

3. The sand spreading and cleaning control system for rail vehicles as described in claim 2, characterized in that, The method for obtaining the sand-spreading start time is as follows: Obtain the first distance between the starting point of the ice zone and the current position of the vehicle; The second distance is the distance the vehicle travels at its current speed for the duration of sand application. The difference between the first distance and the second distance is used as the lead distance between the sand-spreading trigger position and the starting point of the ice zone; The ratio of this lead distance to the current vehicle speed is taken as the time lead. The sum of the vehicle's current time and the time advance is taken as the sand-spreading start time; If the sand-spreading signal fails before the sand-spreading start time, the sand-spreading operation will be cancelled.

4. The sand spreading and cleaning control system for rail vehicles as described in claim 1, characterized in that, The content of the trigger compensation action is as follows: During the sand spreading process, the actual sand spreading flow rate in the sand spreading pipe is collected in real time through a flow sensor; Obtain the inherent fluctuation range of the benchmark sand-spreading flow rate under the calibrated pressure, and combine it with the ratio of the current gas source pressure to the calibrated pressure to dynamically determine the theoretical allowable fluctuation range of the target sand-spreading flow rate; Calculate the absolute value deviation between the actual sand spreading flow rate and the target sand spreading flow rate, compare the absolute value deviation with the theoretical allowable fluctuation range, and determine that the flow rate deviation exceeds the limit when the absolute value deviation is greater than the theoretical allowable fluctuation range; When the flow rate deviation is determined to be excessive, the opening of the air source control valve is adjusted according to the direction and magnitude of the deviation. If the actual sand spreading flow rate is less than the target sand spreading flow rate, the valve opening is increased; if the actual sand spreading flow rate is greater than the target sand spreading flow rate, the valve opening is decreased. After adjusting the valve opening, continue monitoring the actual sand-spreading flow rate until the absolute value deviation within N consecutive sampling periods is less than the theoretically permissible fluctuation range, thus completing the compensation action.

5. A sand spreading and cleaning control system for rail vehicles as described in claim 1, characterized in that, The sand cleaning operation includes the following: When it is determined that the nozzle pressure difference change rate deviates from the normal trend, the current sand spreading operation should be immediately interrupted, the sand spreading material supply should be turned off, the airflow pulse and negative pressure suction should be started, and the nozzle pressure difference change rate should be re-tested after the sand is cleaned. If the re-detected nozzle differential pressure change rate returns to the normal trend, then return to the original sand-spreading operation; If the situation does not return to normal, extend the sand cleaning time and repeat the sand cleaning procedure. If the situation still does not recover after a maximum of three sand cleaning procedures, issue a nozzle blockage alarm signal.

6. A method for controlling sand spreading and cleaning on rail vehicles, comprising the following steps performed by a rail vehicle sand spreading and cleaning control system as described in claim 1, characterized in that, Includes the following steps: Based on real-time acquired images of the ice surface and the track cross slope angle of the preceding track section, cluster analysis is performed on the ice surface images. The results of the cluster analysis are combined with the track cross slope angle to determine whether to output a sand-spreading signal and ice surface distribution feature labels for the ice area. When a sand-spreading signal is received, the spraying angle of the sand-spreading nozzles is dynamically set based on the current vehicle speed and ice surface distribution feature tags, and the track cross slope angle is integrated. The target sand-spreading flow rate is set based on the air source pressure, and the sand-spreading duration and start time are determined. Sanding is performed at the spraying angle at the start of sanding. During the sanding process, the actual sanding flow rate is compared with the target sanding flow rate in real time. If the deviation exceeds the tolerance range dynamically determined based on the flow calibration error, a compensation action is triggered. During and after sand spreading, the nozzle pressure differential change rate is monitored in real time. If the change rate deviates from the normal trend, a sand cleaning operation is performed.

Images