Agv fork positioning method and device based on ranging sensor analysis and medium

Abstract

The application discloses an AGV fork positioning method and device based on ranging sensor analysis, and a medium, and relates to the technical field of data processing.The method comprises the following steps: configuring ranging, vision and orientation sensors on a target AGV, and constructing a sensor collection device group; activating the device group, collecting multi-dimensional sensing data of the target AGV, including distance, image and azimuth angle data; using the image data to position the environment, and traversing a three-dimensional simulation space of a target scene to extract a local reference space; combining the multi-dimensional sensing data to determine the first relative position of the target AGV fork in the local reference space; analyzing the local reference space to obtain the second relative position of the target object, comparing the difference between the two to obtain positioning difference information, and optimizing the positioning of the target AGV fork based on the difference information, thereby achieving the technical effects of improving the positioning efficiency of the fork and the safety of the operation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to an AGV fork positioning method based on ranging sensor analysis, a device and a medium. BACKGROUND

[0002] In the docking of a new energy backend (shelf picking and placing side fork AGV), the number of storage locations is about 1000-5000. Due to the left and right deviation in the walking process of the AGV, there is a relative deviation when the side fork AGV picks and places the material on the guide block, so that the material occasionally falls on the fork guide block, which easily leads to the convexity of the AGV walking out of the vehicle body, causing secondary scratching of the equipment or uneven placement of the material on the shelf, and there are technical problems of low fork positioning accuracy, low positioning efficiency and influence on operation safety. SUMMARY

[0003] The present application provides an AGV fork positioning method based on ranging sensor analysis, a device and a medium to solve the technical problems of low fork positioning accuracy, low positioning efficiency and influence on operation safety in the prior art, and to achieve the technical effect of improving the positioning efficiency and operation safety of the fork.

[0004] In a first aspect, the present application provides an AGV fork positioning method based on ranging sensor analysis, wherein the method comprises:

[0005] A ranging sensor, a vision sensor and a direction sensor are arranged on a target AGV, and a sensing acquisition device group is configured.

[0006] The sensing acquisition device group is activated to collect multi-dimensional ranging sensor data of the target AGV, wherein the multi-dimensional ranging sensor data includes distance data, image data and azimuth angle data.

[0007] The environment position is located based on the image data, and according to the environment position location result, the three-dimensional simulation space of the target scene is traversed to extract a local reference space.

[0008] The distance data, the image data and the azimuth angle data are combined to determine the first relative spatial position of the fork of the target AGV in the local reference space.

[0009] The local reference space is analyzed to obtain the second relative spatial position of the target object, and the first relative spatial position and the second relative spatial position are compared to obtain the difference positioning information.

[0010] The fork positioning of the target AGV is optimized according to the difference positioning information.

[0011] In a feasible manner, a ranging sensor, a vision sensor and a direction sensor are arranged on a target AGV, and a sensing acquisition device group is configured, comprising:

[0012] Determine sensor specification information based on the positioning requirement information of the target scene.

[0013] According to the sensor specification information, the target AGV matches and filters the existing sensors, and defines a sensor layout list according to the matching and filtering result.

[0014] Update the sensor layout of the target AGV with the sensor layout list, and connect the ranging sensor, vision sensor and azimuth sensor to the sensor collection sink to obtain the sensor collection device group.

[0015] In a feasible manner, based on the image data, the environmental position is located, and according to the environmental position locating result, the three-dimensional simulation space of the target scene is traversed, and the local reference space is extracted, including:

[0016] Image denoising and image enhancement are performed on the image data, and key feature information in the image is extracted.

[0017] The environmental feature recognition based on the key feature information is performed through the computer vision method, and the environmental feature vector is generated.

[0018] The environmental position locating result of the target AGV is determined by traversing the environmental feature vector database of the target scene.

[0019] Combined with the environmental position locating result and the preset space extraction constraint, the local space extraction is performed in the three-dimensional simulation space to obtain the local reference space.

