Multifunctional antifreeze intelligent spraying system and method

Abstract

The present application belongs to the technical field of railway coal transportation, and specifically relates to a multifunctional antifreeze intelligent spraying system and method. The system comprises a multi-axis robot unit, which is arranged on one side or both sides of a railway track; a multifunctional spray head assembly, which is installed at the execution end of the multi-axis robot unit; a transmission control system, which is drivingly connected with the multi-axis robot unit; a carriage interior condition sensing unit, which is used for acquiring the internal structure information and coal loading state information of a coal transport carriage in real time; and an intelligent control unit, which is electrically connected with the transmission control system, the multifunctional spray head assembly, the carriage interior condition sensing unit, a carriage number recognition unit and a data storage unit. The multi-axis robot unit is provided with the multifunctional spray head assembly with multiple independently controllable spray nozzles, thereby breaking through the limitation of the movement freedom degree of the traditional fixed or swing arm type spraying mechanism.

Description

Technical Field

[0001] This invention belongs to the field of railway coal transportation technology, specifically a multifunctional intelligent antifreeze spraying system and method. Background Technology

[0002] In my country's northern coal-producing regions, winter temperatures are extremely low, making coal highly susceptible to freezing during long-distance rail transport. This freezing causes coal to adhere to the car walls and floor, leading to unloading difficulties, reduced vehicle turnover efficiency, and even transportation safety hazards. Currently, spraying antifreeze on the inner walls of coal train cars and the surface of the coal is the main technical means to solve the problem of frozen cars.

[0003] Existing antifreeze spraying devices have the following main drawbacks: First, the spraying actuators are mostly fixed or simple swing-arm structures with limited freedom of movement, making it difficult to adapt to the spraying needs of different vehicle models and loading conditions, and dead zones are easily formed in the four corners and gaps of the vehicle compartment; Second, the spraying control is crude and cannot adjust the spraying strategy according to the actual situation inside the vehicle compartment, resulting in insufficient spraying in some areas and excessive spraying in other areas; Third, after the spraying operation is completed, the residual antifreeze is very easy to freeze in the low temperature environment, clogging the nozzles and affecting the normal operation of the next spraying operation; Fourth, there is a lack of precise control over the mapping relationship between spraying points, coverage area and spray volume, making it difficult to achieve precise antifreeze spraying on demand.

[0004] In recent years, multi-axis robot technology has been widely used in the field of industrial automation. Its characteristics such as multiple degrees of freedom, high flexibility, and programmable control provide an ideal solution for complex trajectory operations. Introducing multi-axis robot technology into the field of antifreeze spraying is expected to completely solve the inherent defects of traditional spraying devices and realize intelligent and precise spraying operations. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a multifunctional intelligent antifreeze spraying system and method.

[0006] This invention adopts the following technical solution: a multifunctional intelligent antifreeze spraying system, comprising: A multi-axis robot unit, which is set on one or both sides of the railway track, is used to carry the nozzle assembly and perform spraying actions; A multi-functional nozzle assembly, which is installed at the end of the execution of a multi-axis robot unit, includes multiple independently controllable nozzles; A transmission control system is connected to the multi-axis robot unit for driving and controlling the multi-axis robot unit to complete trajectory movement and attitude adjustment in three-dimensional space. The internal condition sensing unit of the coal car is used to acquire the internal structure information and coal loading status information of the coal car in real time. The intelligent control unit is electrically connected to the transmission control system, the multi-functional nozzle assembly, the carriage interior situation sensing unit, the train number recognition unit, and the data storage unit. The intelligent control unit is used to synchronously generate the contour motion trajectory of the multi-axis robot unit, the switching sequence of each nozzle, and the real-time spraying flow parameters based on the carriage interior information and the real-time speed of the train, and drive each unit to work together to complete the full coverage spraying of the five sides of the carriage under moving conditions. An automatic cleaning unit, which is connected to a multi-functional nozzle assembly, is used to automatically remove the antifreeze residue inside the multi-functional nozzle assembly and at the nozzle after the spraying operation of a single carriage is completed. The train number identification unit is used to identify the train number information of passing trains; the data storage unit is used to store the spraying operation records of each carriage.

[0007] In some embodiments, the vehicle interior situation sensing unit includes: A machine vision camera is installed at the end of a multi-axis robot unit to acquire images of the interior of the carriage and extract the contour of the carriage wall and the morphological features of coal accumulation. A lidar sensor, which is mounted on a support beside the track, is used to scan the three-dimensional contour of the interior of the carriage and obtain the carriage size parameters and coal loading height data; The intelligent control unit fuses the data collected by the machine vision camera and the lidar sensor to generate a three-dimensional geometric model of the interior of the vehicle.

[0008] In some embodiments, the intelligent control unit integrates a trajectory planning module, a timing control module, and a fuzzy adaptive PID controller. The trajectory planning module is used to generate the contour motion trajectory of the multi-axis robot unit based on the position and size information of the carriage and the real-time speed of the train, so that the nozzle maintains a preset spraying distance and angle along the inner wall contour of the carriage. The timing control module is used to control the opening and closing timing of each nozzle according to the position signal of the carriage, so as to achieve sequential spraying of the five sides of the carriage in coordination with the robot's movement; The fuzzy adaptive PID controller is used to adjust the spraying flow rate in real time according to the train speed, ambient temperature and coal moisture content to ensure a constant spraying amount per unit area.

