Marking robot control system based on satellite positioning and orientation technology

Abstract

The invention discloses a lineation robot control system based on a satellite positioning and orientation technology, which relates to the technical field of robot motion control and comprises a satellite positioning and orientation module, an inertial navigation module, a fusion positioning and orientation software module, a motor control software module, a motion control software module and a pattern planning software module. The satellite positioning and orientation module is used for acquiring latitude and longitude coordinates of the current position of the robot by receiving multi-frequency satellite signals; the inertial navigation module is used for collecting posture and motion data of the robot in real time through a gyroscope and an accelerometer. According to the lineation robot control system provided by the invention, through collaborative design of the multi-frequency-point GNSS receiver and the anti-multipath interference antenna array, centimeter-level continuous positioning capability in a complex urban environment is realized, and satellite signal reflection and shielding interference in scenes such as viaducts and avenues are effectively overcome; and satellite positioning and inertial navigation data are deeply fused through a tight coupling Kalman filtering algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field of robot motion control, and specifically to a control system for a line marking robot based on satellite positioning and orientation technology. Background Art

[0002] As an important part of traffic infrastructure, the construction quality of road markings directly affects road traffic safety and traffic efficiency. Traditional road marking construction mainly relies on manual operation or semi-automatic line marking equipment, which has problems such as low construction efficiency, poor pattern consistency, and insufficient accuracy in night operations. With the development of intelligent construction technology, automated line marking equipment based on machine vision or laser navigation has been gradually applied, but it still faces significant challenges in large-scale outdoor construction scenarios: the machine vision system is severely interfered by light changes and road surface stains, the performance of lidar drops sharply in rainy and foggy weather, and existing equipment generally lacks high-precision absolute positioning ability, making it difficult to meet the construction specification requirements of complex national standard patterns. Although the Global Navigation Satellite System (GNSS) has shown high-precision positioning potential in fields such as agriculture and surveying and mapping, its technical adaptability in dynamic control, multi-source data fusion, and complex path planning has not been effectively verified in the field of line marking robots.

[0003] In the prior art, although the positioning scheme based on RTK-GNSS can achieve centimeter-level static positioning, it has three major defects in the dynamic control scenario of mobile robots: First, when satellite signals are blocked by complex environments such as urban canyons and viaducts, positioning jumps are likely to occur, resulting in inaccurate path tracking; Second, the loose coupling fusion method of the inertial navigation unit (IMU) and satellite positioning data fails to effectively suppress error accumulation. Especially when the continuous satellite lock loss exceeds 20 seconds, the pose estimation error increases exponentially; Third, the existing system has insufficient geometric analysis ability for complex marking patterns and is difficult to meet the 1 cm shape tolerance requirements specified in the GB / T 16311 standard. In addition, the traditional electro-mechanical control architecture has a response lag in the coordinated control of the steering motor and the traveling motor, resulting in serrated burrs in the curve section markings, and the equipment lacks an adaptive connection mechanism, and frequent shutdown calibration is required when switching between multiple patterns, seriously affecting the construction continuity. In response to this, we propose a control system for a line marking robot based on satellite positioning and orientation technology. Summary of the Invention

[0004] To solve the above technical problems and provide a control system for a line marking robot based on satellite positioning and orientation technology, the present technical solution solves the above problems.

[0005] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] The line - drawing robot control system based on satellite positioning and orientation technology includes: a satellite positioning and orientation module, an inertial navigation module, a fusion positioning and orientation software module, a motor control software module, a motion control software module, and a pattern planning software module;

[0007] The satellite positioning and orientation module is used to obtain the longitude and latitude coordinates of the current position of the robot by receiving multi - frequency satellite signals;

[0008] The inertial navigation module is used to collect the attitude and motion data of the robot in real - time through gyroscopes and accelerometers;

[0009] The fusion positioning and orientation software module is used to fuse and process the satellite positioning data and inertial navigation data through algorithms to generate fusion positioning data;

