Multi-sensor fusion traffic line marking unmanned vehicle

Abstract

In order to solve the problem that the road marking device is low in automation degree and pipeline is easy to be blocked, the utility model provides a kind of traffic marking unmanned vehicle of multi-sensor fusion, comprising: unmanned vehicle main body, steering wheel and driving device are arranged on unmanned vehicle main body;Marking device, by setting up paint tank, paint pump and spray gun on unmanned vehicle main body, paint pump and inlet and outlet are connected with paint tank and spray gun respectively;Control system, by controller, camera, positioning module, IMU linear measurement unit, laser ranging unit is formed, camera, positioning module, IMU linear measurement unit, laser ranging unit are connected with controller signal input end, paint pump, driving device are connected with controller signal output end, by installing laser radar, camera and other sensors are combined, complete surrounding environment modeling and perception, high-precision automation completes marking work.

Description

Technical Field

[0001] This utility model relates to the field of road marking device technology, and in particular to an unmanned vehicle for traffic marking using multi-sensor fusion. Background Technology

[0002] Road marking is a crucial aspect of road construction. With the continuous expansion and upgrading of highways worldwide, traditional manual marking methods can no longer meet current demands. Therefore, developing efficient and automated marking machinery has become an urgent need for the industry. Furthermore, research on the structure, performance, and efficiency improvements of marking machines is ongoing. To ensure standardized road markings and socio-economic benefits, the use of vehicles is essential for completing the marking work. Utilizing vehicle-assisted marking can significantly reduce labor burden and improve the quality and efficiency of road markings.

[0003] However, the road marking devices currently used in my country are relatively outdated, mostly manual or semi-automatic machines, and have many defects, such as: the heated paint is prone to solidification after cooling, causing blockages; the position and number of nozzles are fixed, lacking flexibility; the marking devices are bulky and difficult to operate; a large amount of manual labor is still required for auxiliary marking, and the degree of automation is low. Utility Model Content

[0004] To address the issues of low automation and easy pipe blockage in road marking devices, this invention provides a multi-sensor fusion-based unmanned traffic marking vehicle, employing the following technical solution:

[0005] A multi-sensor fusion-based unmanned traffic marking vehicle, characterized in that it includes:

[0006] The unmanned vehicle body is equipped with steering wheels and a drive device for driving the unmanned vehicle body forward and turning.

[0007] The marking device consists of a paint tank, a paint pump, and a spray gun mounted on the main body of the unmanned vehicle. The paint pump is connected to the paint tank and the spray gun via its inlet and outlet, respectively.

[0008] The control system consists of a controller, a camera, a positioning module, an IMU linear measurement unit, and a laser ranging unit. The camera, positioning module, IMU linear measurement unit, and laser ranging unit are all connected to the signal input terminal of the controller, and the paint pump and drive device are connected to the signal output terminal of the controller.

[0009] Furthermore, the drive device includes a first drive motor, a differential, and a steering shaft. The first drive motor is fixedly connected to the body of the unmanned vehicle, and the first drive motor is connected to the steering wheel via the differential and the steering shaft.

[0010] Furthermore, the paint tank is equipped with a stirring assembly, which includes a second drive motor, a stirring rod, and stirring blades. The stirring rod is rotatably connected to the inside of the paint tank, and the stirring blades are fixedly connected to the surface of the stirring rod. The second drive motor is fixedly connected to the outside of the paint tank, and the output shaft of the second drive motor is drively connected to the stirring rod.

[0011] Furthermore, a heating component is provided inside the paint tank, which includes a heating plate disposed inside the paint tank, and a temperature sensor is also provided inside the paint tank, which is connected to the signal input terminal of the controller.

[0012] Furthermore, a vent pipe is provided above the paint tank, and an air filter and a switch valve are installed inside the vent pipe.

[0013] Furthermore, the paint tank is also connected to an air compressor, the air outlet of which is connected to the inside of the paint tank and located above the liquid level inside the paint tank.

[0014] Furthermore, a guide frame is fixedly connected to the lower end of the unmanned vehicle body, and a scribing drive assembly is provided on the guide frame. The spray gun is connected to the scribing drive assembly and moves along the width direction of the vehicle through the scribing drive assembly.

