A Method and System for Optimizing Digital Driving Control Commands Based on Multi-Sensor Fusion

Abstract

This invention provides a method and system for optimizing digital vehicle control commands based on multi-sensor fusion. The method includes: generating initial vehicle control commands based on the operational digital signals of various vehicle operating mechanisms in response to magnetic encoder responses; identifying obstacles using 3D point cloud data of the working environment collected by lidar to obtain obstacle location information; identifying targets using image data of the working environment collected by a vision sensor to obtain load placement location information and target placement location information; optimizing the path planning of the initial vehicle control commands based on the obstacle location information, load placement location information, and target placement location information to obtain a first optimized vehicle control command; and controlling each vehicle operating mechanism based on the first optimized vehicle control command combined with real-time vehicle operation data collected by status sensors. This invention improves the safety and accuracy of vehicle operation.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method and system for optimizing digital driving control commands based on multi-sensor fusion. Background Technology

[0002] In industrial production and logistics, overhead cranes, as core lifting equipment, are traditionally controlled primarily by manual on-site operation combined with signal acquisition from a single sensor. This signal is then transmitted to the control unit via multi-core cables to execute corresponding actions. The main drawback of this traditional control method is low precision in executing operating commands. Specifically, the environmental or equipment status data collected by a single sensor has limited dimensions, making it difficult to fully reflect the actual working conditions. This results in insufficient matching between the operating commands generated by the control unit and actual operational needs, significant load positioning deviations, potential safety risks, and an inability to meet the precision requirements of modern industry. Summary of the Invention

[0003] The present invention provides a digital vehicle control command optimization method and system based on multi-sensor fusion, which aims to achieve accurate positioning and smooth transfer of vehicle load, and improve the safety and accuracy of vehicle operation.

[0004] In a first aspect, the present invention provides a method for optimizing digital vehicle control commands based on multi-sensor fusion, applicable to digital vehicles, wherein the digital vehicle integrates multiple sensors, including magnetic encoders, lidar, vision sensors, and status sensors; the method includes:

[0005] The operation digital signals of each traveling mechanism are responded to by the magnetic encoder, and an initial traveling control command is generated based on the operation digital signals.

[0006] Obstacle identification is performed based on the 3D point cloud data of the working environment collected by LiDAR to obtain the location information of obstacles in the working area. Target identification is performed based on the image data of the working environment collected by the vision sensor to obtain the load placement location information and target placement location information in the working area.

[0007] Based on the obstacle location information, the load placement location information, and the target placement location information, the initial driving control command is optimized by path planning to obtain the first optimized driving control command;

[0008] Based on the first optimized driving control command and the real-time driving operation data of each driving mechanism collected by the status sensor, the driving mechanism is controlled.

[0009] Secondly, the present invention also provides a digital vehicle control command optimization system based on multi-sensor fusion, used to implement the digital vehicle control command optimization method based on multi-sensor fusion as described in the first aspect; applied to digital vehicles, wherein the digital vehicle integrates multiple sensors, including magnetic encoders, lidar, vision sensors, and status sensors; the system includes:

[0010] The operation response module is used to respond to the operation digital signals of each traveling mechanism based on the magnetic encoder, and to generate initial traveling control commands based on the operation digital signals.

[0011] The object recognition module is used to identify obstacles based on the 3D point cloud data of the working environment collected by the LiDAR, and to obtain the location information of obstacles in the working area. It also performs target recognition based on the image data of the working environment collected by the vision sensor, and obtains the load placement location information and target placement location information in the working area.

[0012] The instruction optimization module is used to optimize the path planning of the initial driving control instruction based on the obstacle location information, the load placement location information and the target placement location information, so as to obtain the first optimized driving control instruction.

[0013] The control module is used to control each traveling mechanism based on the first optimized traveling control command and the real-time traveling operation data of each traveling mechanism collected by the status sensor.

[0014] Thirdly, the present invention provides an electronic device, comprising: a memory for storing computer software programs; and a processor for reading and executing the computer software programs, thereby realizing the digital driving control command optimization method based on multi-sensor fusion as described above.

[0015] Fourthly, the present invention provides a non-transitory computer-readable storage medium storing a computer software program, which, when executed by a processor, implements the digital vehicle control instruction optimization method based on multi-sensor fusion as described above.

[0016] Fifthly, the present invention provides a computer program product, including a computer program, which, when executed by a processor, implements the above-mentioned digital driving control instruction optimization method based on multi-sensor fusion.

[0017] The digital vehicle control command optimization method based on multi-sensor fusion provided in this invention optimizes the path planning of the initial vehicle control command by using obstacle location information obtained from LiDAR and load placement and target placement information obtained from visual sensors. This multi-sensor collaboration achieves comprehensive perception, avoiding obstacle risks and realizing optimal path planning from the current position to the target position. It avoids the problems of unreasonable command paths, collisions, or excessively long paths caused by the limited data dimensions of a single sensor. Furthermore, the method dynamically adjusts the first optimized vehicle control command based on the real-time operating status judgment results of each operating mechanism. This allows for real-time monitoring of the execution accuracy of the vehicle command, ensuring the stability of vehicle operation during command execution. It solves the problem of low execution accuracy of operating commands, achieves precise positioning and smooth transfer of the vehicle load, reduces collision risks and operational errors, and improves the safety and accuracy of vehicle operations. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the digital vehicle control command optimization method based on multi-sensor fusion provided in an embodiment of the present invention.

[0019] Figure 2 This is a schematic diagram of the structure of the digital driving control command optimization system based on multi-sensor fusion provided in an embodiment of the present invention;

[0020] Figure 3 An embodiment diagram of the electronic device provided in this invention;

[0021] Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. "A plurality of" means two or more, unless otherwise explicitly specified.