[0010] The pattern planning software module is used to decompose complex patterns into a set of simple graphic units such as straight - line segments, arc segments, and polygon units according to the input national standard road marking pattern codes or coordinate information, and generate corresponding path planning instructions based on the fusion positioning data;

[0011] The motion control software module is used to obtain the fusion positioning data in real - time according to the path planning instructions, combine with the attitude information of the robot, and dynamically adjust the path deviation at a frequency of more than 10 times per second to generate motion control signals;

[0012] The motor control software module is used to drive the steering motor and the traveling motor of the robot chassis according to the motion control signals to control the robot to complete the marking drawing along the planned path;

[0013] During the drawing process, the connection modes for different graphic units are divided into forward connection and reverse connection. Among them, in the forward connection mode, continuous drawing is achieved through a path smoothing algorithm, and in the reverse connection mode, it is necessary to pause and recalibrate the attitude before continuing to draw to ensure that the pattern accuracy error is not greater than 1 cm.

[0014] Preferably, the satellite positioning and orientation module includes a multi - frequency GNSS receiver, an anti - multipath interference antenna array, and an RTK differential correction unit;

[0015] Among them, the multi - frequency GNSS receiver is configured to receive multi - band satellite signals simultaneously and achieve millimeter - level positioning accuracy through carrier - phase measurement;

[0016] The anti - multipath interference antenna array is composed of a four - element helical antenna group to form a spatial filtering structure, and eliminates the reflected wave interference through signal arrival angle estimation;

[0017] The RTK differential correction unit receives the reference station correction data in real - time through a 4G communication module and a radio station, and fuses and resolves the original observation values and correction data in a tightly - coupled manner, and the positioning data output frequency is not less than 10 Hz;

[0018] This module has a built-in coordinate conversion engine, which supports real-time dynamic conversion between the WGS84 coordinate system and the local coordinate system of the construction site. The conversion parameters are calibrated through ground control points to ensure that the plane positioning error is less than 8 mm.

[0019] Preferably, the inertial navigation module includes a three-axis MEMS gyroscope, a three-axis accelerometer, and a temperature compensation unit; the range of the gyroscope is ±2000° / s, and the zero-bias stability is ≤0.5° / h;

[0020] The range of the accelerometer is ±16g, and the non-linear error compensation adopts a sixth-order polynomial fitting algorithm;

[0021] The temperature compensation unit collects the temperature distribution data of the sensor through the built-in thermocouple array, and establishes a temperature-drift characteristic look-up table for real-time correction;

[0022] This module integrates a data preprocessing unit, which performs timestamp synchronization, wild value rejection, and sliding window filtering processing, and outputs attitude angle data and three-dimensional acceleration values at a frequency of 100 Hz. The attitude angle data includes: roll angle, pitch angle, and yaw angle, and attitude calculation is performed by the quaternion method to ensure that the heading angle error is less than 0.1° under dynamic conditions.

[0023] Preferably, the integrated positioning and orientation software module implements a tightly coupled Kalman filtering algorithm, and establishes a system model including a 12-dimensional state vector: 3 dimensions of position error, 3 dimensions of velocity error, 3 dimensions of attitude angle error, and 3 dimensions of gyroscope zero bias;

[0024] The observation model integrates GNSS pseudorange / carrier phase observations and inertial navigation output;

[0025] The filtering process is divided into two stages: time update and measurement update. In the time update stage, the system state is estimated based on inertial navigation data, and the Runge-Kutta fourth-order method is used to solve the differential equation;

[0026] In the measurement update stage, when the GNSS signal is valid, the satellite positioning result is used as the observation value input, and the Kalman gain is adaptively adjusted through the innovation covariance matrix;

[0027] During the period when the GNSS signal is lost, the pure inertial navigation mode is adopted and the error compensation mechanism is activated to ensure that the cumulative amount of positioning error is less than 2 cm within 30 seconds.