[0020] In a feasible manner, combined with the distance data, the image data and the azimuth angle data, the first relative spatial position of the fork of the target AGV in the local reference space is determined, including:

[0021] Taking the azimuth angle data as the first registration basis, combined with the initial space orientation coordinate system of the target AGV, the first-level registration is performed.

[0022] Taking the distance data as the second registration basis, the position of the first-level registration result is corrected based on the triangulation method to obtain the second-level registration result.

[0023] Parse the image data, extract the contour information and texture information of the target object as auxiliary registration basis, and verify the second-level registration result through the auxiliary registration basis.

[0024] If the verification is passed, the first relative spatial position is determined based on the second-level registration result and the inboard relative position of the fork of the target AGV.

[0025] In a feasible manner, the target object is the operation target object of the target AGV fork, including a transportation target object and a placement target object, and the second relative spatial position is the relative spatial position of the operation task control point corresponding to the target object in the local reference space.

[0026] In one feasible embodiment, the method further includes:

[0027] The target AGV obtains the fork operation record, and filters and determines the non-empty operation area from the fork operation record, and outputs it as a feedback monitoring area set.

[0028] The feedback monitoring area set is continuously monitored by the scene monitoring device of the target scene, and feedback correction is performed on the target object according to the continuous monitoring result.

[0029] In one feasible embodiment, the method further includes:

[0030] The number of feedback corrections is accumulated through the accumulator.

[0031] If the number of feedback corrections exceeds a first preset number or the number of continuous feedback corrections exceeds a second preset number, the decision on fork positioning optimization is corrected according to the feedback correction record, wherein the first preset number is greater than the second preset number.

[0032] In one feasible embodiment, the method further includes:

[0033] Get the safety fence of the target space.

[0034] The relative distance between the target AGV fork and the safety fence is monitored in real time.

[0035] The relative distance is weighted and corrected according to the safety fence level and the real-time task level of the target AGV. If the correction result is less than a preset safety threshold, the target AGV is automatically stopped.

[0036] In a second aspect, the present invention further provides an electronic device, comprising: a memory for storing executable instructions; and a processor for implementing the AGV fork positioning method based on ranging sensor analysis provided by the present invention when executing the executable instructions stored in the memory.

[0037] In a third aspect, the present invention further provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the AGV fork positioning method based on ranging sensor analysis provided by the present invention.

[0038] The application discloses an AGV fork positioning method based on ranging sensor analysis, equipment and a medium, comprising: setting ranging sensors, visual sensors and direction sensors on a target AGV, and configuring a generated sensor collection equipment group; activating the sensor collection equipment group to collect multi-dimensional ranging sensor data of the target AGV, wherein the multi-dimensional ranging sensor data includes distance data, image data and azimuth angle data; performing environmental position positioning based on the image data, and traversing a three-dimensional simulation space of a target scene according to the environmental position positioning result to extract a local reference space; combining the distance data, the image data and the azimuth angle data to determine a first relative spatial position of the fork of the target AGV in the local reference space; analyzing the local reference space to obtain a second relative spatial position of a target object, and comparing the first relative spatial position and the second relative spatial position to obtain difference positioning information; and performing fork positioning optimization of the target AGV according to the difference positioning information, the AGV fork positioning method based on ranging sensor analysis, equipment and medium disclosed by the application solve the technical problems of low fork positioning precision, low positioning efficiency and influence on operation safety, and achieve the technical effects of improving fork positioning efficiency and operation safety. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 It is a flowchart of the AGV fork positioning method based on ranging sensor analysis of the application.

[0040] Figure 2 It is a structural schematic diagram of an exemplary electronic device of the application.

[0041] Explanation of reference numerals: processor 31, memory 32, input device 33, output device 34. DETAILED DESCRIPTION

[0042] The above technical solutions will be described in detail below in combination with the drawings and specific embodiments, so as to better understand the above technical solutions. Obviously, the described embodiments are only part of the embodiments of the application, not all embodiments of the application, and it should be understood that the application is not limited to the example embodiments for explaining the application. Based on the embodiments of the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application. In addition, it should be noted that, for convenience of description, only parts related to the application are shown in the drawings, not all.

