Straight driving calibration method, device, equipment and storage medium

Abstract

The application discloses a kind of linear travel calibration method, device, equipment and storage medium of intelligent mobile device.Intelligent mobile device includes intelligent driving system and steering control system.The method comprises: intelligent driving system determines target straight path based on environmental perception information and controls device to travel along target straight path;Intelligent driving system determines whether device meets target straight path tracking condition according to actual motion trajectory information;When meeting, straight calibration trigger information is sent to steering control system;Steering control system acquires the current position data of steering execution mechanism in corresponding calibration time window, and determines zero deviation parameter accordingly, and updates the straight travel reference of steering execution mechanism.The application utilizes the target straight tracking state confirmed by intelligent driving system to trigger the self-calibration of steering control system, which can reduce manual debugging and improve the consistency of straight travel reference and actual travel direction.

Description

Technical Field

[0001] This invention relates to the field of driving control and calibration technology for intelligent mobile devices, specifically to a method, apparatus, device, and storage medium for calibrating the straight-line driving of an intelligent mobile device. Background Technology

[0002] Intelligent mobile devices such as smart wheelchairs, smart mobility scooters, mobile robots, nursing transport equipment, and logistics handling equipment are typically equipped with steering mechanisms to adjust the orientation of the front wheels, steering wheels, steering wheels, or other walking wheels. To achieve autonomous navigation, assisted driving, path tracking, or remote control operation, intelligent mobile devices can also be equipped with intelligent driving systems. These systems can determine the target path based on information such as vision, radar, positioning, maps, odometer readings, and inertial measurement, and control the device to move along that path.

[0003] In actual products, there is often a deviation between the theoretical zero position of the steering actuator and the actual straight-line driving direction of the equipment. This deviation may originate from assembly errors of the steering motor or servo motor, encoder installation errors, wheel assembly angle deviations, mechanical transmission clearances, tire wear, structural deformation, and assembly differences after repairing or replacing parts. When the steering control system still uses the theoretical zero position as the straight-line driving reference, even if the control command is to go straight, the equipment may still experience problems such as veering, slow yaw, and increased straight-line tracking errors.

[0004] Traditional methods typically require manual adjustment of the mechanical structure or manual input of compensation parameters during the factory, repair, or commissioning stages. For example, commissioning personnel drive the vehicle a distance on the ground to observe whether it veers off course, and then repeatedly adjust the steering zero point or mechanical connections. This method is inefficient, relies on operational experience, and is difficult to correct new deviations caused by wear, collisions, or maintenance after long-term use of the equipment.

[0005] Existing technologies also include some automatic vehicle steering zero-point calibration or zero-bias compensation schemes. These schemes typically determine whether the vehicle is traveling in a straight line based on the vehicle's own steering wheel angle, wheel speed difference, yaw rate change, yaw rate, or steering motor feedback signals, and update the steering zero point accordingly. While these schemes have some applicability to vehicle chassis or specific steering systems, their judgment logic usually relies directly on the data loop of the steering system or vehicle motion sensors, failing to fully utilize the judgment results of the intelligent driving systems already present in smart mobile devices regarding the target path and actual trajectory.

[0006] For intelligent mobile devices equipped with intelligent driving systems, these systems typically generate a target straight path based on environmental perception information and determine whether the device has stably tracked this path based on the actual motion trajectory. This determination reflects whether the device is truly traveling along the target straight line from both environmental and trajectory perspectives, and is not entirely identical to the encoder zero position or servo motor neutral position of the steering actuator itself. If the target straight path tracking status confirmed by the intelligent driving system can be used to trigger zero-position calibration of the steering control system, then the path determination capability of the intelligent driving system can be transformed into a calibration basis for the steering actuator.

[0007] Therefore, it is necessary to provide a straight-line driving calibration scheme suitable for intelligent mobile devices. This scheme allows the intelligent driving system to provide calibration trigger information to the steering control system when it confirms that the device meets the target straight-line path tracking conditions. The steering control system then collects the current position data of the steering actuator within the corresponding calibration time window and determines the zero-position deviation parameter accordingly, thereby updating the straight-line driving reference. This scheme reduces manual adjustments, improves the consistency between the straight-line driving reference and the actual driving state of the device, and helps improve the path tracking accuracy and reliability of intelligent mobile devices. Summary of the Invention

[0008] The purpose of this invention is to provide a method, apparatus, device, and storage medium for calibrating straight-line driving of intelligent mobile devices, in order to solve the problems that the theoretical zero position of the steering actuator of existing intelligent mobile devices is inconsistent with the actual straight-line driving direction due to assembly errors, transmission errors, wear, or maintenance and replacement, which in turn causes straight-line driving deviation, increased path tracking error, and low efficiency of manual calibration.

[0009] To achieve the above objectives, this invention provides a straight-line driving calibration method for an intelligent mobile device. The intelligent mobile device includes an intelligent driving system and a steering control system, the latter controlling a steering actuator. The method includes: the intelligent driving system determining a target straight-line path based on environmental perception information and controlling the intelligent mobile device to travel along the target straight-line path; the intelligent driving system acquiring actual motion trajectory information of the intelligent mobile device during its travel along the target straight-line path and determining whether the intelligent mobile device meets the target straight-line path tracking conditions based on the actual motion trajectory information; when the intelligent mobile device meets the target straight-line path tracking conditions, the intelligent driving system sending a straight-line calibration trigger message to the steering control system; the steering control system responding to the straight-line calibration trigger message and collecting current position data of the steering actuator within a calibration time window corresponding to the target straight-line path tracking conditions; the steering control system determining a zero-position deviation parameter of the steering actuator relative to the straight-line driving direction based on the current position data collected within the calibration time window; and updating the straight-line driving reference of the steering actuator based on the zero-position deviation parameter.