[0028] Preferably, the pattern planning software module includes a graphic parsing engine and a path optimizer;

[0029] The graphic parsing engine converts the dotted line, solid line, and arrow road marking patterns defined in the GB5768.3-2009 standard into a set of geometric primitives composed of straight line segments, circular arc segments, and cubic Bezier curves through a feature point extraction algorithm;

[0030] The path optimizer implements the Delaunay triangulation algorithm to establish topological relationships and uses the A* search algorithm to plan the optimal drawing order;

[0031] For curved road segments, an adaptive segmentation strategy is implemented: when the curvature radius is less than 5m, the arc length parameterization method is used to generate dense interpolation points (spacing ≤ 10cm), and the second-order geometric continuity at the junction of adjacent primitives is ensured through curvature continuity detection;

[0032] The planning result is output as an instruction set containing the position sequence, traveling speed, and steering angle, and the data format uses a dual encoding mechanism of WGS84 coordinates and the local coordinate system.

[0033] Preferably, the motion control software module constructs a three-level control architecture: the upper-level path tracking controller uses the preview distance adaptive PID algorithm to calculate the desired steering angle based on the lateral deviation and heading deviation between the current fused positioning data and the planned path;

[0034] The middle-level kinematic model calculator is based on the Ackermann steering principle and converts the path deviation into left and right wheel speed difference commands;

[0035] The lower-level dynamic compensator designs a sliding mode control law through Lyapunov stability analysis to cancel the disturbance caused by the change of the ground friction coefficient in real time;

[0036] The control period is set to 100ms, and within each control period, the following operations are performed: the Frenet coordinate system projection method is used to calculate the path deviation, the PID operation with integral separation is used to solve the control quantity, and the actuator command generates the steering motor angle pulse code;

[0037] For sharp turn road segments, a speed-curvature coupling control strategy is implemented to ensure that the maximum centripetal acceleration does not exceed 0.3g.

[0038] Preferably, the motor control software module includes a double closed-loop servo control system: the steering motor uses a position closed-loop control with an absolute encoder feedback, the resolution is 0.01°, and the response bandwidth ≥ 50Hz;

[0039] The traveling motor implements a speed-current double closed-loop control with feedforward compensation, the speed loop sampling frequency is 1kHz, and the current loop uses space vector pulse width modulation technology;

[0040] A cross-coupling controller is established between the two motors, and real-time status data is transmitted through the CAN bus. When steering lag is detected, the traveling speed is automatically reduced;

[0041] The drive unit integrates an overheat protection mechanism: when the motor winding temperature exceeds 85°C, the derating operation mode is started and the output torque is adjusted according to an exponential curve;

[0042] The synchronous controller of the spraying device dynamically adjusts the pressure of the paint pump according to the traveling speed to ensure that the marking width error ≤ ±5 mm.

[0043] Preferably, the same-direction connection mode adopts a cubic Bezier curve transition algorithm to establish a transition zone with a length three times the current line width at the connection of adjacent primitives;

[0044] The control points of the transition curve are composed of the end point P0 of the previous segment, the virtual extension point P1, the starting point P3 of the subsequent segment, and the virtual reverse point P2, where:

[0045] P1 = P0 + 0.5L·T0

[0046] In the formula, L is the marking length, and T0 is the tangent vector of the previous segment;

[0047] Among them:

[0048] P2 = P3 - 0.5L·T1

[0049] In the formula, T1 is the tangent vector of the subsequent segment;

[0050] The path smoothing algorithm calculates the transition curve parameters in real time and ensures the continuity of the curvature change rate through constraint optimization;

[0051] In the reverse connection mode, the braking controller starts the deceleration program 0.5 m away from the connection point. When the speed drops to 0.1 m / s, it triggers the full stop mechanism. The calibration process includes: verification of the validity of satellite positioning data, the number of received satellites ≥ 6, PDOP ≤ 2, inertial navigation zero-bias calibration, and laser scanning comparison of the end pose of the robotic arm. The operation can continue only after calibration is completed.