[0024] Optionally, see Figure 1 , Figure 1 This is a flowchart illustrating the digital vehicle control command optimization method based on multi-sensor fusion provided by the present invention. In this embodiment of the invention, the executing subject of the digital vehicle control command optimization method based on multi-sensor fusion is a digital vehicle. Therefore, the digital vehicle control command optimization method based on multi-sensor fusion includes:

[0025] Step 10: Based on the operation digital signals of each traveling mechanism in response to the magnetic encoder, and based on the operation digital signals, generate initial traveling control commands.

[0026] Optionally, the digital vehicle integrates a perception layer module, a control layer module, a cloud layer module, and a power supply module; the perception layer module, control layer module, and cloud layer module are sequentially connected in communication, and the power supply module provides power to each module; the perception layer module includes a magnetic encoder, a lidar, a vision sensor, and a status sensor.

[0027] Optionally, the digital crane uses the AS5600 magnetic encoder in its sensing layer module as the core component to respond to the digital signals of operation from various crane operating mechanisms (including the main trolley operating mechanism, the auxiliary trolley operating mechanism, and the hoisting mechanism), and generates initial crane control commands based on these digital signals. The AS5600 magnetic encoder, integrated into the digital crane's operating unit, uses a design where a permanent magnet and encoder are magnetically coupled to identify the operating angle. This operating unit employs a switchless design, effectively avoiding accuracy degradation caused by physical wear. Combined with a dustproof and waterproof sealed structure, it is suitable for industrial environments such as industrial plants and logistics warehouses where dust and humidity are present. When the operator operates the operating unit (e.g., the operating handle), the angle of the operating handle changes, and the permanent magnet coupled to the AS5600 magnetic encoder shifts position accordingly. The AS5600 magnetic encoder senses this magnetic change and converts the mechanical angle change of the operating handle into a corresponding digital signal, which is the operating angle digital signal.

[0028] Secondly, the operating digital signals include operating angle digital signals and operating speed digital signals. The operating angle digital signal is directly generated by the AS5600 magnetic encoder and reflects the direction and angle range of the crane's movement as desired by the operator. For example, if the operator rotates the operating handle 30° eastward towards the crane's traveling mechanism, the AS5600 magnetic encoder will generate an operating angle digital signal corresponding to the eastward direction and a 30° angle. The operating speed digital signal is generated based on the operating angle digital signal and the preset speed mapping rules within the digital crane. These preset speed mapping rules are set according to the performance parameters of each traveling mechanism (such as maximum motor power and maximum operating speed) and the general requirements of industrial operations. For example, when the operating angle is set to 0°-10°, the corresponding speed indicated by the operating speed digital signal is 1 m / s; when the operating angle is set to 11°-20°, the corresponding speed is 2 m / s, and so on.

[0029] Optionally, after the AS5600 magnetic encoder generates the operating angle digital signal, the control layer module receives the signal and automatically generates the corresponding operating speed digital signal according to the preset speed mapping rules, thus obtaining the operating digital signal. Further, after acquiring the operating digital signal, the control layer module performs a preliminary validity check on the signal. The check includes whether the operating angle digital signal is within the allowable angle operating range of each operating mechanism and whether the operating speed digital signal is within the allowable speed operating range of each operating mechanism. If the check passes, the control layer module generates an initial trolley control command based on the operating digital signal. This command is used to initially instruct each trolley operating mechanism to act according to the direction, angle, and speed desired by the operator. For example, the initial trolley control command might be "Control the trolley operating mechanism to move eastward at a speed of 2 m / s, and simultaneously control the hoisting mechanism to rise at a speed of 1 m / s."

[0030] In one embodiment, an operator controls the trolley's running mechanism and lifting mechanism of a digital crane to perform a goods-grabbing operation. The operator needs to control the trolley's running mechanism to move towards the shelf and control the lifting mechanism to descend to grab goods. When the operator rotates the operating handle controlling the trolley's running mechanism in the operating unit 15° towards the shelf, the AS5600 magnetic encoder integrated in the operating unit senses the angle change of the operating handle through the magnetic coupling design of the permanent magnet and the encoder, and generates a digital signal of the operating angle corresponding to the shelf direction and the 15° angle. Subsequently, the control layer module receives this digital signal of the operating angle and, according to a preset speed mapping rule (assuming the rule is that the operating angle of 0°-10° corresponds to a speed of 1m / s, and 11°-20° corresponds to a speed of 2m / s), generates a digital signal of operating speed of 2m / s. At this time, the digital signal of operation includes the digital signal of the operating angle of 15° towards the shelf and the digital signal of operating speed of 2m / s.

[0031] Next, the operator rotates the control handle of the hoisting mechanism downwards by 8°. The corresponding AS5600 magnetic encoder generates a downward direction, 8° angle operation digital signal. The control layer module generates an operation speed digital signal of 0.5m / s according to the speed mapping rule (assuming that 0°-10° corresponds to a speed of 0.5m / s), forming the operation digital signal corresponding to the hoisting mechanism.

[0032] The control layer module verifies the validity of the two operation digital signals mentioned above, confirming that the 15° operating angle of the trolley traveling mechanism is within its allowable operating range of 0°-30°, and the speed of 2m / s is within its allowable operating range of 0-3m / s; the 8° operating angle of the lifting mechanism is within its allowable operating range of 0°-20°, and the speed of 0.5m / s is within its allowable operating range of 0-1m / s. After the verification is passed, the initial trolley control command is generated: "Control the trolley traveling mechanism to move towards the shelf at a speed of 2m / s, and control the lifting mechanism to move downward at a speed of 0.5m / s."

[0033] Step 20: Obstacle identification is performed based on the 3D point cloud data of the working environment collected by the lidar to obtain the obstacle location information in the working area, and target identification is performed based on the image data of the working environment collected by the vision sensor to obtain the load placement location information and target placement location information in the working area.