[0010] This invention does not simply collect steering angles and update zero positions during device operation. Instead, it utilizes the target straight path generation and actual trajectory tracking capabilities already possessed by intelligent driving systems. When the intelligent driving system confirms that the device is stably traveling along the target straight path, this state can serve as an external basis for "real straight-line driving." The steering control system then collects the current position data of the steering actuator within the corresponding calibration time window, thereby deducing the current zero-position deviation of the steering actuator relative to the real straight-line driving direction.

[0011] In some embodiments, environmental perception information may include at least one of visual image information, depth image information, lidar information, millimeter-wave radar information, ultrasonic information, positioning information, road boundary information, wall boundary information, and channel boundary information. The target straight path may be generated by the intelligent driving system based on environmental boundaries, channel direction, target point location, map information, or navigation tasks.

[0012] In some embodiments, the actual motion trajectory information may include at least one of the following: position change information of the smart mobile device, heading angle change information, odometer information, inertial measurement information, wheel speed information, and trajectory curvature information. Target straight path tracking conditions may include: the lateral deviation of the actual motion trajectory relative to the target straight path being less than a first preset threshold; the trajectory curvature of the actual motion trajectory being less than a second preset threshold; the heading angle change being less than a third preset threshold; and the duration or distance traveled along the target straight path reaching a preset requirement.

[0013] In some embodiments, the straight-line calibration trigger information may include at least one of the following: target straight-line path tracking success information, calibration start information, calibration end information, straight-line driving confidence level, target straight-line path information, and calibration time window information. The calibration time window can be the time interval during which the intelligent driving system determines that the intelligent mobile device continuously meets the target straight-line path tracking conditions, or it can be the sampling time period determined by the steering control system based on the straight-line calibration trigger information.

[0014] In some embodiments, the steering control system collects multiple current position data within a calibration time window and performs averaging, weighted averaging, filtering, outlier removal, or stability assessment on the multiple current position data to obtain a zero-position deviation parameter. The current position data may include at least one of the following: steering motor encoder position, servo motor position, steering angle sensor information, front wheel steering angle information, and steering actuator feedback position. The zero-position deviation parameter may include at least one of the following: steering zero-position offset, encoder zero-point compensation value, servo motor center position compensation value, and front wheel straight-line angle compensation value.

[0015] In some embodiments, before the steering control system determines the zero-position deviation parameter, it can also determine whether the intelligent mobile device meets the calibration validity conditions. For example, the driving speed is within a preset speed range, no obstacle avoidance steering command has been received, it is not in a state of rapid acceleration, it is not in a state of rapid deceleration, and the fluctuation of the current position data is less than a preset threshold. By using the calibration validity conditions, unreliable sampling caused by temporary obstacle avoidance, sudden stop, ground disturbance, or sensor malfunction can be eliminated.

[0016] In some embodiments, after updating the straight-line driving reference, the intelligent driving system can control the intelligent mobile device to drive along the target straight path again, and determine whether the straight-line driving deviation meets the preset calibration accuracy requirements based on the updated actual motion trajectory information. If the preset calibration accuracy requirements are not met, the zero-position deviation parameter can be redefined, or the straight-line driving reference before the update can be restored.

[0017] The present invention also provides a straight-line driving calibration device for a smart mobile device, a smart mobile device, and a computer-readable storage medium. The device, apparatus, and storage medium are capable of performing the above-described straight-line driving calibration method.

[0018] Compared with existing technologies, the present invention has at least the following beneficial effects: First, by using the target straight-line path tracking status confirmed by the intelligent driving system as the calibration basis, the actual straight-line driving state of the equipment can be more accurately characterized; Second, by sending straight-line calibration trigger information to the steering control system through the intelligent driving system, collaborative calibration between the intelligent driving system and the steering control system can be achieved; Third, the steering control system collects the current position data of the steering actuator within the corresponding calibration time window, which can reflect the actual position of the steering actuator under the actual straight-line driving state; Fourth, by updating the straight-line driving reference through the zero-position deviation parameter, manual calibration can be reduced, and the consistency of straight-line driving after long-term use of the equipment can be improved; Fifth, this solution is particularly suitable for intelligent mobile devices such as intelligent wheelchairs, intelligent mobility scooters, and mobile robots that have intelligent driving systems and independent steering control systems. Attached Figure Description

[0019] Figure 1 A schematic diagram of the system structure of a smart mobile device provided in an embodiment of the present invention.

[0020] Figure 2 This is a flowchart illustrating the straight-line driving calibration method provided in an embodiment of the present invention.

[0021] Figure 3 This is a schematic diagram illustrating the relationship between the target straight path and the actual motion trajectory in an embodiment of the present invention.

[0022] Figure 4 This is a schematic diagram illustrating the relationship between the calibration time window and the current location data acquisition in an embodiment of the present invention.

[0023] Figure 5 This is a schematic diagram of the straight-line driving calibration device provided in an embodiment of the present invention.

[0024] Labeling Explanation: 100, Intelligent Mobile Device; 110, Intelligent Driving System; 120, Steering Control System; 130, Steering Actuator; 140, Memory; 150, Motion Detection Component; 160, Environmental Perception Component; L, Target Straight Path; T, Actual Motion Trajectory; W, Calibration Time Window. Detailed Implementation

[0025] The embodiments of the present invention will now be described with reference to the accompanying drawings. It should be understood that the following embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. Where there is no conflict, the technical features in the following embodiments can be combined with each other.

[0026] like Figure 1 As shown, this embodiment provides an intelligent mobile device 100. The intelligent mobile device 100 can be an intelligent wheelchair, intelligent mobility scooter, mobile robot, nursing transport equipment, logistics handling equipment, etc. The intelligent mobile device 100 includes an intelligent driving system 110, a steering control system 120, a steering actuator 130, a memory 140, and a motion detection component 150.