[0052] Preferably, the system includes an exception handling mechanism: when it is detected that the satellite signal is lost for more than 30 seconds and the cumulative error of the inertial navigation exceeds 3 cm, the emergency stop program is started and the acoustic and optical alarm is activated;

[0053] For the case of motor overload, the control parameters are dynamically adjusted: when the current continuously exceeds 150% of the rated value for 500 ms, it switches to the torque limit mode;

[0054] When the path tracking deviation exceeds 2 cm, the replanning mechanism is triggered: save the current state point, and backtrack to the nearest 3 path points to recalculate the local trajectory;

[0055] The system is built-in with a self-check module, which performs sensor zero calibration, communication link test, and actuator stroke verification every time it is started. The calibration data is stored in the FRAM non-volatile memory;

[0056] The abnormal event records are stored in a circular buffer, saving the last 100 status logs. The data format includes a timestamp, an exception code, and a snapshot of key parameters.

[0057] Preferably, the system implements a multi-source data fusion and calibration method: When starting for the first time every day, execute the site adaptive calibration process: Control the robot to travel along a predetermined rectangular path, and calculate the site magnetic declination compensation amount and wheel diameter calibration coefficient by comparing the satellite positioning trajectory with the inertial navigation deduced trajectory;

[0058] During the operation, perform dynamic calibration every 30 minutes: Select the current position as a temporary reference point, perform a 360° rotation scan, fit the center coordinate using the least squares method, and correct the heading angle drift error;

[0059] The calibration algorithm includes an integrity verification step: When the Euclidean distance difference between the satellite positioning data and the inertial navigation deduced position exceeds 2 cm for 5 consecutive seconds, it is determined as an abnormal state and triggers the sensor cross-check process. The data sources for cross-check include the analysis of the gravity field component of the accelerometer and the consistency test of the gyroscope angular velocity integration.

[0060] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0061] The line marking robot control system proposed by the present invention realizes centimeter-level continuous positioning ability in complex urban environments through the collaborative design of a multi-frequency GNSS receiver and an anti-multipath interference antenna array, effectively overcoming satellite signal reflection and occlusion interference in scenarios such as viaducts and tree-lined roads; through the tightly coupled Kalman filtering algorithm, deeply fuses satellite positioning and inertial navigation data, maintains high-precision pose deduction in scenarios with intermittent satellite signals, and ensures the continuous construction stability in fully enclosed spaces such as tunnels and underground garages; through geometric primitive intelligent parsing and adaptive path planning technology, accurately decomposes the geometric features of national standard line marking patterns, meeting the construction specification requirements for complex patterns such as solid and dashed line combinations and three-dimensional deceleration markings; through a three-level control architecture and a sliding mode dynamic compensation mechanism, improves the path tracking accuracy of the robot under high-speed movement, avoiding paint splashing and line marking deformation caused by centrifugal force in sharp turn sections; through the Bezier curve transition algorithm and the hierarchical braking calibration strategy, realizes the smooth transition of curvature continuity at the joints of different graphic units, greatly shortening the downtime calibration time during multi-pattern switching; through multi-source sensor cross-check and adaptive environment calibration process, ensures the operation reliability of the system under harsh working conditions such as extreme temperatures and electromagnetic interference, reducing the risk of construction interruption caused by equipment anomalies; through electromechanical collaborative control and paint pressure dynamic adjustment technology, synchronously optimizes the response speed of the steering motor and the paint spraying uniformity, improving the edge clarity and width consistency of the line marking. Description of the Drawings

[0062] Figure 1 It is a module framework diagram of the line marking robot control system based on satellite positioning and orientation technology. Detailed Embodiments

[0063] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments in the following description are only examples, and those skilled in the art can think of other obvious variations.