[0043] Embodiment one

[0044] Figure 1 It is a flowchart of the AGV fork positioning method based on ranging sensor analysis of the application, wherein the method comprises:

[0045] The ranging sensor, the vision sensor, and the orientation sensor are arranged on the target AGV, and a sensing collection device group is configured.

[0046] Specifically, the ranging sensor is used to measure the distance between the AGV and the surrounding obstacles or target points, and is mainly used for obstacle avoidance, navigation, and path planning, including LiDAR, ultrasonic sensor, and infrared sensor. The vision sensor is used for environment perception, target recognition, path tracking, and understanding of complex scenes, including RGB camera, depth camera, and panoramic camera. The orientation sensor is used to obtain the attitude, direction, and position information of the AGV, supports navigation and path planning, and can include IMU (Inertial Measurement Unit), electronic compass, GPS module, and UWB (Ultra-Wideband) module.

[0047] Specifically, the ranging sensor, the vision sensor, and the orientation sensor are integrated on the AGV to form a complete sensing collection device group, including a data processing unit (such as an embedded main control board) corresponding to the collection device group composed of the ranging sensor, the vision sensor, and the orientation sensor. Through the above configuration, the sensing collection device group of the AGV can realize efficient data collection and environment perception, and meet the task execution requirements in complex scenes.

[0048] In some embodiments, the ranging sensor, the vision sensor, and the orientation sensor are arranged on the target AGV, and a sensing collection device group is configured, including:

[0049] Based on the positioning requirement information of the target scene, the sensor specification information is determined;

[0050] According to the sensor specification information, the target AGV is matched and filtered with the existing sensors, and the sensor layout list is defined according to the matching and filtering result;

[0051] The sensor layout list is used to update the sensor layout of the target AGV, and the ranging sensor, the vision sensor, and the orientation sensor are connected to the sensing collection sink to obtain the sensing collection device group

[0052] Specifically, first, according to the positioning requirement information of the target scene, the required sensor types and performance indicators are analyzed, including ranging range, accuracy, response time, etc., to determine the sensor specification information that meets these requirements, such as the performance parameters of LIDAR, camera, and gyroscope; then, the existing sensors on the target AGV are checked for matching and filtering to determine the available sensor models and specifications, and according to the matching and filtering result, a sensor layout list is defined for the sensors that need to be added or replaced, wherein the sensor layout list includes the sensor model, quantity, and installation position.

[0053] Specifically, according to the sensor layout list, the physical layout of the sensors on the target AGV is carried out, including fixing, wiring and interface connection, etc., so as to connect the ranging sensors, vision sensors and orientation sensors to the sensing and collecting sink through appropriate interfaces, and the sensing and collecting sink is a centralized data processing unit or control module; then, the software system of the AGV is updated, including the sensor driver and configuration parameters, and the function test of the sensing and collecting device group is carried out to ensure the accuracy and real-time performance of the sensor data.

[0054] Through the above process, it can be ensured that the target AGV is equipped with sensors suitable for the target scene, and can accurately position and navigate.

[0055] Activating the sensing and collecting device group, collecting multi-dimensional ranging sensor data of the target AGV, wherein the multi-dimensional ranging sensor data includes distance data, image data and azimuth angle data.

[0056] Specifically, all sensors installed on the AGV are started, including ranging sensors (such as laser radar or ultrasonic sensor), vision sensors (such as camera) and orientation sensors (such as gyroscope or compass), the sensor system is initialized to ensure that all sensors work synchronously, and the correct working mode is adjusted, that is, according to the operation demand and processing capacity of the AGV, the appropriate data acquisition sampling rate is set to ensure the consistency and availability of the data; then, the distance data between the AGV and the surrounding obstacles is collected in real time using the ranging sensor, the image data within the field of view of the AGV is captured using the vision sensor (such as camera) for environment recognition and navigation, and the current azimuth angle data of the AGV is obtained through the orientation sensor (such as IMU, gyroscope or compass) for determining the orientation of the AGV.