[0015] Furthermore, the marking drive assembly includes a third drive motor, a lead screw, and a nut. The two ends of the lead screw are rotatably connected to the guide frame via bearings, and the length of the lead screw is distributed along the width direction of the vehicle. The output shaft of the third drive motor is connected to the lead screw drive. The nut is threadedly connected to the lead screw and fixedly connected to the spray gun.

[0016] Furthermore, a guide block is provided at the lower center of the guide frame.

[0017] The beneficial effects of this utility model are as follows:

[0018] 1. By setting up multiple sets of sensors, including lidar, cameras, IMU linear measurement units and GNSS modules, the vehicle perceives the surrounding environment and controls the unmanned vehicle to complete road marking based on the detection signals, which effectively improves the automation level of road marking.

[0019] 2. The paint tank is equipped with a heating device, which heats the paint inside the tank, thus preventing the paint from condensing and clogging the spray pipes due to low external ambient temperature.

[0020] 3. An exhaust pipe is installed above the paint tank, and an air filter is installed inside the exhaust pipe to filter out the harmful gases generated by heating the paint, thus improving the environmental protection effect. In addition, a switch valve is installed on the exhaust pipe. When the air compressor starts, the switch valve is closed to keep the paint tank sealed, ensuring that the air compressor can force the paint into the delivery hose and spray gun.

[0021] 4. A guide frame is installed under the unmanned vehicle body, and a screw-nut mechanism driven by a motor drives the spray gun to move along the width of the vehicle, making the lateral position of the spray gun adjustable. This structure increases the flexibility of the marking device, and can adjust the position of the marking as needed. It overcomes the defects of fixed nozzle position and lack of flexibility in traditional equipment, and further improves the adaptability and work efficiency of marking construction. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the structure of the present utility model.

[0023] Figure 2 Schematic diagram of guide frame position

[0024] Figure 3 This is a schematic diagram of the bottom structure of the device of this utility model.

[0025] Figure 4 Schematic diagram of heating device and scribing drive assembly

[0026] Among them, 1-unmanned vehicle body, 2-steering wheel, 3-omnidirectional wheel, 4-paint tank, 5-first drive motor, 6-differential, 7-steering shaft, 8-controller, 9-GNSS module, 10-IMU linear measurement unit, 11-camera, 12-LiDAR, 13-support plate, 14-filter, 15-air compressor, 16-stirring rod, 17-stirring blade, 18-second drive motor, 19-heating plate, 20-temperature sensor, 21-guide frame, 22-third drive motor, 23-lead screw, 24-nut, 25-spray gun, 26-alignment block. Detailed Implementation

[0027] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0028] In the description of the utility model, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the view direction or positional relationship, and are only for the convenience of describing the utility model, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on the utility model.

[0029] This embodiment provides a multi-sensor fusion-based autonomous vehicle for traffic marking, such as... Figures 1-4 As shown, it includes the unmanned vehicle body 1, the line marking device, and the control system.

[0030] Specifically, the unmanned vehicle body 1 has steering wheels 2 on both sides of the middle of the chassis, and universal wheels 3 at the front and rear. The steering wheels 2 are connected to a drive device, which includes a first drive motor 5, a differential 6, and a steering shaft 7. The first drive motor 5 is located in the middle of the unmanned vehicle body 1 and is connected to the differential 6. The differential 6 is connected to the steering wheels 2 through the steering shaft 7.