[0064] Referring to Figure 1 As shown, the line marking robot control system based on satellite positioning and orientation technology includes: a satellite positioning and orientation module, an inertial navigation module, a fusion positioning and orientation software module, a motor control software module, a motion control software module, and a pattern planning software module;

[0065] The satellite positioning and orientation module is used to obtain the longitude and latitude coordinates of the current position of the robot by receiving multi-frequency satellite signals;

[0066] The inertial navigation module is used to collect the attitude and motion data of the robot in real time through a gyroscope and an accelerometer;

[0067] The fusion positioning and orientation software module is used to fuse and process the satellite positioning data and inertial navigation data through an algorithm to generate fusion positioning data;

[0068] The pattern planning software module is used to decompose the complex pattern into a set of simple graphic units such as straight line segments, arc segments, and polygon units according to the input national standard road marking pattern code or coordinate information, and generate corresponding path planning instructions based on the fusion positioning data;

[0069] The motion control software module is used to obtain the fusion positioning data in real time according to the path planning instructions, combine the attitude information of the robot, and dynamically adjust the path deviation at a frequency of more than 10 times per second to generate a motion control signal;

[0070] The motor control software module is used to drive the steering motor and the traveling motor of the robot chassis according to the motion control signal to control the robot to complete the line marking along the planned path;

[0071] During the drawing process, the connection modes for different graphic units are divided into forward connection and reverse connection. Among them, in the forward connection mode, continuous drawing is realized through the path smoothing algorithm. In the reverse connection mode, it is necessary to pause and recalibrate the attitude before continuing to draw to ensure that the pattern accuracy error is not greater than 1 cm.

[0072] The specific implementation process is as follows: After the robot is started, it first performs initialization self-check. It obtains the initial longitude and latitude coordinates through the multi-frequency GNSS receiver of the satellite positioning and orientation module. The receiver is equipped with a four-element helix antenna array to perform spatial filtering on multi-band signals to eliminate the interference of multipath effects. At the same time, it receives the RTK differential correction data of the reference station through the 4G communication module and radio communication, and uses the carrier phase dynamic differential technology to solve the original satellite observation values into a positioning result with centimeter-level accuracy. And through the coordinate conversion engine, the WGS84 coordinate system is converted into the local coordinate system of the construction site. The conversion parameters are pre-calibrated through at least three control points arranged on the ground to ensure that the plane positioning error is controlled within 8 mm; the three-axis MEMS gyroscope and accelerometer in the inertial navigation module collect attitude data at a frequency of 100 Hz. The temperature compensation unit monitors the temperature distribution state of the sensor in real time, compensates and corrects the original data according to the pre-stored thermal drift characteristic look-up table, and uses the quaternion method to solve the current roll angle, pitch angle and heading angle. Under dynamic conditions, the heading angle error is less than 0.1°; the integrated positioning and orientation software module runs the tightly coupled Kalman filter algorithm to deeply integrate the satellite positioning data and inertial navigation data at the observation layer. The system state vector includes 12 dimensions of position, velocity, attitude error and sensor zero bias. When the GNSS signal is stable, it is measured and updated at a frequency of 10 Hz. When the satellite signal is interrupted due to tree occlusion, it automatically switches to the pure inertial navigation mode and activates the error compensation mechanism to suppress position drift through kinematic constraints to ensure that the cumulative error does not exceed 2 cm within 30 seconds.

[0073] After the pattern planning software module receives the marking design file in the standard format of GB / T 16311, the graphic parsing engine decomposes the complex pattern into straight line segments, arc segments and cubic Bezier curve units through the feature point extraction algorithm. The path optimizer establishes a topological relationship based on Delaunay triangulation, and uses the improved A* algorithm to plan the optimal drawing path. It performs adaptive segmentation processing on the curve area with a curvature radius less than 5 m, generates an instruction set including coordinate sequences, traveling speeds and steering angles, and ensures data compatibility through the dual coding mechanism of WGS84 coordinates and local coordinates; the motion control software module constructs a three-level control architecture. The upper-layer path tracking controller uses the preview distance adaptive PID algorithm to dynamically adjust the preview distance according to the lateral deviation between the integrated positioning data and the planned path, and calculates the expected steering angle in the Frenet coordinate system. The middle-layer kinematic model calculator converts the path deviation into left and right wheel speed difference instructions based on the Ackerman steering principle. The lower-layer dynamic compensator uses the sliding mode control algorithm to cancel the ground friction disturbance in real time. The control period is set to 100 ms, and the speed-curvature coupling strategy is automatically started on sharp curve sections to limit the centripetal acceleration not to exceed 0.3g.