[0057] Based on the image data, the environment position is located, and according to the environment position locating result, the three-dimensional simulation space of the target scene is traversed to extract the local reference space.

[0058] Specifically, the environment position locating refers to determining the position and direction of the current AGV in the environment through the environment data collected by the sensor combined with the known environment information (such as map or three-dimensional model), such as which shelf the AGV is in and the current orientation.

[0059] Specifically, the three-dimensional simulation space of the target scene refers to the three-dimensional digital representation of the real scene, which contains a virtual three-dimensional model of all geometric information of the target scene, can reflect the structure and layout of the actual environment, and is used for simulation, analysis and planning; the local reference space refers to a small range of three-dimensional area related to the current device position extracted from the three-dimensional simulation space of the target scene, which is a local subset of the environment, and is used for real-time perception, decision and operation of the AGV.

[0060] In some embodiments, the environmental position is located based on the image data, and according to the environmental position locating result, a local reference space is extracted by traversing the three-dimensional simulation space of the target scene, including:

[0061] The image data is denoised and enhanced, and the key feature information in the image is extracted;

[0062] The environmental feature recognition based on the key feature information is performed by a computer vision method to generate an environmental feature vector;

[0063] The environmental feature vector database of the target scene is traversed for matching to determine the environmental position locating result of the target AGV;

[0064] The local space extraction is performed in the three-dimensional simulation space in combination with the environmental position locating result and the preset space extraction constraint to obtain the local reference space.

[0065] Specifically, first, image preprocessing is performed, a denoising algorithm is applied to the collected image data to reduce noise in the image and improve the accuracy of subsequent processing; the contrast, brightness, etc. of the image are enhanced to make the key features in the image more obvious; then, the key feature information in the image, such as edges, corners, and textures, is extracted through feature engineering.

[0066] Specifically, the key feature information extracted from the image is analyzed using computer vision technology to identify and describe the features of the environment. For example, based on known template images, the appearance position of the target object is found in the target image; a deep learning object detection method (such as YOLO, Faster R-CNN) is used to detect specific objects or regions in the environment; a deep convolutional neural network (CNN) is used to classify the image at the pixel level to extract semantic information of different regions (such as roads, buildings, and plants) in the environment; then, the environmental features in the image are converted into numerical vectors, i.e. environmental feature vectors, which are used as inputs for subsequent calculations or modeling.

[0067] Specifically, the environmental feature vector database of the target scene is traversed, and the generated environmental feature vector is matched with the vectors in the database to determine the environmental position locating result of the AGV, such as the position of the shelf row; then, in combination with the environmental position locating result and the preset space extraction constraint (such as distance, angle, height, etc.), the local space region that needs to be extracted is determined in the three-dimensional simulation space to obtain the local reference space.

[0068] Through this process, the image data is used to accurately locate the environmental position, and the local reference space related to the position is extracted in the three-dimensional simulation space, which improves the accuracy and environmental adaptability of the AGV, enabling it to operate safely and efficiently in complex environments.

[0069] determining a first relative spatial position of the forks of the target AGV in the local reference space in combination with the distance data, the image data and the azimuth data.

[0070] In some embodiments, determining the first relative spatial position of the forks of the target AGV in the local reference space in combination with the distance data, the image data and the azimuth data comprises:

[0071] performing a first registration in combination with an initial spatial orientation coordinate system of the target AGV with the azimuth data as a first registration basis;

[0072] performing a position correction on the first registration result based on a triangulation method to obtain a second registration result with the distance data as a second registration basis;

[0073] analyzing the image data to extract contour information and texture information of the target object as an auxiliary registration basis, and verifying the second registration result through the auxiliary registration basis;

[0074] if the verification is passed, determining the first relative spatial position based on the second registration result and an internal relative position of the forks of the target AGV.