[0031] The marking device includes a paint tank 4 and an air compressor 15. The paint tank 4 has a hollow internal structure with a stirring device rotatably connected inside. Two sets of spray guns 25 are connected to the bottom via a conveying hose. The stirring device includes a second drive motor 18, a stirring rod 16, and stirring blades 17. The second drive motor 18 is fixedly connected to the outer wall of the paint tank 4. The stirring rod 16 is rotatably connected to the inside of the paint tank 4 in a horizontal direction, and one end is connected to the second drive motor 18 for transmission. Stirring blades 17 are provided on the stirring rod 16. The paint tank 4 also has a heating plate 19 and a temperature sensor 20. The temperature sensor 20 is connected to the signal input terminal of the controller 8, and the switching control circuit of the heating plate 19 is connected to the signal output terminal of the controller 8. The temperature of the paint inside the paint tank 4 is detected by the temperature sensor 20, and the temperature is controlled by the heating plate 19. The temperature of the paint inside the paint tank 4 can be adjusted to meet the process and ambient temperature requirements. An air compressor 15 is installed at the rear end of the paint tank 4. The air outlet of the air compressor 15 is connected to the upper part of the paint tank 4. An exhaust pipe 27 is installed at the upper end of the paint tank 4. An air filter is installed inside the exhaust pipe 27. In order to prevent gas from flowing out of the exhaust pipe 27 when the air compressor 15 is working, a switch valve for controlling the on and off is also installed inside the exhaust pipe 27. This switch valve can be a manual switch valve or a solenoid switch valve. If a solenoid switch valve is selected, it is connected to the signal output terminal of the controller 8. After the paint in the paint tank 4 has been heated and exhausted, the controller 8 controls the solenoid switch valve to close and turns on the air compressor 15 to pressurize the paint tank 4. Under the pressure, the paint flows to the spray gun 25 through the delivery hose. Before marking, the paint is stirred and heated by a stirring device and a heating device. The harmful gases generated during the heating of the paint are filtered and discharged through the exhaust pipe 27 and the air filter, which effectively improves the environmental protection of the overall structure.

[0032] Additionally, a guide frame 21 is fixedly connected to the lower end of the unmanned vehicle body 1. Two sets of guide rods are horizontally distributed along the width direction of the unmanned vehicle body 1 on the guide frame 21. Bearings are provided at both ends of the guide rods along their length direction. The two ends of the lead screw 23 are rotatably connected to the guide frame 21 through the bearings and its length is distributed along the width direction of the unmanned vehicle body 1. A third drive motor 22 is fixedly connected to the guide frame 21. The third drive motor 22 is connected to one end of the lead screw 23. A nut 24 is threadedly connected to the lead screw 23. The main body of the spray gun 25 is fixed to the nut 24. The third drive motor 22 can drive the nut 24 and the two sets of spray guns 25 to be distributed along the width direction of the unmanned vehicle body 1. The spray gun 25 is driven by the third drive motor 22 to adjust the spraying position along the lead screw 23.

[0033] A alignment block 26 is also fixedly connected to the middle position below the guide frame 21. The alignment block 26 is used for the calibration of the spray gun 25.

[0034] The control system is as follows Figure 1As shown, specifically, a support plate 13 is fixedly connected to the upper end of the unmanned vehicle body 1. A GNSS module 9 is installed above the support plate 13. An IMU linear measurement unit 10, a camera 11, and a lidar 12 are installed at the front end. A controller 8 is also installed inside the unmanned vehicle body 1. The camera 11 is connected to the signal output terminal of the controller 8 to identify whether there are markings in the area to be marked on the ground ahead, as well as the shape and angle of the markings. The GNSS module 9 is connected to the signal input terminal of the controller 8 to provide unmanned vehicle positioning information. The lidar 12 and the IMU linear measurement unit 10 are connected to the signal input terminal of the controller to capture vehicle and surrounding environment information, including vehicle speed, acceleration, and position information. The continuous attitude and motion data provided by the IMU linear measurement unit 10 helps to update the map in real time. Long-term accumulated errors can be corrected by the positioning of the GNSS module 9. Finally, an accurate vehicle dynamics model is constructed using the MPC model predictive control algorithm to simulate and predict future behavior. Through continuous state updates and feedback corrections, constraints such as vehicle dynamics constraints and collision avoidance constraints are set to ensure safety, feasibility, and environmental adaptability, thereby realizing the dynamic adjustment and optimization of vehicle behavior.

[0035] The working principle of this device is as follows:

[0036] First, before the road markings are applied, the controller 8 opens the electromagnetic switch valve on the exhaust pipe 27 and controls the second drive motor 18 to drive the stirring rod 16 and stirring blades 17 to stir the paint. It also controls the heating plate 19 to heat the paint to prevent condensed paint from clogging the spraying pipe. The heated gas is filtered through the air filter in the exhaust pipe 27 and then discharged to the outside. After the paint tank 4 completes the gas discharge, the electromagnetic switch valve on the exhaust pipe 27 is closed. Next, the unmanned vehicle is moved to the marked starting point and the starting point is calibrated using the alignment block 26. Alternatively, the unmanned vehicle can identify the incomplete markings and then use the alignment block 26 to complete the calibration. Then, based on the surrounding environment measured by the GNSS module 9, IMU linear measurement unit 10, camera 11, and lidar 12, the controller 8 controls the unmanned vehicle body 1 to move forward or turn while controlling the third drive motor 22 to move and adjust the spray gun 25 along the guide rod. At the same time, the controller 8 controls the air compressor 15 to start and deliver paint to the spray gun 25 through the material delivery hose to complete the pattern drawing.