[0074] The motor control software module synchronously controls the steering motor and the traveling motor through the CAN bus protocol. The steering motor uses an absolute encoder to construct a position closed-loop to achieve precise steering with a resolution of 0.01°. The traveling motor implements a double closed-loop control with feedforward compensation, and the sampling frequency of the speed loop reaches 1 kHz. A cross-coupling controller is established between the two motors. When steering lag is detected, the traveling speed is automatically reduced and the control parameters are readjusted. For the connection requirements of different graphic units, in the same-direction connection mode, a cubic Bézier curve transition algorithm is adopted to generate a smooth transition area with a length three times the line width at the connection of adjacent graphic elements, and the curvature continuity is ensured by virtual extension of the control points. In the reverse connection mode, a hierarchical braking program is started at a distance of 0.5 m from the connection point. After the robot comes to a complete stop, a multi-source calibration process is executed, including verification of satellite positioning effectiveness, reset of inertial navigation zero bias, and comparison of laser scanning pose. The operation can continue only after the calibration is completed. The system is built-in with an exception handling mechanism to monitor the operating status of each module in real time. When satellite lock loss timeout or path deviation exceeds the limit, a hierarchical response strategy is triggered, including local path replanning, derating operation of the drive system, and emergency stop protection. At the same time, the abnormal event data is recorded through the FRAM memory to provide complete log support for subsequent fault diagnosis.

[0075] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification is only the principle of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A marking robot control system based on satellite positioning and orientation technology, characterized in that: include: Satellite positioning and orientation module, inertial navigation module, fusion positioning and orientation software module, motor control software module, motion control software module and pattern planning software module; The satellite positioning and orientation module is used to obtain the longitude and latitude coordinates of the robot's current position by receiving multi-frequency satellite signals; The inertial navigation module is used to collect the robot's attitude and motion data in real time through gyroscopes and accelerometers; The fusion positioning and orientation software module is used to fuse satellite positioning data and inertial navigation data through algorithms to generate fusion positioning data; The pattern planning software module is used to decompose the complex pattern into a simple graphic unit set of straight line segments, arc segments and polygonal units according to the input national standard road marking pattern code or coordinate information, and generate corresponding path planning instructions based on the fused positioning data; The motion control software module is used to obtain fused positioning data in real time according to the path planning instructions, and dynamically adjust the path deviation at a frequency of more than 10 times per second in combination with the robot's posture information to generate motion control signals; The motor control software module is used to drive the steering motor and travel motor of the robot chassis according to the motion control signal, and control the robot to complete the marking line drawing along the planned path; During the drawing process, the connection modes for different graphic units are divided into same-direction connection and reverse connection. In the same-direction connection mode, continuous drawing is achieved through a path smoothing algorithm. In the reverse connection mode, the drawing needs to be paused and the posture needs to be recalibrated before continuing to ensure that the pattern accuracy error is no more than 1cm.

2. The marking robot control system based on satellite positioning and orientation technology according to claim 1 is characterized in that: The satellite positioning and orientation module includes a multi-frequency GNSS receiver, an anti-multipath interference antenna array and an RTK differential correction unit; Among them, the multi-frequency GNSS receiver is configured to simultaneously receive multi-band satellite signals and achieve millimeter-level positioning accuracy through carrier phase measurement; The anti-multipath interference antenna array uses a quad-helix antenna group to form a spatial filtering structure, and eliminates reflected wave interference through signal arrival angle estimation; The RTK differential correction unit receives the base station correction data in real time through the 4G communication module and radio communication, and fuses the original observation value with the correction data in a tightly coupled manner. The positioning data output frequency is not less than 10Hz. The module has a built-in coordinate conversion engine that supports real-time dynamic conversion between the WGS84 coordinate system and the local coordinate system of the construction site. The conversion parameters are calibrated through ground control points to ensure that the plane positioning error is less than 8mm.