[0075] Specifically, first, a first registration (i.e., a coarse registration) is performed in combination with an initial spatial orientation coordinate system of the AGV with the azimuth data as a first registration basis, such as using a rotation matrix or a quaternion to transform the position of the forks from the AGV coordinate system to the local reference space to determine the initial orientation of the AGV relative to the local reference space, wherein the initial spatial orientation coordinate system is used for the preliminary position and orientation of the AGV in the local reference space.

[0076] Specifically, on the basis of the first registration result, a second registration result is obtained by using a triangulation method to calculate the accurate position of the forks of the AGV relative to the known fixed point with the distance data as a second registration basis.

[0077] Further, the image data is analyzed to extract the contour information and texture information of the target object, and the extracted contour information and texture information are used as an auxiliary registration basis; then, the extracted image features are compared with the known feature templates in the local reference space according to the second registration result to verify the accuracy of the second registration result, and if the position error of the image features and the position of the second registration result is less than a preset threshold, it is determined that the verification is passed.

[0078] Optionally, if the verification is not passed, the second registration result is further refined by using the image features, for example, by using an optimization algorithm (such as an ICP algorithm) to fine-tune to ensure high-precision matching of the position and orientation.

[0079] Specifically, after the verification passes, the secondary registration result is integrated with the in-machine relative position parameters of the target AGV fork, thereby calculating the final position of the fork in the local reference space.

[0080] Through the above process, the position of the AGV fork can be accurately determined in the local reference space. The primary registration and secondary registration provide position information based on sensor data, while the image data analysis provides visual verification to ensure the accuracy of the position data. This method improves the operation accuracy of the AGV, helps to reduce operation errors, and enhances the adaptability and reliability of the AGV in complex environments.

[0081] Analyzing the local reference space to obtain a second relative spatial position of the target object, and comparing the first relative spatial position and the second relative spatial position to obtain difference positioning information.

[0082] Specifically, the data in the local reference space is analyzed, including pre-set markers, known object positions or environmental features, to determine the position of the target object in the local reference space, i.e. to calculate the second relative spatial position of the target object relative to the AGV fork.

[0083] Specifically, the first relative spatial position (position of the AGV fork) and the second relative spatial position (position of the target object) are compared to calculate the difference between the two positions, including distance difference, angle difference, etc., to obtain difference positioning information, which is used to adjust the position and orientation of the AGV fork to ensure accurate alignment or handling of the target object.

[0084] In some embodiments, the target object is a target AGV fork operation target object, including a transport target object and a placement target object, and the second relative spatial position is the relative spatial position of the operation task control point corresponding to the target object in the local reference space.

[0085] Specifically, the target object specifically refers to the AGV fork operation target object, including target objects that need to be transported and target objects that need to be placed (such as shelves), and the second relative spatial position refers to the relative position of key control points in the operation task, such as the grabbing point, the placement point or other key operation points in the local reference space.

[0086] According to the difference positioning information, the AGV fork positioning is optimized.

[0087] Specifically, the positioning optimization of the target AGV fork according to the differential positioning information is to ensure that the AGV can accurately perform the carrying and placing tasks. Optionally, the fork positioning optimization includes adjusting the target AGV fork to the preset work position (ideal work position) to compensate for the position deviation, optimizing the trajectory of the target AGV fork work through path planning methods, thereby improving the work efficiency and accuracy and ensuring the safety of the goods.

[0088] Through the above process, the positioning accuracy of the AGV fork is improved, the success rate and efficiency of the carrying and placing tasks are increased, and the optimized positioning strategy enables the AGV to better adapt to complex and variable working environments, reduces operation errors, and improves the reliability and stability of the automated logistics system.

[0089] In some embodiments, the method further comprises:

[0090] The target AGV records the fork work record and selects the non-empty work area from the fork work record to output as a feedback monitoring area set.

[0091] The feedback monitoring area set is continuously monitored by the scene monitoring device of the target scene, and the target object is corrected according to the continuous monitoring result.