[0037] Additionally, it should be noted that the MPC model predictive control algorithm is existing technology. However, to provide a clearer explanation of the solution in this application, the principle of the MPC model predictive control algorithm in conjunction with the technical solution of this application will be explained in detail below:

[0038] The first step is to establish the mapping relationship between the kinematic model of the two-wheel differential speed and the sensor parameters:

[0039] Since the operating speed of autonomous vehicles is quite limited, and the vehicle structure can be approximated as a two-wheel differential model, the vehicle is modeled as a simple two-wheel differential kinematic model:

[0040] Where x and y represent the vehicle's position in the global coordinate system, with global positioning data provided by GNSS module 9 as the reference for path tracking; θ represents the vehicle's heading angle, which is measured by IMU linear measurement unit 10 and its attitude is updated in real time.

[0041] Discretized model (using Euler forward difference method):

[0042] ;

[0043] Where v is the vehicle linear velocity and ω is the vehicle angular velocity, both of which are measured by the IMU linear measurement unit 10 to provide instantaneous motion state. The control period is set to T=0.02s to ensure high-frequency state updates and real-time control of the unmanned vehicle.

[0044] The second step is to construct the linearization and prediction equations:

[0045] To simplify the calculation, the model is linearized using a first-order Taylor expansion: ;

[0046] Where A and B are coefficient matrices derived from the discrete model, reflecting the linear relationship between the state and the control quantity.

[0047] k is the discrete-time index, used to mark a specific moment in the control cycle. Camera 11 provides the datum deviation—lateral deviation. heading deviation As the desired state variable, the autonomous vehicle path tracking is achieved through a linear model;

[0048] Predictive equation extension (combined state and control variables): ;

[0049] Among them, ψ, The coefficient matrix for prediction in the time domain is recursively generated from the discretized model;

[0050] U is the control sequence, and the constraints, such as maximum speed, depend on the real-time speed feedback from the IMU.

[0051] The third step is the objective function and multi-sensor weight allocation:

[0052] The objective function is designed as a weighted sum of tracking error and control energy consumption:

[0053] ;

[0054] in:

[0055] Lateral deviation of the markings, identified by the camera, serves as the tracking target;

[0056] As a weight matrix, priority is given to ensuring horizontal accuracy (such as the fit of the marking lines).

[0057] To control the weighting of quantities, abrupt changes in speed and steering are limited;

[0058] This is a relaxation factor to avoid the optimization becoming unsolvable due to the obstacle avoidance constraints of the lidar.

[0059] After GNSS module 9 and IMU linear measurement unit 10 are fused using Kalman filtering, the update... And the camera extracts the road marking information and calculates... and Then, the lidar 12 generates a local obstacle map and constructs obstacle avoidance constraints.

[0060] Step 4: Embedding constraints and sensor data:

[0061] Dynamic constraints:

[0062] ;

[0063] in The speed is monitored in real time by the IMU linear measurement unit 10, which also provides feedback on the limit.

[0064] Obstacle avoidance constraints (linear inequality form):

[0065] ;

[0066] Implementation method: LiDAR 12 detects the distance to obstacles. Generate a safe distance Dynamic construction and .

[0067] Prediction equations generated from the discrete model Solve the quadratic programming problem: ,in , Output the optimal control sequence U* and drive the vehicle to adjust v and ω to achieve dynamic path tracking and obstacle avoidance.

[0068] The device parameters are configured as follows:

[0069] Prediction time domain Np=20: Covers the future 0.4 seconds, balancing computational load and prediction accuracy;

[0070] Control time domain Nc=10: Only optimize the control quantities in the first 10 steps to reduce the solution complexity;

[0071] Sensor frequencies: IMU linear measurement unit 10 (100Hz), lidar 12 (20Hz), camera 11 (30Hz).