3. The marking robot control system based on satellite positioning and orientation technology according to claim 1 is characterized in that: The inertial navigation module includes a three-axis MEMS gyroscope, a three-axis accelerometer and a temperature compensation unit; the gyroscope range is ±2000° / s, and the zero bias stability is ≤0.5° / h; The accelerometer range is ±16g, and the nonlinear error compensation uses a sixth-order polynomial fitting algorithm; The temperature compensation unit collects the sensor temperature distribution data through the built-in thermocouple array and establishes a temperature-drift characteristic lookup table for real-time correction; The module integrates a data preprocessing unit, performs timestamp synchronization, outlier removal and sliding window filtering, and outputs attitude angle data and three-dimensional acceleration values ​​at a frequency of 100Hz. The attitude angle data includes: roll angle, pitch angle, yaw angle, and attitude calculation is performed through the quaternion method to ensure that the heading angle error is less than 0.1° under dynamic conditions.

4. The marking robot control system based on satellite positioning and orientation technology according to claim 1 is characterized in that: The fusion positioning and orientation software module implements a tightly coupled Kalman filter algorithm to establish a system model containing a 12-dimensional state vector: 3 dimensions of position error, 3 dimensions of velocity error, 3 dimensions of attitude angle error, and 3 dimensions of gyroscope zero bias; The observation model integrates GNSS pseudorange / carrier phase observations and inertial navigation outputs; The filtering process is divided into two stages: time update and measurement update. In the time update stage, the system state is estimated based on the inertial navigation data, and the Runge-Kutta fourth-order method is used to solve the differential equation. In the measurement update phase, when the GNSS signal is valid, the satellite positioning result is input as the observation value, and the Kalman gain is adaptively adjusted through the new information covariance matrix; During the period when the GNSS signal is lost, the pure inertial navigation mode is adopted and the error compensation mechanism is activated to ensure that the accumulated positioning error is less than 2cm within 30 seconds.

5. The marking robot control system based on satellite positioning and orientation technology according to claim 1 is characterized in that: The pattern planning software module includes a graphics parsing engine and a path optimizer; The graphics parsing engine converts the dotted line, solid line, and arrow road marking patterns defined in the GB5768.3-2009 standard into a set of geometric primitives consisting of straight line segments, circular arc segments, and cubic Bezier curves through a feature point extraction algorithm; The path optimizer implements the Delaunay triangulation algorithm to establish topological relationships and uses the A* search algorithm to plan the optimal drawing order; For curved sections, an adaptive segmentation strategy is implemented: when the radius of curvature is less than 5m, the arc length parameterization method is used to generate dense interpolation points (spacing ≤ 10cm), and the curvature continuity test is used to ensure the second-order geometric continuity of the adjacent primitives. The planning result is output as an instruction set including position sequence, travel speed, and steering angle. The data format adopts a dual encoding mechanism of WGS84 coordinates and local coordinate system.

6. The marking robot control system based on satellite positioning and orientation technology according to claim 1 is characterized in that: The motion control software module builds a three-level control architecture: the upper-level path tracking controller uses a preview distance adaptive PID algorithm to calculate the desired steering angle based on the lateral deviation and heading deviation of the current fused positioning data and the planned path; The middle-level kinematic model calculator converts the path deviation into the left and right wheel speed difference command based on the Ackerman steering principle; The underlying dynamic compensator designs the sliding mode control law through Lyapunov stability analysis to offset the disturbance caused by the change of ground friction coefficient in real time; The control cycle is set to 100ms, and in each control cycle, the following are executed: the path deviation is calculated using the Frenet coordinate system projection method, the control quantity is solved by the PID operation with integral separation, and the steering motor angle pulse code is generated by the actuator instruction; For sharp bends, a speed-curvature coupling control strategy is implemented to ensure that the maximum centripetal acceleration does not exceed 0.3g.