[0034] Optionally, the lidar is deployed on the bottom of the digitized crane vehicle. Its function is to scan the working environment and collect three-dimensional point cloud data of the working environment. The three-dimensional point cloud data is generated by the laser beam emitted by the lidar reflecting back after contacting objects (including obstacles, loads, shelves, etc.) in the working environment. The lidar calculates the three-dimensional spatial coordinates of each point on the object by recording information such as the emission angle and propagation time of the laser beam. The set of three-dimensional spatial coordinates constitutes the three-dimensional point cloud data of the working environment.

[0035] Optionally, after receiving the 3D point cloud data acquired by the LiDAR, the control layer module processes the data using a preset obstacle recognition algorithm. This preset obstacle recognition algorithm is based on point cloud clustering and feature extraction techniques. First, it denoises the 3D point cloud data, removing noisy point clouds caused by LiDAR measurement errors, ambient light interference, and other factors. Then, it performs cluster analysis on the denoised point cloud data, aggregating point clouds with similar spatial locations and features into a cluster, with each cluster representing an object in the working environment. Next, it extracts features from each point cluster, including its volume, shape, and spatial location range. Finally, it compares the extracted features with pre-stored data within the digital vehicle. The obstacle feature template (which is pre-established based on the features of common obstacles in industrial operation scenarios, such as columns, other equipment, and piled-up debris) is compared. If the feature of a certain point cloud cluster matches the obstacle feature template with a preset threshold (the preset threshold is usually set to above 80%, which can be adjusted according to the complexity of the operation environment), then the object corresponding to the point cloud cluster is determined to be an obstacle. The center coordinates and outline range of the obstacle are calculated based on the three-dimensional spatial coordinates of the point cloud cluster, thus obtaining the obstacle location information in the operation area.

[0036] Optionally, the vision sensor is also deployed on the bottom of the digital vehicle and works in conjunction with the lidar. Its function is to collect image data of the working environment. The image data of the working environment is two-dimensional image information generated by the vision sensor through its internal image sensor (such as CCD or CMOS sensor) converting the light signals in the working environment into electrical signals, and then through analog-to-digital conversion, image processing and other steps. This image information can clearly reflect the appearance, color, texture and other features of objects in the working environment.

[0037] Optionally, after receiving the operational environment image data collected by the vision sensor, the control layer module processes it using a preset target recognition algorithm. This preset target recognition algorithm, based on deep learning and image feature matching technologies, first preprocesses the operational environment image data, including image deblurring, image enhancement, and color correction, to improve image clarity and quality. Then, it performs target region detection on the preprocessed image, using methods such as sliding windows and region proposal to filter candidate regions from the image that may contain loads or target placement locations (such as specific shelf locations). Next, it extracts features from each candidate region, including target shape features (such as the cuboid shape of the load or the rectangular shape of the shelf location), color features (such as the specific color of the load packaging or the color of the shelf location label), and texture features (such as the pattern on the load surface or the material texture of the shelf surface). Finally, it compares the extracted features with pre-stored load feature templates and target placement location feature templates (the load feature template is pre-established based on common cargo appearance features, while the target placement location feature template is pre-established based on the appearance and labeling features of shelf locations).

[0038] If the matching degree between the candidate region features and the load feature template exceeds a preset threshold (e.g., above 85%), the candidate region is determined to be the location of the load. Based on the coordinates of the region in the image and the installation parameters of the vision sensor (e.g., installation height, shooting angle, focal length, etc.), the three-dimensional spatial coordinates of the load in the working area are calculated, thus obtaining the load placement location information. If the matching degree between the candidate region features and the target placement location feature template exceeds a preset threshold, the candidate region is determined to be the target placement location, and its three-dimensional spatial coordinates are calculated, thus obtaining the target placement location information.

[0039] In one embodiment, the work area includes pillars (obstacles), goods to be transferred (loads), and designated storage shelves (target placement locations). A lidar unit deployed on the bottom of the vehicle scans the work area, collecting 3D point cloud data containing objects such as pillars, goods, and shelves. After receiving this data, the control layer module first performs denoising to remove noisy point clouds; then it performs cluster analysis, aggregating pillars, goods, and shelves into different point cloud clusters; finally, it extracts the features of each point cloud cluster. The pillar point cloud cluster exhibits cylindrical shape, large volume, and high height characteristics, and is compared with a pre-stored obstacle feature template (cylindrical shape, volume greater than 1m). 3 The column was compared with a height greater than 3m. The matching degree reached 90%, and it was determined to be an obstacle. Based on its point cloud data, the center coordinates of the column were calculated to be (10m, 5m, 4m), and the outline range was 1m in diameter and 4m in height, thus obtaining the obstacle location information.

[0040] Optionally, the vision sensor simultaneously acquires image data of the work area. The control layer module preprocesses the images, removing blurry parts and enhancing image contrast. Then, target region detection is performed, identifying two regions in the image that match candidate region characteristics: one region is cuboid in shape with a blue surface (goods packaging color), and the other region is rectangular with yellow markings (shelf location markings). Next, features of these two candidate regions are extracted, and the cuboid shape and blue surface candidate region features are compared with a pre-stored load feature template (cuboid, blue surface color, dimensions 2m*1m*). The image was compared with a 1m image, and the matching degree was 88%, which was determined to be the load. Based on its coordinates in the image and the installation parameters of the vision sensor (installation height 5m, shooting angle 30°, focal length 10mm), the three-dimensional spatial coordinates of the load were calculated as (5m, 3m, 1m), and the load placement location information was obtained. The rectangular shape and yellow-marked candidate area features were compared with the pre-stored target placement location feature template (rectangle, yellow marking on the surface, size 2.5m*1.5m), and the matching degree was 92%, which was determined to be the target placement location. Its three-dimensional spatial coordinates were calculated as (15m, 8m, 2m), and the target placement location information was obtained.