[0009] In some embodiments, the automatic cleaning unit includes: An air source, which is connected to the liquid path of the multi-functional nozzle assembly via an air pipeline; A pneumatically controlled valve, wherein the pneumatically controlled valve is installed on a pneumatic pipeline; A heating device, which is wrapped around the outside of the air passage and covered with insulation material, is used to heat the compressed air used for purging. A temperature sensor is installed in the gas pipeline downstream of the heating device to detect the temperature of the purging gas. Antifreeze recovery line, which is connected to the inlet of the multi-functional nozzle assembly, is used to return the residual antifreeze that has been blown out; A recovery control valve is provided on the antifreeze recovery pipeline; The pneumatic control valve, recovery control valve, and temperature sensor are all electrically connected to the intelligent control unit. The intelligent control unit controls the heating device to start and stop according to the temperature feedback, and controls the valve to open and close according to the preset sequence to complete the purging and cleaning.

[0010] In some embodiments, the multi-axis robot unit is a six-degree-of-freedom industrial robot, including a base fixed to a foundation next to the track, a waist joint rotatably mounted on the base, an upper arm hinged to the waist joint, a forearm hinged to the upper arm, and a wrist joint located at the end of the forearm and having three degrees of rotational freedom. The multi-functional nozzle assembly includes a nozzle body with at least five nozzles. The spray direction and angle of each nozzle correspond to the front panel, rear panel, side panels, and floor panel of the vehicle body, respectively. Each nozzle has an independently installed electrically controlled valve that is electrically connected to the intelligent control unit.

[0011] A spraying method for a multifunctional intelligent antifreeze spraying system includes the following steps: S1: Before a coal train enters the spraying area, the train number is identified by the train number recognition unit and transmitted to the intelligent control unit. S2: The internal images and three-dimensional contour data of the compartment to be sprayed are collected in real time by the compartment internal situation perception unit, and after fusion processing, they are transmitted to the intelligent control unit to obtain information on the compartment structure and coal loading status. S3: The intelligent control unit synchronously generates the contour motion trajectory of the multi-axis robot unit, the switching sequence of each nozzle, and the initial spray flow parameters based on the carriage information and the real-time speed of the train. S4: When the front of the carriage enters the spraying operation area, the intelligent control unit drives the multi-axis robot unit to move along the contour trajectory, and at the same time controls the corresponding nozzles to open according to the time sequence, and dynamically adjusts the spraying flow according to the real-time speed of the train to complete the full coverage spraying of the five sides inside the carriage. S5: When the truck leaves the spraying area, the intelligent control unit closes all nozzles and starts the automatic cleaning unit to blow and clean the multi-functional nozzle assembly to remove residual antifreeze. S6: The intelligent control unit records the spraying operation parameters to the data storage unit and completes the spraying operation of a single carriage.

[0012] In some embodiments, step S2 specifically includes: S21: The robot acquires images of the interior of the carriage through a machine vision camera installed at the end of the robot, and extracts the outline of the inner wall of the carriage and the morphological features of coal accumulation through image recognition. S22: The interior of the car is scanned by lidar sensors installed beside the track to obtain three-dimensional data on the dimensions of the car and the height of the coal loading. S23: The visual image features are fused with the 3D data of the LiDAR to generate a 3D geometric model of the interior of the vehicle, which is then transmitted to the intelligent control unit.

[0013] In some embodiments, step S3 specifically includes: S31: The trajectory planning module generates the contour motion trajectory of the multi-axis robot unit based on the position and size information of the carriage and the coal loading pattern; S32: The timing control module generates the opening and closing timing of each nozzle based on the position signal of the carriage, and coordinates with the robot's movement to achieve sequential spraying of the carriage's five sides; S33: The fuzzy adaptive PID controller calculates the real-time spraying flow rate based on the train speed, ambient temperature, and coal moisture content to ensure a constant spraying rate per unit area.

[0014] In some embodiments, step S33 specifically includes: S331: Establish the target spray flow rate function Q target =q0×v×W×f(T)×g(M), Where q0 is the baseline spraying rate per unit area, v is the real-time speed of the train, W is the effective spraying width, f(T) is the temperature correction coefficient corresponding to the ambient temperature, and g(M) is the moisture content correction coefficient corresponding to the coal moisture content; the lower the ambient temperature and the higher the coal moisture content, the larger the value of the corresponding correction coefficient. S332: Input the detected instantaneous train speed, real-time ambient temperature and coal moisture content parameters into the fuzzy adaptive PID controller to calculate the real-time target spray flow rate; S333: Real-time acquisition of actual spray flow rate, using the deviation between target flow rate and actual flow rate and the rate of change of deviation as input, online tuning of PID control parameters through fuzzy inference, and output of control quantity to adjust valve opening or liquid supply pump speed; S334: It adopts a feedforward-feedback composite control structure, which generates the feedforward target flow rate through the speed feedforward channel and corrects the flow rate deviation by combining the PID feedback channel to ensure a constant spraying rate per unit area.

[0015] In some embodiments, in step S5, the purging and cleaning are performed in the following sequence: First, close all the electronically controlled valves of the nozzles. Then, open the recovery control valve of the antifreeze recovery line, and then open the pneumatic control valve of the air line. Use heated compressed air to purge the residual antifreeze inside the nozzles and return it for recovery. After the preset purging time, close the pneumatic control valve and the recovery control valve in sequence to complete the cleaning.