[0027] The intelligent driving system 110 is used to perform path planning, environmental perception, navigation control, or assisted driving control. The intelligent driving system 110 may include a processor, a perception sensor interface, a positioning module, a map module, a path planning module, and a trajectory tracking module. The intelligent driving system 110 can determine a target straight path L based on environmental perception information and control the intelligent mobile device 100 to travel along the target straight path L.

[0028] The steering control system 120 controls the steering actuator 130. The steering actuator 130 may include a steering motor, a servo motor, a steering wheel module, a front wheel steering mechanism, a steering wheel mechanism, or other actuators capable of changing the orientation of the wheels. The steering control system 120 can control the steering actuator 130 to operate according to the target steering angle, target wheel orientation, or path tracking control commands from the intelligent driving system 110.

[0029] In some products, the intelligent driving system 110 and the steering control system 120 can be deployed in different controllers, connected via CAN bus, serial port, Ethernet, LIN bus, wireless communication, or other communication methods. In other products, the intelligent driving system 110 and the steering control system 120 can also be deployed in different software modules of the same controller. This invention is applicable as long as the intelligent driving system 110 can provide straight-line calibration trigger information to the steering control system 120.

[0030] like Figure 2 As shown, the straight-line driving calibration method in this embodiment includes the following steps.

[0031] In step S101, the intelligent driving system 110 determines the target straight path L based on environmental perception information and controls the intelligent mobile device 100 to drive along the target straight path L.

[0032] Environmental perception information can come from cameras, depth cameras, LiDAR, millimeter-wave radar, ultrasonic sensors, positioning modules, map modules, etc. This information can include visual image information, depth image information, LiDAR information, millimeter-wave radar information, ultrasonic information, positioning information, road boundary information, wall boundary information, and passage boundary information. The intelligent driving system 110 can determine a target straight-line path L based on environmental boundaries, passage direction, target point location, map information, or navigation tasks.

[0033] For example, in an indoor corridor scenario, the intelligent driving system 110 can determine the direction of corridor extension based on the boundaries of the side walls or the passageway, and generate a target straight path L extending along the corridor direction. In an intelligent wheelchair scenario, the target straight path L can be a straight path from the current position to the target point. In a mobile robot scenario, the target straight path L can also be a straight line segment in a navigation task.

[0034] In step S102, the intelligent driving system 110 obtains the actual motion trajectory information of the intelligent mobile device 100 during its journey along the target straight path L, and determines whether the intelligent mobile device 100 meets the target straight path tracking conditions based on the actual motion trajectory information.

[0035] Actual motion trajectory information can be provided by the motion detection component 150, positioning module, odometer, inertial measurement unit, wheel speed sensor, visual odometer, or radar positioning module. This actual motion trajectory information may include position change information, heading angle change information, odometer information, inertial measurement information, wheel speed information, trajectory curvature information, etc.

[0036] like Figure 3 As shown, the intelligent driving system 110 can compare the target straight path L with the actual trajectory T. When the lateral deviation of the actual trajectory T relative to the target straight path L is less than a first preset threshold, the trajectory curvature of the actual trajectory T is less than a second preset threshold, the change in the heading angle of the intelligent mobile device 100 is less than a third preset threshold, and the duration or distance traveled along the target straight path L reaches a preset requirement, it can be determined that the intelligent mobile device 100 meets the target straight path tracking conditions.

[0037] The target straight-line path tracking condition does not require the actual motion trajectory T to completely coincide with the target straight-line path L, but rather to confirm that the intelligent mobile device 100 has been stably traveling along the target straight-line path L. In other words, this condition is used to confirm from the path tracking perspective of the intelligent driving system 110 that the device is in a reliable actual straight-line driving state.

[0038] In step S103, when the intelligent mobile device 100 meets the target straight path tracking conditions, the intelligent driving system 110 sends straight calibration trigger information to the steering control system 120.

[0039] The straight-line calibration trigger information may include at least one of the following: target straight-line path tracking success information, calibration start information, calibration end information, straight-line driving confidence level, target straight-line path information, and calibration time window information. For example, when the intelligent driving system 110 determines that the device has continuously met the target straight-line path tracking conditions for a preset time, it can send calibration start information to the steering control system 120; when the conditions are no longer met or the sampling time ends, it can send calibration end information.

[0040] In some embodiments, the straight-line calibration trigger information can directly carry the start and end times of the calibration time window W. In other embodiments, the steering control system 120 can also determine the calibration time window W based on the time when the calibration start information and calibration end information are received.

[0041] In step S104, the steering control system 120 responds to the straight line calibration trigger information and collects the current position data of the steering actuator 130 within the calibration time window W corresponding to the target straight line path tracking conditions.

[0042] like Figure 4 As shown, the calibration time window W is the time interval during which the intelligent driving system 110 determines that the intelligent mobile device 100 continuously meets the target straight path tracking conditions. Since this time interval corresponds to the state in which the device is stably traveling along the target straight path L, the current position data of the steering actuator 130 collected within this time interval can reflect the actual position of the steering actuator 130 when the device is actually traveling in a straight line.

[0043] Current position data may include the position of the steering motor encoder, the position of the servo motor, information from the steering angle sensor, the front wheel angle information, and the feedback position of the steering actuator. The steering control system 120 can collect multiple current position data at a preset sampling frequency within a calibrated time window W.

[0044] In step S105, the steering control system 120 determines the zero-position deviation parameter of the steering actuator 130 relative to the straight driving direction based on the current position data collected within the calibration time window W.