[0041] Step 30: Optimize the path planning of the initial driving control command based on obstacle location information, load placement location information, and target placement location information to obtain the first optimized driving control command.

[0042] Optionally, the control layer module uses the approximate running direction planned by the initial driving control command as a basis, combined with obstacle location information, to determine whether the initial running path will collide with the obstacle. If there is a risk of collision, it replans an optimal running path from the load placement position to the target placement position based on the load placement position information and the target placement position information, which can avoid the obstacle.

[0043] Optionally, the control layer module adjusts and optimizes the running speed and running angle in the initial driving control command according to the replanned optimal running path, and generates the first optimized driving control command to ensure that the driving mechanism can safely complete the operation task, as in steps 301 to 304.

[0044] Step 40: Based on the first optimized driving control command and the real-time driving operation data of each driving mechanism collected by the status sensor, control each driving mechanism.

[0045] Optionally, the control layer module receives real-time travel data (including vibration frequency, bearing temperature, motor current, travel speed, and travel position of each travel mechanism) collected by multi-dimensional state sensors in the perception layer module. Then, based on the real-time travel data, the control layer module determines the operating status of each travel mechanism and decides whether to adjust the first optimized travel control command. If no adjustment is needed, the drive submodule directly drives and controls each travel mechanism using the first optimized travel control command. If adjustment is required, the drive submodule drives and controls each travel mechanism using the adjusted travel control command, as described in steps 401 to 404.

[0046] This invention utilizes obstacle location information acquired by LiDAR and load and target placement information acquired by visual sensors to optimize the path planning of initial vehicle control commands. Therefore, comprehensive perception is achieved through multi-sensor collaboration, avoiding obstacle risks and realizing optimal path planning from the current position to the target position. This avoids the problems of unreasonable command paths, collision risks, or excessively long paths caused by the limited data dimensions of a single sensor. Furthermore, the optimized vehicle control commands are dynamically adjusted based on the real-time operating status of each mechanism, enabling real-time monitoring of the execution accuracy during command execution. This ensures the stability of vehicle operation during command execution, solves the problem of low command execution accuracy, achieves precise positioning and smooth transfer of the vehicle load, reduces collision risks and operational errors, and improves the safety and accuracy of vehicle operations.

[0047] Optionally, the processes of steps 301 to 304 include:

[0048] Step 301: Construct obstacle constraint boundaries using obstacle location information for each obstacle; construct an initial straight path using load placement location information and target placement location information; perform spatial interference prediction based on obstacle constraint boundaries and the initial straight path for each obstacle; and identify target obstacles that have spatial interference with at least one path point in the initial straight path.

[0049] Optionally, the control layer module constructs obstacle constraint boundaries based on the obstacle location information of each obstacle. The obstacle location information includes the center coordinates and outline range of the obstacle. The obstacle constraint boundary refers to the virtual boundary formed by extending a preset boundary compensation distance outward from the obstacle based on the outline range of the obstacle. The preset boundary compensation distance is a fixed value set according to the maximum external dimensions of the digital crane operating mechanism (such as the width of the trolley operating mechanism, the length of the trolley operating mechanism, the diameter of the hoisting mechanism, etc.) and industrial operation safety standards, for example, set to 0.5 meters, to avoid the collision risk caused by obstacle position measurement errors.

[0050] Furthermore, the control layer module constructs an initial straight-line path based on the load placement location information (including the load's three-dimensional spatial coordinates) and the target placement location information (including the target placement location's three-dimensional spatial coordinates). The initial straight-line path refers to the path formed by drawing a straight line through two points in three-dimensional space, with the center coordinates of the load placement location as the starting point and the center coordinates of the target placement location as the ending point. This path is used to initially indicate the direction of travel of the digital crane from the load location to the target placement location.

[0051] Furthermore, spatial interference prediction is performed based on the obstacle constraint boundaries and the initial straight path for each obstacle. Spatial interference prediction refers to determining, through spatial geometric calculations, whether any path point on the initial straight path is within the spatial range enclosed by the obstacle constraint boundaries of a certain obstacle. If the three-dimensional spatial coordinates of a path point satisfy the spatial range conditions of the obstacle constraint boundaries (i.e., the x, y, and z coordinates of the path point are all within the x-axis, y-axis, and z-axis ranges corresponding to the obstacle constraint boundaries), then it is determined that the path point has spatial interference with the obstacle; if at least one path point in the initial straight path has spatial interference with an obstacle, then that obstacle is identified as the target obstacle.

[0052] In one embodiment, the obstacle (column) has the following obstacle location information: center coordinates (10 meters, 5 meters, 4 meters), outline range of diameter 1 meter and height 4 meters, load placement location information is center coordinates (5 meters, 3 meters, 1 meter), and target placement location information is center coordinates (15 meters, 8 meters, 2 meters).

[0053] The control layer module constructs obstacle constraint boundaries: the preset boundary compensation distance is 0.5 meters, and the outline range of the column is 9.5 meters-10.5 meters in the x-axis direction (center x-coordinate 10 meters ± 0.5 meters radius), 4.5 meters-5.5 meters in the y-axis direction (center y-coordinate 5 meters ± 0.5 meters radius), and 0 meters-4 meters in the z-axis direction (height 0 meters-4 meters). After expanding the boundary compensation distance, the obstacle constraint boundary ranges from 9 meters to 11 meters in the x-axis direction, 4 meters to 6 meters in the y-axis direction, and 0 meters to 4.5 meters in the z-axis direction.