[0016] Compared with the prior art, the present invention has the following beneficial effects: This invention utilizes a multi-axis robotic unit equipped with a multi-nozzle, independently controllable, multi-functional spray head assembly. This overcomes the limitations of motion freedom in traditional fixed or swing-arm spraying mechanisms, allowing it to adapt to different vehicle types and coal loading conditions, achieving full coverage spraying of the five sides of the carriage without blind spots. Combining machine vision and lidar fusion perception mechanisms, along with intelligent control unit contour trajectory planning, nozzle timing control, and fuzzy adaptive PID flow adjustment, the spraying strategy can be dynamically adjusted according to the actual conditions inside the carriage and the real-time speed of the train. This ensures a uniform and constant spray volume per unit area, effectively reducing antifreeze waste while maintaining antifreeze effectiveness. The accompanying automatic cleaning unit automatically blows away and recovers residual antifreeze from the nozzles after the operation of a single carriage, effectively solving the problem of nozzle freezing and clogging in low-temperature environments. This improves system reliability and operational continuity, ultimately realizing intelligent, precise, and efficient antifreeze spraying operations for railway coal transportation. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall structure; Figure 2 This is a schematic diagram of the structure of a multi-axis robot unit; Figure 3 This is a schematic diagram of the structure of a multi-functional nozzle assembly; Figure 4 This is a block diagram of the control system. Figure 5 This describes the working principle of the automatic cleaning unit. In the diagram: 1-Multi-axis robot unit, 2-Multi-functional nozzle assembly, 3-Transmission control system, 4-Cargo compartment interior perception unit, 5-Intelligent control unit, 6-Automatic cleaning unit, 7-Base, 8-Waist joint, 9-Upper arm, 10-Lower arm, 11-Wrist joint, 12-Nozzle body, 13-Nozzle, 14-Electrically controlled valve, 15-Machine vision camera, 16-LiDAR sensor, 17-Air source, 18-Air pipeline, 19-Pneumatic control valve, 20-Heating device, 21-Temperature sensor, 22-Recovery pipeline, 23-Recovery control valve, 24-Vehicle number recognition unit, 25-Data storage unit. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are some embodiments of the present invention, but 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.

[0019] like Figure 1 As shown, a multi-functional antifreeze intelligent spraying system includes: a multi-axis robot unit 1, which is disposed on one or both sides of the railway track and is used to carry the nozzle assembly and complete all spraying actions; a multi-functional nozzle assembly 2, which is installed at the end effector of the multi-axis robot unit 1; a transmission control system 3, which is driven and connected to the multi-axis robot unit 1 and is used to control the multi-axis robot unit 1 to complete trajectory movement and attitude adjustment in three-dimensional space; and a car interior situation sensing unit 4, which is used to acquire real-time internal structure information and loading status information of the coal car. The intelligent control unit 5 is electrically connected to the transmission control system 3, the multi-functional nozzle assembly 2, the carriage interior situation sensing unit 4, the car number recognition unit 24, and the data storage unit 25, respectively, and is used to generate the motion trajectory of the multi-axis robot unit 1 and the nozzle control parameters based on the carriage interior information; the automatic cleaning unit 6 is connected to the multi-functional nozzle assembly 2 and is used to automatically remove the antifreeze residue inside the multi-functional nozzle assembly 2 and at the nozzle after the spraying operation is completed; the car number recognition unit 24 is used to identify the car number information of the passing train; and the data storage unit 25 is used to store the spraying records of each carriage.

[0020] A multi-axis robot unit 1 is positioned on one or both sides of the railway track, serving as the main actuator for spraying operations. It possesses multiple degrees of freedom and carries the nozzle assembly, completing all spraying actions. A multi-functional nozzle assembly 2 is installed at the end effector of the multi-axis robot unit, comprising the nozzle body and multiple independently controlled nozzles. These nozzles have different spray directions and angles, corresponding to the front, rear, side, and bottom panels of the car body, respectively. A transmission control system 3 is integrated inside or outside the multi-axis robot unit and is connected to its drive system. It controls the multi-axis robot unit's trajectory movement and attitude adjustment in three-dimensional space. A car body interior sensing unit 4 acquires real-time information on the internal structure and loading status of the coal car body. An intelligent control unit 5 is electrically connected to the transmission control system, the multi-functional nozzle assembly, and the car body interior sensing unit, generating the motion trajectory of the multi-axis robot unit. An automatic cleaning unit is connected to the multi-functional nozzle assembly and automatically removes residual antifreeze from inside the nozzle assembly and at the nozzle nozzles after spraying operations.

[0021] In specific embodiments, such as Figure 2 As shown, the multi-axis robot unit 1 is a six-degree-of-freedom industrial robot, including a base 7, a waist joint 8, an upper arm 9, a forearm 10, and a wrist joint 11. The base 7 is fixedly mounted on a foundation beside the track, connected by high-strength bolts to ensure stability during high-speed robot movement. The waist joint 8 is rotatably mounted on the base 7, driven by a servo motor and a precision reducer to achieve 360° continuous rotation, enabling the robot to cover a larger working area in the horizontal plane. The upper arm 9 is hinged to the waist joint 8, allowing for pitch motion around a horizontal axis. The forearm 10 is hinged to the upper arm 9, allowing for swinging relative to the upper arm 9. The wrist joint 11 is located at the end of the forearm 10, possessing three degrees of rotational freedom, allowing for arbitrary adjustment of the nozzle posture. All joints of the multi-axis robot unit 1 are driven by absolute encoder servo motors, ensuring high repeatability and meeting the positioning requirements for precise spraying.

[0022] In terms of installation, this embodiment adopts a side-mounted installation, with the robot base fixed to the side of the track foundation, and the robot arm extending horizontally above the track to perform spraying operations. This method is simple to install and provides ample maintenance space, but the robot's working range is limited by its arm span. For wider carriages, it may be necessary to configure two robots for collaborative operation.

[0023] In a specific embodiment, the multi-functional nozzle assembly 2 includes a nozzle body 12, on which multiple nozzles 13 are provided. Each nozzle 13 has an independently provided electrically controlled valve 14 on its pipeline, and the electrically controlled valve 14 is electrically connected to the intelligent control unit 5.