[0045] Specifically, the steering control system 120 can perform averaging, weighted averaging, filtering, outlier removal, or stability assessment on multiple current position data to obtain zero-point deviation parameters. For example, when the encoder position corresponding to the theoretical straight-line reference is P0, and the average value of multiple encoder positions within the calibration time window W is P1, the difference between P1 and P0 can be determined as the encoder zero-point compensation value. Alternatively, P1 can be directly used as the new straight-line driving reference position, or the servo center position compensation value, front wheel straight-line angle compensation value, etc., can be calculated based on P1.

[0046] Before determining the zero-point deviation parameter, it is also necessary to determine whether the intelligent mobile device 100 meets the calibration validity conditions. Valid calibration conditions may include the driving speed being within a preset speed range, no obstacle avoidance steering command being received, not being in a state of rapid acceleration, not being in a state of rapid deceleration, and the fluctuation of the current position data being less than a preset threshold. By using valid calibration conditions, incorrect calibration can be avoided when the device is performing obstacle avoidance, emergency stop, skidding, collision, or abnormal trajectory.

[0047] Step S106: Update the straight-line driving reference of the steering actuator 130 according to the zero-position deviation parameter.

[0048] The steering control system 120 can write the zero-position deviation parameter into its storage unit, parameter table, configuration file, or non-volatile storage area. In subsequent straight-line driving control, the steering control system 120 compensates for the target position of the steering actuator 130 based on the zero-position deviation parameter, so that the actual output of the steering actuator 130 is closer to the actual straight-line driving direction of the equipment.

[0049] For example, when the control system subsequently receives a straight-line command, the steering control system 120 no longer simply controls the steering actuator 130 to return to the theoretical zero position, but instead controls it to return to the straight-line driving reference position corrected according to the zero-position deviation parameter. This can compensate for deviation caused by assembly errors, encoder zero-point errors, servo motor center position errors, or wheel assembly installation angle errors.

[0050] In some embodiments, after updating the straight-line driving reference of the steering actuator 130, a verification step can also be performed. The intelligent driving system 110 controls the intelligent mobile device 100 to drive along the target straight path L again, obtains the updated actual motion trajectory information, and determines whether the straight-line driving deviation of the intelligent mobile device 100 meets the preset calibration accuracy requirements based on the updated actual motion trajectory information. When the straight-line driving deviation does not meet the preset calibration accuracy requirements, the zero-position deviation parameter can be redefined, or the straight-line driving reference before the update can be restored.

[0051] The key to the above solution lies in the fact that the intelligent driving system 110 does not merely provide ordinary motion state data, but rather provides the target straight-line path tracking results obtained based on environmental perception and trajectory tracking; the steering control system 120 does not independently determine whether the vehicle is traveling straight based on its own steering angle or encoder information, but rather collects the position data of its own actuators within the calibration time window W confirmed by the intelligent driving system 110. Thus, cross-system calibration collaboration between the intelligent driving system 110 and the steering control system 120 can be formed.

[0052] like Figure 5 As shown, this embodiment also provides a straight-line driving calibration device for a smart mobile device. The device includes a path determination module, a trajectory judgment module, a triggering module, a position acquisition module, a deviation determination module, and a reference update module.

[0053] The path determination module is used by the intelligent driving system 110 to determine the target straight path L based on environmental perception information, and to control the intelligent mobile device 100 to travel along the target straight path L. The trajectory judgment module is used by the intelligent driving system 110 to obtain the actual motion trajectory information of the intelligent mobile device 100 during its travel along the target straight path L, and to determine whether the intelligent mobile device 100 meets the target straight path tracking conditions based on the actual motion trajectory information. The triggering module is used by the intelligent driving system 110 to send straight calibration trigger information to the steering control system 120 when the intelligent mobile device 100 meets the target straight path tracking conditions.

[0054] The position acquisition module is used by the steering control system 120 in response to the straight-line calibration trigger information to acquire the current position data of the steering actuator 130 within the calibration time window W corresponding to the target straight-line path tracking conditions. The deviation determination module is used to determine the zero-position deviation parameter of the steering actuator 130 relative to the straight-line driving direction based on the current position data acquired within the calibration time window W. The reference update module is used to update the straight-line driving reference of the steering actuator 130 based on the zero-position deviation parameter.

[0055] This embodiment also provides an intelligent mobile device 100. The intelligent mobile device 100 includes an intelligent driving system 110, a steering control system 120, a steering actuator 130, and a memory 140. The memory 140 stores a program that, when executed, implements the straight-line driving calibration method of any of the above embodiments.

[0056] This embodiment also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the straight-line driving calibration method in any of the above embodiments. The computer-readable storage medium can be a read-only memory, random access memory, flash memory, hard disk, solid-state drive, memory card, or other medium capable of storing programs.

[0057] Furthermore, the environmental perception component 160 may include one or more of the following: a camera, a depth camera, a LiDAR, a millimeter-wave radar, an ultrasonic sensor, and a positioning module. The intelligent driving system 110 can identify and fuse the data output by the environmental perception component 160 to obtain wall boundaries, passage boundaries, ground boundaries, obstacle boundaries, road edges, passable areas, target point locations, or map positioning results. All of the above information can be used as environmental perception information to determine the target straight-line path L.

[0058] In one embodiment, the intelligent driving system 110 determines the target straight path L based on the channel boundary. Specifically, the intelligent driving system 110 can identify the walls or boundary lines on both sides of the channel, calculate the channel centerline, and use the channel centerline or a line segment parallel to the channel centerline as the target straight path L. This implementation is suitable for straight-line calibration of intelligent wheelchairs in scenarios such as corridors, hospital ward passages, and home walkways.

[0059] In another embodiment, the intelligent driving system 110 determines the target straight path L based on the target point's location. Specifically, when there are no obstacles between the target point and the current position that need to be bypassed, the intelligent driving system 110 can generate a straight line segment from the current position to the target point as the target straight path L. This implementation is suitable for mobile robots, intelligent mobility scooters, or nursing transport equipment to perform straight-line calibration in open areas.