[0054] Next, an initial straight path is constructed: starting from the center coordinates of the load placement position (5m, 3m, 1m) and ending at the center coordinates of the target placement position (15m, 8m, 2m), a straight line is formed in three-dimensional space. The coordinates of any path point on this line can be calculated using the equation of the line. For example, the coordinates of path point 1 are (7m, 4m, 1.2m), the coordinates of path point 2 are (9m, 5m, 1.4m), and the coordinates of path point 3 are (11m, 6m, 1.6m), etc.

[0055] Then, spatial interference prediction is performed: it is determined whether each path point is within the obstacle constraint boundary (x: 9m-11m, y: 4m-6m, z: 0m-4.5m). For path point 2 (9m, 5m, 1.4m), the x-coordinate of 9m falls within the 9m-11m range, the y-coordinate of 5m falls within the 4m-6m range, and the z-coordinate of 1.4m falls within the 0m-4.5m range. Therefore, it is determined that path point 2 has spatial interference with the obstacle, and thus the pillar is identified as the target obstacle.

[0056] Step 302: For each target obstacle, connect the path points that have spatial interference in the initial straight path to obtain the initial path intersection line segment, and extend the preset driving safety distance in a direction perpendicular to the initial path intersection line segment as the central axis to obtain the spatial safety buffer zone.

[0057] Optionally, for each target obstacle, the control layer module filters out all path points that spatially interfere with the target obstacle from the initial straight path. These path points are those determined to be within the obstacle constraint boundary of the target obstacle in the spatial interference prediction step 301, and their three-dimensional spatial coordinates all satisfy the spatial range conditions of the obstacle constraint boundary. Next, the control layer module connects all the filtered path points with spatial interference to obtain the initial path intersection segment. The initial path intersection segment refers to the line segment formed by connecting the first and last interfering path points along the direction of the initial straight path among all the path points that spatially interfere with the target obstacle. This line segment represents the part of the initial straight path that spatially interferes with the target obstacle.

[0058] Then, the control layer module extends a preset safe travel distance in a direction perpendicular to the initial path intersection segment, using the initial path intersection segment as the central axis, to obtain a spatial safety buffer zone. The direction perpendicular to the initial path intersection segment includes all directions forming a 90-degree angle with the initial path intersection segment within a plane perpendicular to the plane containing the initial path intersection segment. The preset safe travel distance is a fixed value set based on the vibration offset, load sway amplitude, and industrial operation safety redundancy requirements that may occur during the operation of the digital crane's operating mechanism. For example, it is set to 0.8 meters to ensure that the digital crane's operating mechanism and load maintain a safe distance from the target obstacle when navigating around it, thus avoiding collisions. The spatial safety buffer zone is a cylindrical virtual area with the initial path intersection segment as the central axis and the preset safe travel distance as the radius.

[0059] In one embodiment, a preset safe driving distance is set to 0.8 meters, the target obstacle is a pillar, and path points in the initial straight path that have spatial interference with the target obstacle are selected, including path point A (9 meters, 5 meters, 1.4 meters), path point B (9.5 meters, 5.25 meters, 1.45 meters), path point C (10 meters, 5.5 meters, 1.5 meters), path point D (10.5 meters, 5.75 meters, 1.55 meters), and path point E (11 meters, 6 meters, 1.6 meters). These path points are arranged sequentially along the direction of the initial straight path.

[0060] The control layer module connects the first interference path point (path point A) and the last interference path point (path point E) along the initial straight path direction to form the initial path intersection line segment. The starting coordinates of this line segment are (9 meters, 5 meters, 1.4 meters) and the ending coordinates are (11 meters, 6 meters, 1.6 meters).

[0061] Next, using the initial path intersection segment as the central axis, extend 0.8 meters in all directions perpendicular to this segment to form a spatial safety buffer zone. This spatial safety buffer zone is a cylindrical virtual area, with the central axis of the cylinder being the initial path intersection segment (starting point (9 meters, 5 meters, 1.4 meters), ending point (11 meters, 6 meters, 1.6 meters)). The radius of the cylinder is 0.8 meters, and the length of the cylinder is the same as the length of the initial path intersection segment (calculated using the two-point distance formula, the length of the initial path intersection segment is approximately 2.24 meters, therefore the length of the cylinder is approximately 2.24 meters). The spatial range of this cylindrical area is the area that subsequent path planning needs to avoid.

[0062] Step 303: Based on the load placement location information and the target placement location information, determine the initial path direction, and with the spatial safety buffer as the constraint boundary and the starting point of the initial path intersection segment as the center, determine the candidate detour path area based on the initial path direction and the preset length parameter along the intersection segment.

[0063] Optionally, the control layer module determines the initial path direction based on the load placement location information and the target placement location information. The initial path direction refers to the direction from the load placement location to the target placement location in three-dimensional space. This direction can be determined by calculating the vector between the center coordinates of the load placement location and the center coordinates of the target placement location. For example, if the center coordinates of the load placement location are (X1, Y1, Z1) and the center coordinates of the target placement location are (X2, Y2, Z2), then the initial path direction vector is (X2-X1, Y2-Y1, Z2-Z1), and the direction of this vector is the initial path direction.

[0064] Then, the control layer module uses the spatial safety buffer as the constraint boundary (i.e., the subsequently planned path must not enter the spatial safety buffer), takes the starting point of the initial path intersection segment obtained in step 302 as the center, and determines the candidate detour path area based on the initial path direction and the preset length parameter along the intersection segment. The preset intersection segment length parameter refers to the length extended along the length direction of the initial path intersection segment to both sides of the starting point of the initial path intersection segment. This value is set according to the size of the spatial safety buffer zone, the spatial range of the work area, and the turning flexibility of the digital gantry crane. For example, it is set to 1.5 times the length of the initial path intersection segment. The candidate detour path area refers to the feasible area in three-dimensional space, centered on the starting point of the initial path intersection segment, along the initial path direction, where the digital gantry crane can detour around the target obstacle while avoiding the spatial safety buffer zone. The boundary of this area includes the boundary of the spatial safety buffer zone, the extension line of the initial path direction, and the range defined by the preset intersection segment length parameter. The candidate detour path area must ensure that when the digital gantry crane plans a path within this area, it can start from the load placement position, detour around the target obstacle, and continue to run towards the target placement position.