[0024] The multi-functional nozzle assembly 2 is mounted on the execution flange of the wrist joint 11 of the multi-axis robot unit 1. For example... Figure 3 As shown, the multi-functional nozzle assembly 2 includes a nozzle body 12 and five independently controlled nozzles 13. The five nozzles 13 are mounted on the nozzle body 12, and their spray directions and spray angles correspond to the five inner surfaces of the carriage, respectively. The multiple nozzles include a front plate nozzle, a rear plate nozzle, a left side plate nozzle, a right side plate nozzle, and a bottom plate nozzle, each corresponding to one of the five inner surfaces of the carriage. Specifically, the axis of the front plate nozzle forms a certain angle with the axis of the nozzle body and is used to spray towards the front plate of the carriage; the rear plate nozzle is used to spray towards the rear plate of the carriage; the left side plate nozzle and the right side plate nozzle spray towards the left and right side plates, respectively; and the bottom plate nozzle is vertically downward or slightly forward-leaning and is used to spray towards the bottom plate of the carriage or the coal surface. Each nozzle 13 has an independently installed electrically controlled valve 14 on its pipeline, which is electrically connected to the intelligent control unit 5 to realize independent on / off control and flow regulation of each nozzle 13.

[0025] In a specific embodiment, the transmission control system 3 is integrated inside the multi-axis robot unit 1, including a servo driver, a motion controller, and an absolute encoder. The servo driver receives instructions from the intelligent control unit 5 and drives the servo motors of each joint to move along a predetermined trajectory; the motion controller is responsible for robot kinematics calculation and interpolation calculation to achieve smooth trajectory control; the absolute encoder provides real-time feedback on the position information of each joint, forming a closed-loop control.

[0026] In a specific embodiment, the carriage interior perception unit 4 includes: a machine vision camera 41, which is installed at the end of the multi-axis robot unit 1 to collect image information inside the carriage and extract the carriage outline and coal accumulation morphology features; and a lidar sensor 42, which is installed on a support beside the track to scan the three-dimensional outline inside the carriage and obtain carriage size data and loading height data; the data from the machine vision camera 41 and the lidar sensor 42 are processed in the intelligent control unit 5.

[0027] In a specific embodiment, the intelligent control unit 5 integrates a trajectory planning module, a timing control module, and a fuzzy adaptive PID controller. The trajectory planning module generates the contour motion trajectory of the multi-axis robot unit 1 based on the carriage position, size information, and real-time train speed. The timing control module controls the opening and closing sequence of each nozzle 13 based on the carriage position signal. The PID controller adjusts the spraying flow rate in real time based on the train speed, ambient temperature, and coal moisture content.

[0028] The intelligent control unit 5 adopts an architecture combining an industrial control computer and a programmable logic controller (PLC). It is electrically connected to the transmission control system 3, the electrically controlled valve 14 of the multi-functional nozzle assembly 2, and the carriage interior situation sensing unit 4. It contains a trajectory planning module, a timing control module, and a fuzzy adaptive PID controller. The trajectory planning module generates the motion trajectory of the multi-axis robot unit 1 based on the carriage position and size information obtained from the carriage interior situation sensing unit 4, combined with the train's real-time speed. That is, the robot drives the nozzles to move along the contour of the carriage's inner wall, always maintaining the optimal spray distance and angle. The timing control module controls the opening and closing sequence of each nozzle 13 based on the carriage position signal transmitted from the car number identification unit 24 and the position sensor, coordinating with the robot's movement to achieve sequential spraying of the carriage's five sides: when the robot moves the nozzles to the front area of ​​the carriage, the front nozzles open; when the robot moves along the side wall of the carriage, the side nozzles open; when the robot moves to the rear area of ​​the carriage, the rear nozzles open; throughout the movement, the bottom nozzles can be continuously opened or intermittently opened depending on the coal accumulation. The PID controller receives the instantaneous train speed signal from the radar speed measurement module. The required spraying volume is calculated based on the ambient temperature, coal moisture content, and train speed. The lower the ambient temperature, the higher the coal moisture content, and the faster the train speed, the greater the required spraying volume.

[0029] In a specific embodiment, the automatic cleaning unit 6 includes: an air source 17, which is connected to the multi-functional nozzle assembly 2 via an air passage pipe 18; an air control valve 19, which is disposed on the air passage pipe 18; a heating device 20, which is wrapped around the outside of the air passage pipe 18 and covered with insulation material; a temperature sensor 21, which is disposed inside the air passage pipe 18 behind the heating device 20; an antifreeze recovery pipe 22, which is connected to the liquid inlet of the multi-functional nozzle assembly 2; and a recovery control valve 23, which is disposed on the antifreeze recovery pipe 22. The air control valve 19, the recovery control valve 23, and the temperature sensor 21 are all electrically connected to the intelligent control unit 5.

[0030] The air source 17 uses a screw air compressor to provide compressed air at 0.6-0.8 MPa, and is equipped with an air tank and a drying and filtering device to ensure the stability and cleanliness of the air source. The air pipeline 18 uses a stainless steel flexible hose, which runs from the air source 17 along the robot arm to the liquid inlet of the multi-functional nozzle assembly 2. The air control valve 19 is located on the air pipeline 18 and is a two-position, two-way solenoid valve used to control the on / off of compressed air.