[0092] Specifically, the fork work record includes the work time, position coordinates, target object information (such as the type, weight, and size of the goods), and related operation states (such as picking up and placing goods); the non-empty work area is defined as a coordinate or area range that explicitly interacts with the target object, i.e., an area where goods exist after work, which needs to be continuously monitored.

[0093] Specifically, real-time data of the non-empty work area set is collected by fixed or mobile monitoring devices (such as cameras and laser radars) in the scene, and the monitoring data content includes the target object state (whether it is moved or damaged) and the area dynamic change information (such as whether there are new obstacles).

[0094] Further, the target object state is compared and analyzed to identify abnormal situations (such as incorrect placement of the target object, target loss, target slip, collapse, etc.), and corresponding correction and disposal are performed, such as re-transporting to the correct position, calling a backup monitoring device or mobile device to search for the target again, using the AGV fork to make slight adjustments to correct the slip state, suspending the work task in the relevant area, and starting manual intervention or collaborative robots for cleaning and recovery.

[0095] Through comprehensive identification and accurate processing of abnormal states, the error rate in logistics or warehousing work can be effectively reduced, the safety of target object storage and management is improved, and the work efficiency and the intelligent level of the automated system are optimized.

[0096] In some implementations, the method further comprises:

[0097] accumulating the number of feedback corrections by an accumulator;

[0098] if the number of feedback corrections exceeds a first preset number or the number of consecutive feedback corrections exceeds a second preset number, making a decision correction for the fork positioning optimization according to the feedback correction records, wherein the first preset number is greater than the second preset number.

[0099] Specifically, an accumulator is set, and the number of feedback corrections each time the fork positioning optimization is performed is accumulated in the accumulator; then, the number of corrections in the accumulator is monitored, and whether the number of corrections exceeds a first preset number or the number of consecutive feedback corrections exceeds a second preset number is compared, wherein the first preset number is greater than the second preset number.

[0100] Specifically, if the number of corrections in the accumulator exceeds the first preset number or the number of consecutive feedback corrections exceeds the second preset number, a decision correction process is triggered, including adjusting parameters of the positioning optimization, such as navigation parameters and control algorithm parameters, according to the feedback correction records; improving the positioning optimization strategy, such as introducing a new algorithm or optimizing the path planning logic; then, the adjusted parameters and the optimized strategy are updated to the AGV control system to achieve more accurate fork positioning.

[0101] Through the above method, the positioning process of the AGV fork can be systematically monitored and optimized. By setting a preset number and making a decision correction according to the feedback correction records, unnecessary corrections are reduced, and the stability and accuracy of positioning are improved.

[0102] In some embodiments, the method further comprises:

[0103] obtaining a safety fence of a target space;

[0104] monitoring the relative distance between the target AGV fork and the safety fence in real time;

[0105] weighting the relative distance according to the safety fence level and the real-time task level of the target AGV, and if the result of the weighting is less than a preset safety threshold, automatically stopping the target AGV.

[0106] Specifically, the safety fence can be physical (such as an actually existing fence, obstacle, generated by real-time scanning of the environment by a sensor) or virtual (safety boundary extracted from a predefined map or scene model), representing an area that the AGV needs to avoid collision or approach during operation.

[0107] Specifically, safety fences have different fence levels, which indicate the safety importance or degree of danger of the fence. For example, low-level ordinary areas do not require special protection; high-level dangerous areas (such as next to high-temperature equipment, the edge of a deep pit, etc.) require higher safety guarantees.

[0108] Specifically, the real-time position and posture of the fork are obtained through sensors (such as encoders, IMUs, lidars, and vision systems), and then the minimum distance between the fork and the safety fence is calculated based on the real-time position of the fork and the position of the safety fence.

[0109] Specifically, the task level indicates the importance or urgency of the AGV's current task. Based on the safety fence level and the task level, a weighted correction is made to the relative distance. If the corrected distance is less than the safety threshold, it is considered that there is a safety risk and measures need to be taken, such as emergency stop of the target AGV. The safety threshold represents the minimum allowable distance between the fork and the safety fence.