[0065] In one embodiment, the center coordinates of the load placement location in step 20 are (5m, 3m, 1m), the center coordinates of the target placement location are (15m, 8m, 2m), the starting coordinates of the initial path intersection line segment in step 302 are (9m, 5m, 1.4m), and the spatial safety buffer zone is a cylindrical area (center axis starting point (9m, 5m, 1.4m), ending point (11m, 6m, 1.6m), radius 0.8m).

[0066] The control layer module determines the initial path direction: the initial path direction vector is calculated as (15m-5m, 8m-3m, 2m-1m) = (10m, 5m, 1m), and the direction pointed to by this vector is the initial path direction (from (5m, 3m, 1m) to (15m, 8m, 2m)).

[0067] Next, set the preset length parameter along the intersection segment: the initial path intersection segment length in step 302 is approximately 2.24 meters, therefore the preset length parameter along the intersection segment is 2.24 meters * 1.5 ≈ 3.36 meters. Taking the starting point (9 meters, 5 meters, 1.4 meters) of the initial path intersection segment as the center, extend along the initial path direction (vector (10 meters, 5 meters, 1 meter) direction) forward (towards the target placement position) and backward (towards the load placement position) by 3.36 meters each. At the same time, use the spatial safety buffer zone (cylindrical area) as the constraint boundary (the path must not enter this cylindrical area), and delineate a feasible area in three-dimensional space. The x-axis range of this area is approximately (9 meters - 3.36 meters * (10 / )) to (9 meters + 3.36 meters * (10 / The y-axis range is approximately (5 meters - 3.36 meters * (5 / )) to (5 meters + 3.36 meters * (5 / The z-axis range is approximately (1.4 meters - 3.36 meters * (1 / )) to (1.4 meters + 3.36 meters * (1 / Furthermore, this area must avoid the cylindrical range of the spatial safety buffer zone, and the resulting feasible area is the candidate detour path area.

[0068] Step 304: Based on the intersection of the spatial boundary of the work area and the candidate detour path area of ​​each target obstacle, the target detour path area is obtained. Based on the target detour path area of ​​each target obstacle, the initial driving control command is optimized by path planning to obtain the first optimized driving control command.

[0069] Optionally, the control layer module obtains the spatial boundary of the work area. The spatial boundary of the work area refers to the spatial range boundary in which the digital crane can perform operations. This boundary is preset based on the actual physical dimensions of the work area (such as the length, width, and height of the factory) and work safety regulations. For example, the spatial boundary of the work area may have an x-axis range of 0-20 meters, a y-axis range of 0-10 meters, and a z-axis range of 0-5 meters. All operating actions of the digital crane must not exceed this boundary range.

[0070] Then, the control layer module performs intersection calculations on the spatial boundary of the work area and the candidate detour path area for each target obstacle to obtain the target detour path area. Intersection calculation refers to taking the shared spatial range of both the candidate detour path area and the spatial boundary of the work area. That is, the target detour path area must simultaneously satisfy the condition of being within both the candidate detour path area and the spatial boundary of the work area, ensuring that the detour path planned by the digital vehicle within this area can both avoid the target obstacle and not exceed the spatial range of the work area. In one embodiment, the preset spatial boundary of the work area is 0-20 meters on the x-axis, 0-10 meters on the y-axis, and 0-5 meters on the z-axis. The candidate detour path area obtained in step 303 has an x-axis range of approximately 5.7-12.3 meters, a y-axis range of approximately 2.7-7.3 meters, and a z-axis range of approximately 1.1-1.7 meters. The control layer module calculates the intersection of the work area's spatial boundary and the candidate detour path area: the candidate detour path area's x-axis range of 5.7m-12.3m is entirely within the work area's x-axis range of 0m-20m, its y-axis range of 2.7m-7.3m is entirely within the work area's y-axis range of 0m-10m, and its z-axis range of 1.1m-1.7m is entirely within the work area's z-axis range of 0m-5m. Therefore, the intersection of these two ranges constitutes the candidate detour path area (x: 5.7m-12.3m, y: 2.7m-7.3m, z: 1.1m-1.7m), which is the target detour path area.

[0071] Furthermore, the control layer module optimizes the initial driving control command based on the target detour path area of ​​each target obstacle to obtain the first optimized driving control command, as described in steps 3041 to 3045.

[0072] This invention, through identifying interference obstacles, delineating avoidance areas, limiting feasible path ranges, and optimizing the logic chain of control commands, ensures the rationality and efficiency of the operation path while guaranteeing the safe operation of the digital crane (avoiding obstacles and not exceeding the operation boundary). This enables the optimized first-order crane control command to accurately guide the digital crane to complete the operation task from load grabbing to target placement, thereby improving the automated operation accuracy and safety performance of the digital crane.

[0073] Optionally, the process of steps 3041 to 3045 includes:

[0074] Step 3041: For each target obstacle, construct path segments based on the order of load placement location information, detour start point, detour end point, and target placement location information to obtain candidate detour paths. The detour start point is the starting point of the intersection line segment of the initial path, and the detour end point is any point within the target detour path area. The detour end point satisfies the following conditions: it is outside the safety buffer zone, and the line connecting it to the detour start point is perpendicular to the direction of the initial path.