[0060] In another embodiment, the intelligent driving system 110 determines the target straight path L based on map information or a navigation task. The map information may include a pre-built indoor map, navigation lines, virtual tracks, channel directions, or positioning coordinates. The intelligent driving system 110 may select a path segment from the navigation task that meets the straight-line driving conditions as the target straight-line path L, and perform straight-line driving calibration on that path segment.

[0061] The intelligent driving system 110 can obtain actual motion trajectory information in multiple ways. Position change information can be obtained by a positioning module, visual odometry, lidar positioning, or a fusion positioning algorithm; heading angle change information can be obtained by an inertial measurement unit, electronic compass, or a fusion attitude algorithm; odometry information can be calculated by wheel speed sensors or motor encoders; trajectory curvature information can be obtained by fitting continuously sampled position points. The above information can be used individually or in combination.

[0062] Specifically, the visual images output by the environmental perception component 160 can undergo distortion correction, perspective transformation, edge extraction, semantic segmentation, or target detection processing to obtain wall boundaries, passage boundaries, road edges, or passable area boundaries. The LiDAR point cloud can undergo noise reduction, ground point removal, line feature extraction, or rasterization processing to obtain passage directions, obstacle boundaries, or wall line features. Ultrasonic or millimeter-wave radar data can be used to assist in confirming lateral boundary distances and the status of obstacles ahead.

[0063] The data output by the motion detection component 150 can also be preprocessed. Wheel speed information can be converted into travel distance and speed based on the encoder pulses of the left and right wheels or multiple traveling wheels; inertial measurement information can undergo zero-bias correction, low-pass filtering, or attitude fusion to obtain heading angle, angular velocity, or heading angle change; positioning information can undergo time synchronization, coordinate transformation, and abnormal jump rejection to obtain a continuous sequence of position points. After the above processing, the intelligent driving system 110 can form the actual motion trajectory T.

[0064] In sensor fusion, the intelligent driving system 110 can fuse wheel speed odometer data, inertial measurement information, and visual or radar positioning results. For example, wheel speed odometer data provides continuous displacement over a short period, inertial measurement information provides heading changes, and visual or radar positioning results are used to correct accumulated drift. After fusion, timestamped position points, heading angles, and velocity information can be obtained, which together constitute the actual motion trajectory information.

[0065] When determining the target straight path tracking conditions, the intelligent driving system 110 can first project the actual motion trajectory T onto the target straight path L and calculate the lateral deviation of the actual motion trajectory T relative to the target straight path L. When the lateral deviation is less than a first preset threshold in multiple consecutive sampling periods, it can be considered that the device is stably tracking the target straight path L in the lateral position.

[0066] The first preset threshold can be set according to the type, speed, and sensor accuracy of the intelligent mobile device 100. For example, for indoor intelligent wheelchairs or low-speed mobile robots, the first preset threshold can be set to 1cm to 10cm, preferably 2cm to 5cm; for larger logistics handling equipment, the first preset threshold can also be set to 5cm to 20cm. The above values ​​are only examples, and those skilled in the art can adjust them according to the equipment size, wheelbase, positioning accuracy, and usage scenario.

[0067] The intelligent driving system 110 can also calculate the trajectory curvature based on the continuous position points of the actual motion trajectory T. When the trajectory curvature is less than a second preset threshold, it indicates that the actual motion trajectory T is close to a straight line rather than an arc or serpentine trajectory. For low-speed mobile devices, the trajectory curvature can be calculated by fitting an arc or a straight line to multiple adjacent position points, or it can be approximated by the ratio of the change in heading angle to the travel distance.

[0068] The second preset threshold can be a trajectory curvature threshold. For example, the second preset threshold can be set to 0.005 m^-1 to 0.1 m^-1, preferably 0.01 m^-1 to 0.05 m^-1; it can also be represented by the equivalent turning radius, for example, requiring the equivalent turning radius corresponding to the actual motion trajectory T to be greater than 10 m, 20 m, or 50 m. For short-distance calibration, the trajectory curvature can also be replaced by the linear fitting residual. When the fitting residual is less than the preset residual threshold, the trajectory is considered to meet the linear requirement.

[0069] The intelligent driving system 110 can also determine the change in heading angle. When the change in heading angle of the intelligent mobile device 100 within the calibration candidate time period is less than a third preset threshold, it indicates that the device's attitude change is small, which is helpful in confirming that the device is in a stable straight-line driving state. The change in heading angle can be the difference in heading angle at the start and end of the calibration candidate time period, or it can be the maximum value, mean square value, or cumulative change in heading angle fluctuation within that time period.

[0070] The third preset threshold can be a heading angle change threshold. For example, in an indoor low-speed calibration scenario, the third preset threshold can be set to 0.5° to 5°, preferably 1° to 3°. When calculating the heading angle change using angular velocity integral, the average yaw rate within the calibration candidate time period can also be required to be less than a preset angular velocity threshold, for example, less than 0.5° / s to 5° / s. This can prevent the equipment from being misjudged as traveling in a straight line during slight turns.

[0071] In some implementations, the target straight-line path tracking condition may also include a duration condition or a travel distance condition. For example, the intelligent driving system 110 determines that the target straight-line path tracking condition is met only when the lateral deviation, trajectory curvature, and heading angle changes all meet the threshold requirements, and this state is maintained continuously for no less than a preset time or a preset distance. This avoids triggering unreliable calibration due to instantaneous data accidentally meeting the threshold.

[0072] The duration condition can be set from 0.5s to 10s, preferably from 1s to 3s; the travel distance condition can be set from 0.2m to 5m, preferably from 0.5m to 2m. For indoor low-speed devices such as intelligent wheelchairs, a shorter distance and lower speed calibration window can be preferred; for logistics handling equipment, a longer straight line segment can be used to improve statistical stability.