[0031] The heating device 20 uses an electric heating strip wrapped around the outside of the air duct 18, and is covered with insulation material to heat the compressed air and prevent residual antifreeze at the nozzle from freezing instantly during purging in low-temperature environments. A temperature sensor 21 is located inside the air duct 18 behind the heating device 20 to detect the compressed air temperature. The temperature sensor 21 is electrically connected to the intelligent control unit 5, which controls the start and stop of the heating device 20 based on temperature feedback to ensure a suitable purging gas temperature.

[0032] The antifreeze recovery line 22 is connected to the inlet of the multi-functional nozzle assembly 2, and the recovery control valve 23 is installed on the antifreeze recovery line 22. Both the pneumatic valve 19 and the recovery control valve 23 are electrically connected to the intelligent control unit 5. After spraying, the intelligent control unit 5 first closes all the electrically controlled valves 14, then controls the recovery control valve 23 to open, and then controls the pneumatic valve 19 to open, using heated compressed air to purge the residual antifreeze inside the nozzle body 12 and at the nozzle 13 opening. The purged residual antifreeze flows back to the storage tank through the recovery line 22 under air pressure. After purging for 3-5 seconds, the intelligent control unit 5 closes the pneumatic valve 19 and the recovery control valve 23, completing the automatic cleaning. The entire cleaning process is completed within 10 seconds after spraying, effectively preventing the residual antifreeze from freezing at low temperatures.

[0033] In a specific embodiment, the vehicle number identification unit 24 uses an RFID reader / writer, installed beside the track, to identify the electronic vehicle number tags of passing trains and obtain vehicle number information. The data storage unit 25 stores the spraying records for each carriage, including spraying time, spray volume, ambient temperature, vehicle number, robot trajectory parameters, and other information. Both the vehicle number identification unit 24 and the data storage unit 25 are electrically connected to the intelligent control unit 5, facilitating subsequent data traceability and spraying effect analysis, and providing data support for optimizing spraying strategies.

[0034] A spraying method based on the multifunctional antifreeze intelligent spraying system according to any one of claims 1-6, comprising: S1: Before the coal train enters the spraying area, the train number identification unit 24 identifies the train number and transmits it to the intelligent control unit 5.

[0035] S2: The vehicle interior situation perception unit 4 collects real-time images and three-dimensional contour data of the vehicle interior that is about to enter the spraying area, and transmits them to the intelligent control unit 5 after fusion processing.

[0036] Specifically, S2 includes: S21: The machine vision camera 41 collects image information inside the carriage and extracts features such as the carriage outline and coal accumulation shape through image recognition algorithms; The carriage contour algorithm first acquires the original images of the carriage interior via visual camera 15, and then performs preprocessing operations (image grayscale conversion and denoising, image enhancement). The acquired carriage interior images are then processed using the U2-Net salient object detection algorithm. This algorithm is based on a deep residual U-shaped three-layer network structure. The input convolutional layer converts the input feature map into an intermediate feature map for local feature extraction; the L-layer U-Net symmetric encoder-decoder structure increases network depth through pooling operations to capture contextual information at different scales; residual connections fuse local and multi-scale features through summation, preserving original information details; multi-level feature fusion achieves accurate extraction of the carriage interior wall contour, outputting the carriage length, width, height, and coordinates of each wall corner point.

[0037] The coal accumulation morphology algorithm performs edge detection within the car body contour, employing the Canny edge detection operator to extract edge features of the coal accumulation. The Canny operator includes Gaussian filtering to smooth the image, calculating gradient magnitude and direction, non-maximum suppression, and double-threshold edge connection calculation steps. Then, an improved Hough transform algorithm is used to extract the edge contour of the coal accumulation. Based on the gradient magnitude and gradient direction angle of each edge point, a voting window centered on the gradient direction angle is determined, and complex voting is performed on the Hough accumulator. After smoothing, local maxima are extracted as line parameters. Based on the edge contour extracted by the Hough transform, a sliding window statistical method is used to determine the coal accumulation morphology: a sliding window with a width of W (W=50 pixels) and a step size of S (S=10 pixels) is set along the length of the car body. Within each window, the distribution density D and average height H of continuous edge pixels are statistically analyzed: D=N edge / W_total. Where N is... edge W represents the number of pixels at the inner edge of the window. 总 This represents the total number of pixels in the window. When D exceeds the threshold T... d (T) d =0.3) and H exceeds the threshold T h When the window area is identified as a raised area of ​​coal accumulation, the statistical results of each window are used to generate a height distribution curve of the coal accumulation, thus identifying the contact interface between the coal and the car wall.

[0038] S22: The lidar sensor 42 scans the three-dimensional outline of the interior of the carriage to obtain accurate carriage size data and loading height data; S23: The data collected by the machine vision camera 41 and the lidar sensor 42 are processed by the front end and then transmitted to the intelligent control unit 5.

[0039] S3: The intelligent control unit 5 generates the contour motion trajectory of the multi-axis robot unit 1 based on the carriage information and the real-time speed of the train, and generates the switching sequence of each nozzle 13 and calculates the initial spray flow rate.