7. The marking robot control system based on satellite positioning and orientation technology according to claim 1 is characterized in that: The motor control software module includes a dual closed-loop servo control system: the steering motor uses position closed-loop control with absolute encoder feedback, with a resolution of 0.01° and a response bandwidth of ≥50Hz; The traveling motor implements speed-current dual closed-loop control with feedforward compensation, the speed loop sampling frequency is 1kHz, and the current loop adopts space vector pulse width modulation technology; A cross-coupling controller is established between the two motors, which transmits real-time status data via the CAN bus and automatically reduces the travel speed when steering lag is detected; The drive unit integrates an overheat protection mechanism: when the motor winding temperature exceeds 85°C, the derating operation mode is activated and the output torque is adjusted according to the exponential curve; The synchronous controller of the spraying device dynamically adjusts the paint pump pressure according to the travel speed to ensure that the marking line width error is ≤±5mm.

8. The marking robot control system based on satellite positioning and orientation technology according to claim 1 is characterized in that: The same-direction connection mode uses a cubic Bezier curve transition algorithm to establish a transition zone with a length of three times the current line width at the connection between adjacent primitives; The control points of the transition curve are composed of the front end point P0, the virtual extension point P1, the back start point P3 and the virtual reverse point P2, where: P1=P0+0.5L·T0 In the formula, L is the length of the marking line, T0 is the tangent vector of the front segment; in: P2=P3-0.5L·T1 Where, T1 is the tangent vector of the rear segment; The path smoothing algorithm calculates the transition curve parameters in real time and ensures the continuity of the curvature change rate through constraint optimization; In the reverse connection mode, the brake controller starts the deceleration program at 0.5m away from the connection point, and triggers the full stop mechanism when the speed drops to 0.1m / s. The calibration process includes: verification of the validity of satellite positioning data, number of received satellites ≥6, PDOP≤2, inertial navigation zero bias calibration, and laser scanning comparison of the end position of the robotic arm. Operation can only be continued after the calibration is completed.

9. The marking robot control system based on satellite positioning and orientation technology according to claim 1, characterized in that: The system includes an exception handling mechanism: when it detects that the satellite signal is out of lock for more than 30 seconds and the cumulative error of the inertial navigation exceeds 3cm, the emergency shutdown procedure is initiated and the sound and light alarm is activated; Dynamically adjust control parameters for motor overload: when the current exceeds 150% of the rated value for 500ms, switch to torque limit mode; When the path tracking deviation exceeds 2cm, the re-planning mechanism is triggered: the current state point is saved, and the last three path points are traced back to recalculate the local trajectory; The system has a built-in self-check module, which performs sensor zero point calibration, communication link test and actuator stroke verification every time it is started, and the calibration data is stored in FRAM non-volatile memory; Abnormal event records are stored in a circular buffer, saving the most recent 100 status logs. The data format includes timestamp, abnormal code and key parameter snapshots.

10. The marking robot control system based on satellite positioning and orientation technology according to claim 1, characterized in that: The system implements a multi-source data fusion calibration method: at the first start of each day, the site adaptive calibration process is executed: the robot is controlled to move along a predetermined rectangular path, and the site magnetic declination compensation and wheel diameter calibration coefficient are calculated by comparing the satellite positioning trajectory with the inertial navigation trajectory; During the operation, dynamic calibration is performed every 30 minutes: the current position is selected as a temporary reference point, a 360° rotation scan is carried out, the coordinates of the circle center are fitted using the least squares method, and the heading angle drift error is corrected; The calibration algorithm includes an integrity verification step: when the Euclidean distance difference between the satellite positioning data and the inertial navigation calculated position exceeds 2 cm for 5 seconds, it is judged as an abnormal state and the sensor cross-verification process is triggered. The verification data source includes accelerometer gravity field component analysis and gyroscope angular velocity integral consistency check.

Images