[0073] The straight-line calibration trigger information can take several forms. In one method, the intelligent driving system 110 sends calibration start information to the steering control system 120. After receiving the calibration start information, the steering control system 120 begins to collect current position data. When the intelligent driving system 110 determines that the target straight-line path tracking conditions are no longer met or the collection time reaches the preset requirement, it sends calibration end information. The steering control system 120 then stops collecting data and calculates the zero-position deviation parameter based on the collected data.

[0074] In another approach, the intelligent driving system 110 sends target straight-line path tracking success information and straight-line driving confidence level to the steering control system 120. The steering control system 120 can enter the calibration sampling state only when the straight-line driving confidence level is greater than a preset confidence level threshold. The confidence level can be calculated comprehensively based on lateral deviation, trajectory curvature, heading angle change, sensor reliability, positioning quality, and driving speed stability.

[0075] For example, the intelligent driving system 110 can normalize the lateral deviation, trajectory curvature, and heading angle change to scores between 0 and 1, and then combine this with the sensor reliability to obtain the straight-line driving confidence score C. The smaller the lateral deviation, the smaller the trajectory curvature, the smaller the heading angle change, and the higher the sensor quality, the higher the confidence score C. When C is greater than 0.7, 0.8, or 0.9, the target straight-line path tracking result can be considered reliable, and calibration can be triggered.

[0076] In another approach, the straight-line calibration trigger information can carry target straight-line path information and calibration time window information. The target straight-line path information may include path direction, path start point, path end point, or path coordinates; the calibration time window information may include window start time, window end time, window duration, or sampling synchronization identifier. The steering control system 120 can use this information to determine which current position data belong to valid calibration samples.

[0077] For the calibration time window W, the steering control system 120 can use either real-time sampling or cache backtracking. Real-time sampling means that the steering control system 120 collects the current position data in real time after receiving the calibration start information; cache backtracking means that the steering control system 120 continuously caches the current position data for a period of time, and extracts the current position data within the window from the cached data after receiving the calibration time window W sent by the intelligent driving system 110.

[0078] The current position data can be the raw feedback value of the steering actuator 130 or the converted angle value. Taking the steering motor as an example, the current position data can be the encoder count value, the encoder mechanical angle, the output angle after deceleration, or the actual angle of the steering wheel; taking the servo motor as an example, the current position data can be the PWM midpoint offset, the servo motor feedback potentiometer position, or the servo motor control angle; taking the front wheel steering mechanism as an example, the current position data can be the output value of the front wheel angle sensor.

[0079] When the current position data is an encoder count value, the steering control system 120 can convert the encoder count value into the output angle of the steering actuator 130 based on the encoder resolution, reduction ratio, and mechanical transmission relationship. For example, the encoder count value can be first converted into a motor angle, and then divided by the reduction ratio to obtain the steering wheel angle. When the current position data is a servo control quantity, the control pulse width or feedback voltage can be converted into the servo output angle based on the correspondence between the servo control pulse width and the angle.

[0080] To ensure accurate data correspondence, the intelligent driving system 110 and the steering control system 120 can synchronize the sampled data in time. For example, the straight-line calibration trigger information can carry a timestamp, and the steering control system 120 selects the current position data within the same time range based on this timestamp; alternatively, the two systems can use a unified system clock to ensure that the actual motion trajectory T and the current position data correspond in time. This avoids using a lagging steering position for calibration after the vehicle has left the straight-line tracking state.

[0081] When determining the zero-position deviation parameter, the steering control system 120 can first remove outliers from multiple current position data within the calibration time window W. For example, if the difference between a sampling point and its adjacent sampling points exceeds a preset jump threshold, or if the difference between the sampling point and the sample average exceeds a preset outlier threshold, the sampling point can be removed. After removing outliers, the remaining data is then averaged or weighted averaged.

[0082] In the weighted averaging method, the steering control system 120 can assign weights to each current position data point based on the straight-line driving confidence level, lateral deviation at the sampling time, trajectory curvature, heading angle change, or driving speed stability provided by the intelligent driving system 110. Sampling points with more stable target straight-line path tracking performance can have higher weights. This results in zero-position deviation parameters that better reflect the position of the steering actuator under real straight-line driving conditions.

[0083] In the filtering method, the steering control system 120 can use moving average filtering, median filtering, low-pass filtering, Kalman filtering, or other filtering methods suitable for low-speed mobile devices to smooth the current position data within the calibration time window W. The filtered stable value can be used as a new straight-line driving reference, or the zero-position deviation parameter can be obtained by subtracting it from the theoretical straight-line reference.

[0084] In the stability assessment method, the steering control system 120 can calculate the range, variance, standard deviation, or rate of change of the current position data. When the fluctuation of the current position data is greater than a preset threshold, it indicates that the steering actuator 130 may still be adjusting or subject to external disturbances within the calibration time window W. In this case, the current calibration can be abandoned, and resampling can be performed after the target straight path tracking conditions are met again.

[0085] The zero-position deviation parameter can take different forms depending on the type of steering actuator 130. For encoder-controlled steering motors, the zero-position deviation parameter can be the encoder zero-point compensation value; for servo motors, the zero-position deviation parameter can be the servo motor mid-position compensation value; for front wheel steering mechanisms with angle sensors, the zero-position deviation parameter can be the front wheel straight-line angle compensation value; for steering wheel modules, the zero-position deviation parameter can be the steering wheel straight-line target angle compensation amount.