[0040] The intelligent control unit 5 is the "brain" of the entire spraying system. Its core task is to synchronously generate three control commands based on the sensor information inside the carriage and the real-time speed of the train: the robot's contour-following motion trajectory, the switching sequence of each nozzle, and the initial spray flow rate. This process involves the collaborative calculation of three control sub-modules, and its specific implementation process is as follows: 1) The intelligent control unit 5 receives the interior information of the carriage collected by the vision camera 15 and the lidar sensor 16, extracts the contour of the inner wall of the carriage through the U²-Net saliency detection algorithm, identifies the height distribution of the coal pile and the position of the contact interface through the improved Hough transform and sliding window statistical method, and fuses them to generate a three-dimensional geometric model of the carriage. 2) Based on the carriage model, spray path points are generated by sampling at fixed intervals along the length of the carriage. The target position and attitude of the nozzle at each path point are calculated according to the coal pile shape. Linear interpolation and circular interpolation are used to generate continuous end motion paths, and S-shaped velocity curves are used for velocity planning. 3) The six-degree-of-freedom robot inverse kinematics model established by the DH parameter method converts the Cartesian space path points into a sequence of joint angular displacements. A fifth-order polynomial is used to interpolate each joint, and a time synchronization strategy is used to achieve multi-axis coordinated motion and generate a sequence of joint space trajectory points. 4) Determine the trigger time for the carriage to enter and leave the spraying area based on the sensor signals, and calculate the opening and closing times of the front panel nozzles, side panel nozzles, bottom panel nozzles and rear panel nozzles in combination with the robot's movement progress. Generate the nozzle switching time and dynamically correct each time parameter proportionally when the train speed changes. 5) Based on the target unit area spraying rate q target Given the real-time velocity v(t) and the effective spray width W, calculate the initial target flow rate Q0=q. target ×v(t)×W, introducing correction coefficients for ambient temperature and coal moisture content, outputs the initial parameters and feedforward control signal of the PID controller; 6) The intelligent control unit (5) outputs the joint space trajectory point sequence to the transmission control system 3 through unified management, outputs the nozzle switching time schedule to the multi-functional nozzle assembly 2, and outputs the initial flow rate and PID parameters to the control unit to ensure the precise synchronization of the three commands.

[0041] S4: When the front of the vehicle enters the spraying area, the intelligent control unit 5 starts the multi-axis robot unit 1 to move along the predetermined trajectory, and at the same time controls the corresponding electronically controlled valves to open according to the time sequence, and dynamically adjusts the spraying flow according to the real-time vehicle speed to complete the full coverage spraying of the five sides of the vehicle.

[0042] Step S3 includes: S31: The trajectory planning module generates the motion trajectory of the multi-axis robot unit based on the position and size information of the carriage; The trajectory planning module plans a smooth and continuous motion path from the initial pose to the target pose for each joint of the multi-axis robot unit, while ensuring that the nozzle remains aligned with the predetermined spray point on the inner wall of the vehicle during spraying. The specific process of motion trajectory generation is as follows: 1) The kinematic model of the multi-axis robot is established using the DH parameter method. The nozzle end pose is calculated by forward kinematics and the target pose is converted into the angular displacement of each joint by inverse kinematics. 2) In Cartesian space, the spraying start point, spraying path point and spraying end point are determined according to the size of the carriage and the coal accumulation shape. Linear interpolation and circular interpolation are used to generate a continuous end motion path, and an S-shaped speed planning strategy is used to plan the path speed to ensure that the motion process is smooth and without impact. 3) The path point sequence obtained by Cartesian space programming is converted into a trajectory point sequence in joint space through inverse kinematics solution. A continuous change function of the angular displacement of each joint over time is generated in joint space, and the trajectory is smoothed by B-spline curve. 4) Verify and segment the generated trajectory to ensure that the speed, acceleration and torque of each joint meet the robot's performance limits, and finally output the motion trajectory that can be executed by the transmission control system.

[0043] S32; The timing control module controls the opening and closing sequence of each nozzle according to the position signal of the carriage, and coordinates with the robot's movement to achieve sequential spraying of the five sides of the carriage; Based on the carriage position signal, the opening and closing of each nozzle is precisely controlled according to a preset timing sequence. Combined with the movement trajectory of the multi-axis robot, this achieves sequential spraying of the five inner surfaces of the carriage (front panel, rear panel, side panels, and floor panel), ensuring full coverage while minimizing antifreeze overflow and waste. The specific process for the opening and closing sequence of each nozzle is as follows: 1) The timing control module receives the position signal of the carriage. When it detects that the front end of the carriage enters the spraying area, it records the trigger time E1; when it detects that the rear end of the carriage leaves the spraying area, it records the trigger time E2. 2) Based on the E1 time and the real-time train speed v(t) provided by the radar speed measurement module, calculate the opening and closing times of each nozzle: the front plate nozzle opens in advance before the E1 time, the side plate and bottom plate nozzles open in advance at the E1 time and close in advance at the E2 time, and the rear plate nozzle opens in advance before the E2 time and closes in advance after the E2 time. 3) The timing control module synchronizes the generated nozzle switching command sequence with the robot motion trajectory of the trajectory planning module, and ensures that the corresponding nozzle is turned on when the robot reaches the target spraying position through a unified system clock; 4) During spraying, the timing control module dynamically corrects various time parameters based on real-time speed detection results: when the actual speed deviates from the preset speed, it adjusts the timing parameters according to the time interval.actual =ti×(v0 / v(t)) performs speed compensation at the switching moment; when the carriage position deviation Δx is detected, press ti. corrected =ti+Δx / v(t) is used for position compensation; 5) When a sudden speed change or abnormal sensor signal is detected, the timing control module immediately activates the safety protection mode, shuts down all nozzles and sends an emergency stop signal to the transmission control system.

[0044] S33: The PID controller adjusts the spraying flow rate in real time according to the train's running speed to ensure a constant spraying volume per unit area.