[0075] Optionally, the detour starting point is the starting point of the intersection line segment of the initial path, which is the first path point in the initial straight path that spatially interferes with the target obstacle; the detour ending point is any point within the target detour path area, and must satisfy two constraints: first, the detour ending point is outside the spatial safety buffer zone constructed in step 302 (i.e., the three-dimensional coordinates of the detour ending point do not fall within the cylindrical virtual area of ​​the spatial safety buffer zone); second, the direction of the line connecting the detour ending point and the detour starting point is perpendicular to the initial path direction determined in step 303 (the initial path direction is the vector direction from the load placement position to the target placement position).

[0076] Optionally, the control layer module constructs path segments in the order of "load placement location information → detour start point → detour end point → target placement location information". Specifically, the path segment "load placement location information → detour start point" is a line segment from the center coordinates of the load placement location to the detour start point, used to instruct the digital vehicle to travel from the load location to the detour start point; the path segment "detour start point → detour end point" is a line segment from the detour start point to the detour end point, used to instruct the digital vehicle to detour around the target obstacle; and the path segment "detour end point → target placement location information" is a line segment from the detour end point to the center coordinates of the target placement location, used to instruct the digital vehicle to travel from the detour end point to the target placement location.

[0077] Therefore, the complete path formed by connecting these three path segments in sequence is the candidate path for bypassing obstacles, and at least one candidate path for bypassing obstacles is generated for each target obstacle.

[0078] Step 3042: Based on the discrete points on each candidate path of the bypass action, traverse all obstacles to determine whether any discrete point belongs to its obstacle constraint boundary.

[0079] Optionally, the control layer module discretizes each candidate bypass path by uniformly selecting multiple discrete points on each segment of the candidate bypass path. The distance between any two adjacent discrete points is less than half of a preset safe driving distance, such as the 0.8 meters set in step 302. Therefore, the distance between adjacent discrete points must be less than 0.4 meters. This setting ensures that the discrete points densely cover the candidate bypass path, avoiding potential interference risks on the path due to excessively large spacing between discrete points.

[0080] Next, the control layer module traverses all obstacles (including target obstacles and non-target obstacles) and obtains the obstacle constraint boundary of each obstacle (the obstacle constraint boundary is the virtual boundary constructed in step 301 with the boundary compensation distance extended).

[0081] Then, it is determined one by one whether each discrete point on the candidate path of the around motion belongs to the obstacle constraint boundary of any obstacle. The determination method is to check whether the three-dimensional coordinates of the discrete point fall within the x-axis range, y-axis range, and z-axis range of the obstacle constraint boundary. If the x-coordinate of the discrete point is between the minimum and maximum values ​​of the x-axis of the obstacle constraint boundary, the y-coordinate is between the minimum and maximum values ​​of the y-axis, and the z-coordinate is between the minimum and maximum values ​​of the z-axis, then the discrete point is determined to belong to the obstacle constraint boundary of the obstacle.

[0082] Step 3043: For each candidate path for bypassing the obstacle, if none of its discrete points belong to the obstacle constraint boundary of any obstacle, then it is determined as the first target path for bypassing the obstacle.

[0083] Optionally, the control layer module summarizes and analyzes the judgment results of all discrete points for each candidate bypass path. If all discrete points on a candidate bypass path are determined to be obstacle constraint boundaries that do not belong to any obstacle, meaning that no discrete point on the candidate path falls within the constraint boundaries of any obstacle, then the candidate path will not spatially interfere with any obstacle during the entire driving process, meeting the safe driving requirements, and the control layer module determines the candidate bypass path as the first target bypass path. If at least one discrete point on a candidate bypass path belongs to the obstacle constraint boundary of any obstacle, then the candidate path has a collision risk and must be eliminated, not included in the subsequent optimization scope.

[0084] Step 3044: Based on the maximum and minimum angle changes of the angle control parameters of each traveling mechanism at all time points during the execution of each first bypass target path, determine the total angle change of each traveling mechanism in each first bypass target path.

[0085] Optionally, the angle control parameters are parameters used to control the rotation angle of each traveling mechanism (trolley traveling mechanism, trolley traveling mechanism, and hoisting mechanism), with each traveling mechanism corresponding to an independent angle control parameter; the time nodes are multiple moments divided according to fixed time intervals during the process of the digital traveling machine executing the first orbital action target path. The time interval is set according to the traveling speed and path length, for example, set to 0.1 seconds, to ensure that changes in the angle control parameters can be accurately captured; the maximum angle change value is the difference between the maximum and minimum values ​​of the angle control parameter of a certain traveling mechanism at all time nodes; the minimum angle change value is the minimum value of the change in the angle control parameter of a certain traveling mechanism between two adjacent moments at all time nodes (here, minimum refers to the minimum absolute value of the change).

[0086] Next, for each first bypass target path, the control layer module obtains the angle control parameters of each traveling mechanism at all time points and calculates the maximum angle change value for each traveling mechanism (i.e., the maximum value minus the minimum value of the angle control parameter for that mechanism). Then, based on the variation pattern of the angle control parameters and the mechanical performance of the traveling mechanism, the control layer module determines the minimum angle change value (usually set to the minimum angle change amount that satisfies the smooth operation of the mechanism, such as 0.1 degrees). Finally, the total angle change of each traveling mechanism in each first bypass target path is calculated using the formula: Total Angle Change = Maximum Angle Change + (Total Number of Time Points - 1) * Minimum Angle Change. The total angle change reflects the total amplitude of angle control by the traveling mechanism during the path execution.

[0087] Step 3045: Based on the total angle change of each traveling mechanism in each first orbital action target path, the initial traveling control command is optimized by path planning to obtain the first optimized traveling control command.

[0088] Optionally, the control layer module optimizes the initial train control command based on the total angle change of each train operating mechanism in each first orbital action target path to obtain the first optimized train control command, as in steps 30451 to 30454.