[0086] In one specific calculation method, let the theoretical straight-line driving reference be P0, and the current position data after filtering within the calibration time window W be P1. Then, the zero-position deviation parameter can be ΔP = P1 - P0. During subsequent straight-line driving control, the steering control system 120 can correct the straight-line control target from P0 to P0 + ΔP. In another method, P1 can also be directly written as the new straight-line driving reference.

[0087] For example, if the theoretical straight-line driving reference P0 is an encoder count value of 10000, and the current position data P1 within the calibration time window W, after outlier removal and averaging, is 10085, then the zero-position deviation parameter ΔP = 85 encoder counts can be obtained. During subsequent straight-line driving control, the steering control system 120 corrects the straight-line target position from 10000 to 10085, or adds a compensation of 85 counts to all steering target positions. If the current position data is expressed in angles, for example, P0 is 0° and P1 is 1.2°, then the zero-position deviation parameter can be a front wheel straight-line angle compensation value of 1.2°.

[0088] In another example, if multiple steering angle data collected within the calibration time window W are 1.0°, 1.1°, 1.2°, 1.1°, and 5.0°, the steering control system 120 can first remove the 5.0° value, which deviates significantly from other samples, as an outlier. Then, it can average the remaining samples to obtain 1.1°, and use 1.1° as the servo center compensation value or the front wheel straight-line angle compensation value. Through this processing, the zero-position deviation parameter can be made closer to the stable steering position under actual straight-line driving conditions.

[0089] When updating the straight-line driving reference, the steering control system 120 can write the zero-position deviation parameter into volatile memory as a temporary compensation parameter, or into non-volatile memory as a long-term compensation parameter. For factory debugging or maintenance calibration, writing into non-volatile memory is preferred; for self-learning correction during operation, temporary parameters can be written first, and the long-term parameters can be updated after multiple verifications of stability.

[0090] To avoid erroneous calibration, the calibration validity conditions can further include a driving speed limit. If the speed is too low, the trajectory information may be significantly affected by positioning noise; if the speed is too high, the device may be in a state of rapid dynamic adjustment. Therefore, the driving speed of the intelligent mobile device 100 can be required to be within a preset speed range. This speed can be determined by wheel speed information, odometer information, positioning changes, or speed estimates from the intelligent driving system 110.

[0091] The calibration validity conditions may also include obstacle avoidance status judgment. When the intelligent driving system 110 is performing obstacle avoidance, detour, emergency braking, collision avoidance, or manual intervention, even if the actual trajectory is close to a straight line for a short period of time, it should not be used as a calibration basis. Therefore, when there is an obstacle avoidance steering command, a rapid acceleration command, a rapid deceleration command, an emergency stop signal, or a manual forced steering signal, the steering control system 120 may prohibit updating the straight-line driving reference.

[0092] During the updated verification process, the intelligent driving system 110 can regenerate the target straight path L and control the intelligent mobile device 100 to travel along the target straight path L. If the lateral deviation, trajectory curvature, or heading angle change of the updated actual motion trajectory T relative to the target straight path L meets the preset calibration accuracy requirements, the calibration is confirmed to be valid; otherwise, the zero-position deviation parameters can be resampled and calculated, or the straight-line driving reference before the update can be restored to ensure equipment safety.

[0093] In some implementations, the intelligent mobile device 100 can perform the above calibration method during the factory commissioning phase, or it can perform the above calibration method after maintenance, after replacement of the steering actuator 130, after replacement of the travel wheels, after long-term deviation is detected, after user-triggered calibration, or during regular system maintenance. Therefore, this invention can be used for both one-time factory calibration and self-calibration during equipment use.

[0094] In some implementations, the intelligent driving system 110 and the steering control system 120 are independent control systems. The intelligent driving system 110 is responsible for determining the target straight path L, acquiring the actual motion trajectory T, and judging the target straight path tracking conditions; the steering control system 120 is responsible for receiving straight-line calibration trigger information, collecting the current position data of the steering actuator 130, determining the zero-position deviation parameter, and updating the straight-line driving reference. This division of labor avoids the steering control system 120 relying solely on its own feedback signals to judge the straight-line state, thereby improving the reliability of the calibration basis.

[0095] In other embodiments, the intelligent driving system 110 and the steering control system 120 can be integrated into the same controller, but they can still perform the above steps as different functional modules. For example, the path planning and trajectory determination functional module outputs straight-line calibration trigger information, and the steering control functional module responds to the straight-line calibration trigger information and updates the straight-line driving reference of the steering actuator 130. This integration does not affect the technical effect of the present invention.

[0096] It should be noted that the intelligent driving system 110 and steering control system 120 mentioned in the above embodiments can be hardware-independent controllers or different functional modules within the same controller; each of these modules can be implemented through hardware circuits, software programs, or a combination of hardware and software. Those skilled in the art can adjust the sensor type, path generation method, trajectory judgment indicators, sampling frequency, filtering method, and compensation parameter form without departing from the concept of this invention, and all such adjustments should fall within the protection scope of this invention.

Claims

1. A method for calibrating the straight-line driving of an intelligent mobile device, characterized in that, The intelligent mobile device includes an intelligent driving system and a steering control system, the steering control system being used to control a steering actuator, and the method comprising: The intelligent driving system determines the target straight path based on environmental perception information and controls the intelligent mobile device to drive along the target straight path; The intelligent driving system acquires the actual motion trajectory information of the intelligent mobile device during its journey along the target straight path, and determines whether the intelligent mobile device meets the target straight path tracking conditions based on the actual motion trajectory information. When the intelligent mobile device meets the target straight path tracking conditions, the intelligent driving system sends straight calibration trigger information to the steering control system; The steering control system responds to the straight line calibration trigger information and collects the current position data of the steering actuator within the calibration time window corresponding to the target straight line path tracking condition; The steering control system determines the zero-position deviation parameter of the steering actuator relative to the straight driving direction based on the current position data collected within the calibration time window. The straight-line driving reference of the steering actuator is updated based on the zero-position deviation parameter.