[0045] The spraying process is a system with large time delays, nonlinearity, and multiple disturbances, making it difficult to accurately establish model parameters. Therefore, this invention employs a fuzzy adaptive PID cascade flow controller, and the specific process of online tuning of PID parameters using fuzzy inference is as follows: 1) Establish the functional relationship between spraying volume and train speed: Q target =q target ×v×W, where q target The target spraying rate is defined as v, the real-time speed of the train, and the effective spraying width. Ambient temperature T and coal moisture content M are introduced as correction factors, and the correction coefficients f(T) and g(M), which are calibrated by experiments, are used to correct the target flow rate. 2) Input the detected instantaneous train speed v(t) into the fuzzy adaptive PID controller, and the controller calculates the current required target flow rate Q. target(t) ; 3) Real-time detection of actual spray flow rate Q_actual(t), calculation of deviation E=Q target -Q actual The rate of change of deviation EC = dE / dt; 4) Fuzzify E and EC into seven variables: NB, NM, NS, ZO, PS, PM, and PB. Use the triangular membership function for fuzzification, perform Mamdani fuzzy inference based on the preset fuzzy rule table, defuzzify using the centroid method, and output the adjustment amount ΔK of the PID parameters. P ΔK I ΔK D ; 5) Update PID parameters: K P =K P0 +ΔK P KI=K I0 +ΔK I K D =K D0 +ΔK D Calculate the control output u(t) and drive the actuator to adjust the valve opening or the speed of the liquid supply pump; 6) A feedforward-feedback composite control structure is adopted. The speed feedforward channel directly calculates the feedforward target flow based on the real-time speed, and the PID feedback channel corrects the deviation between the actual flow and the target flow, thereby improving the system response speed and control accuracy.

[0046] S5: When the vehicle leaves the spraying area, the intelligent control unit 5 closes all electronically controlled valves and activates the automatic cleaning unit 6 to purge and clean the multi-functional nozzle assembly 2. The purging and cleaning is performed in the following sequence: First, close all the electronically controlled valves of the nozzles. Then, open the recovery control valve of the antifreeze recovery line, and then open the pneumatic control valve of the air line. Use heated compressed air to purge the residual antifreeze inside the nozzles and return it for recovery. After the preset purging time, close the pneumatic control valve and the recovery control valve in sequence to complete the cleaning.

[0047] S6: The intelligent control unit 5 records all parameters of this spraying operation to the data storage unit 25, completing the spraying operation of a single carriage.

[0048] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multifunctional intelligent antifreeze spraying system, characterized in that, include: A multi-axis robot unit (1) is provided on one or both sides of the railway track to carry the nozzle assembly and perform spraying actions; Multifunctional nozzle assembly (2), which is installed at the execution end of a multi-axis robot unit (1) and includes multiple independently controllable nozzles (13). The transmission control system (3) is connected to the multi-axis robot unit (1) for driving and controlling the multi-axis robot unit (1) to complete trajectory movement and posture adjustment in three-dimensional space; The internal condition sensing unit (4) of the car body is used to obtain the internal structure information and coal loading status information of the coal car body in real time. The intelligent control unit (5) is electrically connected to the transmission control system (3), the multi-functional nozzle assembly (2), the carriage interior situation perception unit (4), the vehicle number recognition unit (24), and the data storage unit (25), respectively. The intelligent control unit (5) is used to generate the contour motion trajectory of the multi-axis robot unit (1), the switching sequence of each nozzle (13), and the real-time spraying flow parameters in sync with the carriage interior information and the real-time speed of the train, and drive each unit to work together to complete the full coverage spraying of the carriage five sides under the moving working condition. Automatic cleaning unit (6), which is connected to multi-functional nozzle assembly (2), is used to automatically remove the antifreeze residue inside the multi-functional nozzle assembly (2) and at the nozzle after the spraying operation of a single carriage is completed; The train number identification unit (24) is used to identify the train number information of the passing train; the data storage unit (25) is used to store the spraying operation records of each carriage.

2. The multifunctional antifreeze intelligent spraying system according to claim 1, characterized in that, The compartment interior situation sensing unit (4) includes: Machine vision camera (41), the machine vision camera (41) is installed at the execution end of the multi-axis robot unit (1) to collect images inside the carriage and extract the outline of the inner wall of the carriage and the coal accumulation morphology features; A lidar sensor (42) is installed on a trackside bracket to scan the three-dimensional outline inside the carriage and obtain carriage size parameters and coal loading height data. The intelligent control unit (5) fuses the data collected by the machine vision camera (41) and the lidar sensor (42) to generate a three-dimensional geometric model of the interior of the carriage.

3. The multifunctional antifreeze intelligent spraying system according to claim 1, characterized in that, The intelligent control unit (5) integrates a trajectory planning module, a timing control module, and a fuzzy adaptive PID controller. The trajectory planning module is used to generate the contour motion trajectory of the multi-axis robot unit (1) based on the position and size information of the carriage and the real-time speed of the train, so that the nozzle maintains a preset spraying distance and angle along the inner wall contour of the carriage. The timing control module is used to control the opening and closing timing of each nozzle (13) according to the position signal of the carriage, and to achieve sequential spraying of the carriage five-sided structure in coordination with the robot's movement; The fuzzy adaptive PID controller is used to adjust the spraying flow rate in real time according to the train speed, ambient temperature and coal moisture content to ensure a constant spraying amount per unit area.

4. The multifunctional antifreeze intelligent spraying system according to claim 1, characterized in that, The automatic cleaning unit (6) includes: Air source (17), which is connected to the liquid path of the multi-functional nozzle assembly (2) through air pipeline (18); A pneumatic control valve (19) is provided on the pneumatic pipeline (18); Heating device (20), which is wrapped around the outside of the air passage pipe (18) and covered with heat insulation material on the outside, is used to heat the compressed air for purging; Temperature sensor (21), the temperature sensor (21) is installed in the gas pipeline (18) downstream of the heating device (20) and is used to detect the temperature of the purging gas; Antifreeze recovery line (22), which is connected to the inlet of the multi-functional nozzle assembly (2) for returning the residual antifreeze that has been blown out; Recovery control valve (23), said recovery control valve (23) is installed on the antifreeze recovery pipeline (22); The pneumatic control valve (19), the recovery control valve (23) and the temperature sensor (21) are all electrically connected to the intelligent control unit (5). The intelligent control unit (5) controls the heating device (20) to start and stop according to the temperature feedback, and controls the valve to open and close according to the preset time sequence to complete the purging and cleaning.