[0110] By combining the safety fence level and task level, the relative distance between the AGV and the fence is weighted and corrected, and the dynamic adjustment mechanism automatically stops the AGV when a safety risk is detected. This improves operational safety while enabling flexible optimization of operational efficiency based on task requirements.

[0111] In summary, the AGV fork positioning method based on distance measurement and sensor analysis provided by the present invention has the following technical effects:

[0112] The method is implemented by setting a distance measuring sensor, a visual sensor and an azimuth sensor on the target AGV, and configuring and generating a sensor acquisition device group; activating the sensor acquisition device group to collect multi-dimensional distance measuring sensor data of the target AGV, wherein the multi-dimensional distance measuring sensor data includes distance data, image data and azimuth data; performing environmental position positioning based on the image data, and traversing the three-dimensional simulation space of the target scene according to the environmental position positioning result to extract the local reference space; determining the first relative spatial position of the fork of the target AGV in the local reference space by combining the distance data, image data and azimuth data; parsing the local reference space to obtain the second relative spatial position of the target object, and comparing the first relative spatial position with the second relative spatial position to obtain differential positioning information; optimizing the positioning of the fork of the target AGV according to the differential positioning information, thereby achieving the technical effect of improving the positioning efficiency of the fork and the safety of the operation.

[0113] Example 2

[0114] Figure 2 The schematic structural diagram of an exemplary electronic device provided for the present invention shows a block diagram of an exemplary electronic device suitable for implementing an embodiment of the present invention. Figure 2The electronic device shown is merely an example and should not impose any limitation on the functions and application scope of the embodiments of the present application. As Figure 2 As shown in the figure, the electronic device includes a processor 31, a memory 32, an input device 33, and an output device 34; the number of processors 31 in the electronic device can be one or more, Figure 2 In the foregoing description, the processor 31 in the electronic device, the memory 32, the input device 33, and the output device 34 are connected through a bus or other means, Figure 2 In the foregoing description, the processor 31 in the electronic device, the memory 32, the input device 33, and the output device 34 are connected through a bus or other means,

[0115] The memory 32 is a computer readable storage medium, which can be used to store software programs, computer executable programs, and modules, such as program instructions / modules corresponding to the AGV fork positioning method based on ranging sensor analysis in the embodiments of the present application. The processor 31 executes the software programs, instructions, and modules stored in the memory 32, thereby performing various function applications and data processing of the computer device, i.e., implementing the AGV fork positioning method based on ranging sensor analysis described above.

[0116] Embodiment Three

[0117] The embodiment provides a computer readable storage medium, which has a computer program stored thereon, and the computer program is executed by a processor to implement the following steps: setting a ranging sensor, a vision sensor, and a direction sensor on a target AGV, and configuring a generated sensor acquisition device group; activating the sensor acquisition device group to collect multi-dimensional ranging sensor data of the target AGV, wherein the multi-dimensional ranging sensor data includes distance data, image data, and azimuth angle data; performing environment position positioning based on the image data, and traversing a three-dimensional simulation space of a target scene according to an environment position positioning result to extract a local reference space; determining a first relative spatial position of a fork of the target AGV in the local reference space in combination with the distance data, the image data, and the azimuth angle data; analyzing the local reference space to obtain a second relative spatial position of a target object, and comparing the first relative spatial position and the second relative spatial position to obtain difference positioning information; and performing fork positioning optimization of the target AGV according to the difference positioning information.

[0118] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0119] It should be understood that the embodiments disclosed herein and the foregoing description thereof are merely exemplary in nature and, thus, that various changes in the details thereof can be implemented by those skilled in the art without departing from the spirit and scope of the present application. Such changes are intended to fall within the scope of the present application as defined by the appended claims.