[0072] Using the aforementioned prediction model, objective function, and constraints, controller 8 transforms the optimization problem into a finite-time quadratic programming (QP) problem and solves for the optimal control sequence in real time during each control cycle. Subsequently, following the rolling optimization principle of model predictive control, controller 8 actually only executes the first control command in the optimal sequence (such as the desired steering angle or wheel speed at the current moment), and then enters the next cycle, using the updated vehicle state as the initial state to predict and optimize again. This process is repeated continuously, enabling the unmanned vehicle to correct its trajectory in real time during travel, gradually approach the target path, and achieve dynamic obstacle avoidance. The above control strategy fully integrates information provided by multiple sensors, ensuring that the unmanned vehicle can still move stably and accurately along the planned trajectory under complex working conditions.

[0073] By deeply embedding the parameters of each sensor into the modeling, prediction, and optimization process of MPC, unmanned vehicles can achieve high-precision lane marking in complex environments while meeting dynamic constraints and safety obstacle avoidance requirements.

[0074] The above description is only a preferred embodiment of the present utility model. It should be noted that for those skilled in the art, several improvements can be made without departing from the principle of the present utility model, and these improvements should also be considered within the protection scope of the present utility model.

Claims

1. A multi-sensor fusion traffic marking unmanned vehicle, characterized in that, include: The unmanned vehicle body is equipped with steering wheels and a drive device for driving the unmanned vehicle body forward and turning. The marking device consists of a paint tank, a paint pump, and a spray gun mounted on the main body of the unmanned vehicle. The paint pump is connected to the paint tank and the spray gun via its inlet and outlet, respectively. The control system consists of a controller, a camera, a positioning module, an IMU linear measurement unit, and a laser ranging unit. The camera, positioning module, IMU linear measurement unit, and laser ranging unit are all connected to the signal input terminal of the controller, and the paint pump and drive device are all connected to the signal output terminal of the controller.

2. The multi-sensor fusion based traffic lane marking autonomous vehicle of claim 1, wherein, The drive unit includes a first drive motor, a differential, and a steering shaft. The first drive motor is fixedly connected to the body of the unmanned vehicle, and the first drive motor is connected to the steering wheel through the differential and the steering shaft.

3. The multi-sensor fusion traffic marking unmanned vehicle according to claim 1, characterized in that, The paint tank is equipped with a stirring assembly, which includes a second drive motor, a stirring rod, and stirring blades. The stirring rod is rotatably connected to the inside of the paint tank, and the stirring blades are fixedly connected to the surface of the stirring rod. The second drive motor is fixedly connected to the outside of the paint tank, and the output shaft of the second drive motor is drivenly connected to the stirring rod.

4. The multi-sensor fusion traffic marking unmanned vehicle according to claim 1, characterized in that, The paint tank is equipped with a heating component, which includes a heating plate disposed inside the paint tank. The paint tank is also equipped with a temperature sensor, which is connected to the signal input terminal of the controller.

5. The multi-sensor fusion traffic marking unmanned vehicle according to claim 4, characterized in that, An air vent pipe is installed above the paint tank, and an air filter and a switch valve are installed inside the air vent pipe.

6. The multi-sensor fusion traffic marking unmanned vehicle according to claim 1, characterized in that, The paint tank is also connected to an air compressor, the air outlet of which is connected to the inside of the paint tank and located above the liquid level inside the paint tank.

7. The multi-sensor fusion traffic marking unmanned vehicle according to claim 1, characterized in that, A guide frame is fixedly connected to the front of the lower part of the unmanned vehicle body. A scribing drive component is provided on the guide frame. The spray gun is connected to the scribing drive component and moves along the width direction of the vehicle through the scribing drive component.

8. The multi-sensor fusion traffic marking unmanned vehicle according to claim 7, characterized in that, The marking drive assembly includes a third drive motor, a lead screw, and a nut. The two ends of the lead screw are rotatably connected to the guide frame through bearings, and the length of the lead screw is distributed along the width direction of the vehicle. The output shaft of the third drive motor is connected to the lead screw drive. The nut is threadedly connected to the lead screw and fixedly connected to the spray gun.

9. The multi-sensor fusion traffic marking unmanned vehicle according to claim 8, characterized in that, A guide block is provided at the lower center of the guide frame.

Images