5. The multifunctional antifreeze intelligent spraying system according to claim 1, characterized in that, The multi-axis robot unit (1) is a six-degree-of-freedom industrial robot, including a base (7) fixed to the foundation next to the track, a waist joint (8) rotatably set on the base (7), an upper arm (9) hinged to the waist joint (8), a forearm (10) hinged to the upper arm (9), and a wrist joint (11) set at the end of the forearm (10) and having three degrees of rotational freedom. The multi-functional nozzle assembly (2) includes a nozzle body (12), on which at least five nozzles (13) are provided. The spray direction and angle of each nozzle (13) correspond to the front panel, rear panel, side panels and bottom panel of the carriage respectively. Each nozzle (13) has an independently provided electrically controlled valve (14) that is electrically connected to the intelligent control unit (5) on its pipeline.

6. A spraying method based on the multifunctional antifreeze intelligent spraying system according to any one of claims 1-5, characterized in that, Includes the following steps: S1: Before the coal train enters the spraying area, the train number is identified by the vehicle number identification unit (24) and transmitted to the intelligent control unit (5). S2: The internal images and three-dimensional contour data of the compartment to be sprayed are collected in real time by the compartment internal situation perception unit (4), and after fusion processing, they are transmitted to the intelligent control unit (5) to obtain the compartment structure and coal loading status information; S3: The intelligent control unit (5) generates the contour motion trajectory of the multi-axis robot unit (1), the switching sequence of each nozzle (13), and the initial spray flow parameters in sync with the carriage information and the real-time speed of the train. S4: When the front end of the carriage enters the spraying operation area, the intelligent control unit (5) drives the multi-axis robot unit (1) to move along the contour trajectory, and at the same time controls the corresponding nozzle (13) to open according to the time sequence, and dynamically adjusts the spraying flow according to the real-time speed of the train to complete the full coverage spraying of the five sides inside the carriage. S5: When the vehicle leaves the spraying area, the intelligent control unit (5) closes all nozzles (13) and starts the automatic cleaning unit (6) to blow and clean the multi-functional nozzle assembly (2) to remove residual antifreeze; S6: The intelligent control unit (5) records the spraying operation parameters to the data storage unit (25) to complete the spraying operation of a single carriage.

7. The multifunctional antifreeze intelligent spraying method according to claim 6, characterized in that, Step S2 specifically includes: S21: The machine vision camera (41) installed at the end of the robot collects images of the interior of the car, and extracts the outline of the inner wall of the car and the coal accumulation morphology features through image recognition; S22: The interior of the car is scanned by a lidar sensor (42) installed next to the track to obtain three-dimensional data of the car dimensions and coal loading height; S23: The visual image features are fused with the three-dimensional data of the lidar to generate a three-dimensional geometric model of the interior of the vehicle and then transmitted to the intelligent control unit (5).

8. The multifunctional antifreeze intelligent spraying method according to claim 6, characterized in that, Step S3 specifically includes: S31: The trajectory planning module generates the contour motion trajectory of the multi-axis robot unit (1) based on the position and size information of the carriage and the coal loading form; S32: The timing control module generates the opening and closing timing of each nozzle (13) according to the position signal of the carriage, and coordinates with the robot movement to realize the sequential spraying of the carriage's five sides; S33: The fuzzy adaptive PID controller calculates the real-time spraying flow rate based on the train speed, ambient temperature, and coal moisture content to ensure a constant spraying rate per unit area.

9. The multifunctional antifreeze intelligent spraying method according to claim 8, characterized in that, Step S33 specifically includes: S331: Establish the target spray flow rate function Q target =q0×v×W×f(T)×g(M), Where q0 is the baseline spraying rate per unit area, v is the real-time speed of the train, W is the effective spraying width, f(T) is the temperature correction coefficient corresponding to the ambient temperature, and g(M) is the moisture content correction coefficient corresponding to the coal moisture content; the lower the ambient temperature and the higher the coal moisture content, the larger the value of the corresponding correction coefficient. S332: Input the detected instantaneous train speed, real-time ambient temperature and coal moisture content parameters into the fuzzy adaptive PID controller to calculate the real-time target spray flow rate; S333: Real-time acquisition of actual spray flow rate, using the deviation between target flow rate and actual flow rate and the rate of change of deviation as input, online tuning of PID control parameters through fuzzy inference, and output of control quantity to adjust valve opening or liquid supply pump speed; S334: It adopts a feedforward-feedback composite control structure, which generates the feedforward target flow rate through the speed feedforward channel and corrects the flow rate deviation by combining the PID feedback channel to ensure a constant spraying rate per unit area.

10. The multifunctional antifreeze intelligent spraying method according to claim 6, characterized in that, In step S5, the purging and cleaning are performed in the following sequence: First, close all the electronically controlled valves of the nozzles. Then, open the recovery control valve of the antifreeze recovery line, and then open the pneumatic control valve of the air line. Use heated compressed air to purge the residual antifreeze inside the nozzles and return it for recovery. After the preset purging time, close the pneumatic control valve and the recovery control valve in sequence to complete the cleaning.

Images