2. The straight-line driving calibration method according to claim 1, characterized in that, The environmental perception information includes at least one of the following: visual image information, depth image information, lidar information, millimeter-wave radar information, ultrasonic information, positioning information, road boundary information, wall boundary information, and passage boundary information.

3. The straight-line driving calibration method according to claim 1, characterized in that, The target straight path is a straight driving path generated by the intelligent driving system based on environmental boundaries, channel direction, target point location, map information, or navigation task.

4. The straight-line driving calibration method according to claim 1, characterized in that, The actual motion trajectory information includes at least one of the following: position change information of the smart mobile device, heading angle change information, odometer information, inertial measurement information, wheel speed information, and trajectory curvature information.

5. The straight-line driving calibration method according to claim 1, characterized in that, The target straight-line path tracking conditions include: The lateral deviation of the actual motion trajectory relative to the target straight path is less than a first preset threshold. And / or, the trajectory curvature of the actual motion trajectory is less than a second preset threshold; And / or, the change in heading angle of the smart mobile device is less than a third preset threshold; And / or, the duration or distance the smart mobile device travels along the target straight path reaches a preset requirement.

6. The straight-line driving calibration method according to claim 1, characterized in that, The straight-line calibration trigger information includes at least one of the following: target straight-line path tracking success information, calibration start information, calibration end information, straight-line driving confidence, target straight-line path information, and calibration time window information.

7. The straight-line driving calibration method according to claim 1, characterized in that, The calibration time window is the time interval during which the intelligent driving system determines that the intelligent mobile device continuously meets the target straight-line path tracking conditions.

8. The straight-line driving calibration method according to claim 1, characterized in that, The steering control system collects multiple current position data within the calibration time window and determines the zero-position deviation parameter based on the multiple current position data.

9. The straight-line driving calibration method according to claim 8, characterized in that, The steering control system performs averaging, weighted averaging, filtering, outlier removal, or stability assessment on multiple current position data to obtain the zero-position deviation parameter.

10. The straight-line driving calibration method according to claim 1, characterized in that, The current position data includes at least one of the following: steering motor encoder position, servo motor position, steering angle sensor information, front wheel steering angle information, and steering actuator feedback position.

11. The straight-line driving calibration method according to claim 1, characterized in that, The zero-position deviation parameter includes at least one of the following: steering zero-position offset, encoder zero-point compensation value, servo motor center position compensation value, and front wheel straight-line angle compensation value.

12. The straight-line driving calibration method according to claim 1, characterized in that, Before the steering control system determines the zero-position deviation parameter, the following is also included: Determine whether the smart mobile device meets the calibration validity conditions; The calibration validity conditions include at least one of the following: the driving speed is within a preset speed range, no obstacle avoidance steering command has been received, the vehicle is not in a state of rapid acceleration, the vehicle is not in a state of rapid deceleration, and the fluctuation of the current position data is less than a preset threshold.

13. The straight-line driving calibration method according to claim 1, characterized in that, Updating the straight-line driving reference of the steering actuator based on the zero-position deviation parameter includes: The zero-position deviation parameter is written into the storage unit of the steering control system; In subsequent straight-line driving control, the target position of the steering actuator is compensated based on the zero-position deviation parameter.

14. The straight-line driving calibration method according to claim 1, characterized in that, After updating the straight-line driving reference of the steering actuator, the following is also included: The intelligent driving system controls the intelligent mobile device to travel along the target straight path again; Obtain updated actual motion trajectory information; Based on the updated actual motion trajectory information, determine whether the straight-line driving deviation of the smart mobile device meets the preset calibration accuracy requirements.

15. The straight-line driving calibration method according to claim 14, characterized in that, When the straight-line driving deviation does not meet the preset calibration accuracy requirements, the zero-position deviation parameter is re-determined, or the straight-line driving reference before the update is restored.

16. The straight-line driving calibration method according to claim 1, characterized in that, The intelligent driving system and the steering control system are independent control systems. The intelligent driving system is used to determine the target straight path and to determine whether the actual motion trajectory meets the target straight path tracking conditions. The steering control system is used to update the straight driving reference of the steering actuator according to the straight calibration trigger information.

17. A linear motion calibration device for an intelligent mobile device, characterized in that, include: The path determination module is used to determine the target straight path based on environmental perception information through the intelligent driving system, and to control the intelligent mobile device to drive along the target straight path; The trajectory determination module is used to obtain the actual motion trajectory information of the intelligent mobile device during the process of driving along the target straight path through the intelligent driving system, and to determine whether the intelligent mobile device meets the target straight path tracking conditions based on the actual motion trajectory information. The trigger module is used to send straight calibration trigger information to the steering control system through the intelligent driving system when the intelligent mobile device meets the target straight path tracking conditions; The position acquisition module is used to acquire the current position data of the steering actuator within a calibration time window corresponding to the target straight path tracking condition, in response to the straight calibration trigger information by the steering control system. The deviation determination module is used to determine the zero-position deviation parameter of the steering actuator relative to the straight driving direction based on the current position data collected within the calibration time window; The reference update module is used to update the straight-line driving reference of the steering actuator according to the zero-position deviation parameter.

18. A smart mobile device, characterized in that, It includes an intelligent driving system, a steering control system, a steering actuator, and a memory, wherein the memory stores a program that, when executed, implements the straight-line driving calibration method as described in any one of claims 1 to 16.

19. The intelligent mobile device according to claim 18, characterized in that, The intelligent mobile device is an intelligent wheelchair, an intelligent mobility scooter, a mobile robot, a nursing transport device, or a logistics handling device.

20. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the straight-line driving calibration method as described in any one of claims 1 to 16.

Images