Claims

1. The AGV fork positioning method based on distance measurement and sensor analysis is characterized by: The method comprises: Set up distance sensors, visual sensors, and orientation sensors on the target AGV, and configure and generate a sensor acquisition device group, including: Determine sensor specifications based on the positioning requirements of the target scene; According to the sensor specification information, the interactive target AGV performs matching screening on existing sensors and defines a sensor deployment list based on the matching screening results; Update the sensor layout of the target AGV using the sensor layout list, and connect the distance sensor, vision sensor and orientation sensor to the sensor acquisition sink to obtain the sensor acquisition device group; Activate the sensor acquisition device group to collect multi-dimensional ranging sensor data of the target AGV, wherein the multi-dimensional ranging sensor data includes distance data, image data and azimuth angle data; Performing environmental position positioning based on the image data, and traversing the three-dimensional simulation space of the target scene according to the environmental position positioning result to extract a local reference space, including: Performing image denoising and image enhancement on the image data, and extracting key feature information from the image; Performing environmental feature recognition based on the key feature information using a computer vision method to generate an environmental feature vector; Traversing the environmental feature vector database of the target scene for matching, and determining the environmental position positioning result of the target AGV; Combining the environmental position positioning result with the preset space extraction constraint, performing local space extraction in the three-dimensional simulation space to obtain the local reference space; Determine a first relative spatial position of a fork of a target AGV in the local reference space by combining the distance data, the image data, and the azimuth data; parsing the local reference space to obtain a second relative spatial position of the target object, and comparing the first relative spatial position with the second relative spatial position to obtain difference positioning information; The fork positioning of the target AGV is optimized based on the differential positioning information.

2. The AGV fork positioning method based on distance measurement and sensor analysis according to claim 1, characterized in that: Determining a first relative spatial position of a fork of a target AGV in the local reference space by combining the distance data, the image data, and the azimuth data includes: Using the azimuth data as the first registration basis, a first-level registration is performed in combination with the initial spatial orientation coordinate system of the target AGV; Using the distance data as a second registration basis, performing position correction on the first registration result based on triangulation to obtain a second registration result; Parsing the image data, extracting contour information and texture information of the target object as auxiliary registration basis, and verifying the secondary registration result by using the auxiliary registration basis; If the verification passes, the first relative spatial position is determined based on the secondary registration result and the relative position of the target AGV fork within the machine.

3. The AGV fork positioning method based on distance measurement and sensor analysis according to claim 2, characterized in that: The target object is the operation target object of the target AGV fork, including a transport target object and a placement target object, and the second relative spatial position is the relative spatial position of the operation task control point corresponding to the target object in the local reference space.

4. The AGV fork positioning method based on distance sensing analysis according to claim 1, characterized in that: The method further comprises: Recording the target AGV's acquisition of fork operation records, and screening and determining non-empty operation areas from the fork operation records, and outputting the result as a feedback monitoring area set; The feedback monitoring area set is continuously monitored by the scene monitoring device of the target scene, and feedback correction is performed on the target object according to the continuous monitoring result.

5. The AGV fork positioning method based on distance measurement and sensor analysis according to claim 4, characterized in that: The method further comprises: The number of feedback corrections is accumulated through the accumulator; If the number of feedback corrections exceeds a first preset number or the number of continuous feedback corrections exceeds a second preset number, the decision on fork positioning optimization is corrected according to the feedback correction record, wherein the first preset number is greater than the second preset number.

6. The AGV fork positioning method based on distance measurement and sensor analysis according to claim 1, characterized in that: The method further comprises: Get the security fence of the target space; Real-time monitoring of the relative distance between the target AGV fork and the safety fence; The relative distance is weighted and corrected according to the level of the safety fence and the real-time task level of the target AGV. If the correction result is less than a preset safety threshold, the target AGV is automatically stopped.

7. An electronic device, characterized in that: The electronic device comprises: a memory for storing executable instructions; The processor is configured to implement the AGV fork positioning method based on ranging sensor analysis as described in any one of claims 1 to 6 when executing the executable instructions stored in the memory.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the AGV fork positioning method based on distance measurement sensor analysis as described in any one of claims 1 to 6 is implemented.

Images