[0089] This invention, through a logical chain of constructing candidate paths, comprehensive safety detection, screening safe paths, and quantifying the merits of paths, enables digital vehicles to complete operations with minimal adjustments to the operating mechanism and the smoothest operation while avoiding all obstacles, thereby improving the safety and stability of vehicle operation.

[0090] Optionally, the processes of steps 30451 to 30454 include:

[0091] Step 30451: For each first orbital target path, if the total angle change of all corresponding traveling mechanisms is less than or equal to their respective angle change limits, then it is determined as the second orbital target path.

[0092] Optionally, the control layer module obtains the total angle change of each traveling mechanism (including the trolley traveling mechanism, the hoisting mechanism, and the crane mechanism) in each first bypassing target path. The total angle change refers to the total change in the angle control parameter of the corresponding traveling mechanism from the starting angle to the ending angle during the execution of the first bypassing target path. This value reflects the angle adjustment range of the traveling mechanism during the path execution process.

[0093] Next, the control layer module calls the internally stored angle change limit values ​​for each traveling mechanism. The angle change limit value refers to the maximum allowable change value of the angle control parameter of the traveling mechanism during a single operation, set according to the mechanical structural strength, motor drive capability, and industrial operation safety standards of the traveling mechanism. For example, the angle change limit value of the main trolley traveling mechanism is set to 60 degrees, the angle change limit value of the auxiliary trolley traveling mechanism is set to 45 degrees, and the angle change limit value of the hoisting mechanism is set to 30 degrees. If the total angle change exceeds the angle change limit value, it may lead to damage to the mechanical components of the traveling mechanism or a decrease in operational stability.

[0094] Then, the control layer module filters each first bypass operation target path: it compares the total angle change of each traveling mechanism under the path with the corresponding angle change limit value. If the total angle change of all traveling mechanisms is less than or equal to their respective angle change limit values, it means that the first bypass operation target path meets the performance constraints and safety requirements of the traveling mechanisms in terms of angle adjustment, and the digital traveling mechanism can safely execute the path. If the total angle change of any traveling mechanism is greater than its corresponding angle change limit value, it is determined that the first bypass operation target path does not meet the safety operation requirements and is eliminated. The first bypass operation target path that meets the requirements is determined as the second bypass operation target path.

[0095] Step 30452: Based on the line segment length of each path segment in each second bypass target path and the maximum speed, deceleration coefficient and deceleration time of each train operating mechanism in different path segments in the initial train control command, the running time is obtained, and the total running time of each second bypass target path is determined based on the running time of each path segment.

[0096] Optionally, the control layer module breaks down each second detour target path into all the path segments contained in that path. A path segment refers to the line segment constructed in step 3041, which is formed by connecting the load placement location information, the detour start point, the detour end point, and the target placement location information in sequence. Each second detour target path typically includes three path segments: the path segment from the load placement location to the detour start point, the path segment from the detour start point to the detour end point, and the path segment from the detour end point to the target placement location.

[0097] Next, the control layer module calculates the line segment length for each path segment. The line segment length refers to the straight-line distance between the two endpoints of the path segment (such as the load placement location and the detour start point, the detour start point and the detour end point, etc.) in three-dimensional space. It can be calculated using the distance formula between two points in three-dimensional space. This length reflects the distance that the traveling mechanism needs to move in this path segment.

[0098] Then, the control layer module obtains the maximum speed, deceleration coefficient, and deceleration time of each traveling mechanism in different path segments from the initial traveling control command. The maximum speed refers to the highest permissible operating speed of the traveling mechanism in the corresponding path segment, which is preset by the traveling machine's motor performance and operational requirements. The deceleration coefficient is the proportionality of speed reduction during deceleration, ranging from 0 to 1; for example, a deceleration coefficient of 0.1 indicates a 10% speed reduction. The deceleration time is the time required for the traveling mechanism to decrease from its maximum speed to a stable operating speed, which is preset by the traveling machine's braking system performance.

[0099] Furthermore, the control layer module calculates the running time of each path segment according to the formula running time = line segment length / (maximum speed * (1 - deceleration coefficient * deceleration time)). The formula in this embodiment of the invention takes into account the impact of the deceleration process of the traveling mechanism on the actual running speed during the running of the path segment, so that the calculated running time is more in line with the actual operation.

[0100] Furthermore, after calculating the running time of each path segment, the control layer module adds up the running times of all path segments under the same second detour target path to obtain the total running time of the second detour target path. The total running time reflects the total time required for the crane to complete the entire task by executing the path. The total running time is a key indicator reflecting the execution efficiency of each second detour target path. The shorter the total running time, the higher the efficiency of the crane in completing the task by executing the path, and the faster the operation process from load grabbing to target placement can be realized.

[0101] Step 30453: Determine the second orbital target path with the shortest total running time as the optimal orbital path.

[0102] Optionally, the control layer module compares and sorts the total running time of all second orbital action target paths. The sorting process is carried out in ascending order of total running time, that is, the path with the shortest total running time is ranked first, and the paths with longer total running times are arranged in order.

[0103] Then, the control layer module selects the second bypass target path with the shortest total running time. This path, while satisfying the limits of the gantry mechanism's angle change and avoiding obstacles, has the highest work execution efficiency, enabling the task to be completed in the shortest time, reducing waiting time, and improving overall production efficiency. Finally, the control layer module determines the selected second bypass target path with the shortest total running time as the optimal bypass path.

[0104] Step 30454: Based on the angle control command sequence and speed control command sequence of each traveling mechanism corresponding to the optimal path of the bypass operation, and combined with the operation logic of each traveling mechanism in the initial traveling control command, update the angle control command sequence and speed control command sequence of each traveling mechanism in the initial traveling control command to obtain the first optimized traveling control command.

[0105] Optionally, the control layer module extracts the angle control command sequence and speed control command sequence for each traveling mechanism corresponding to the optimal path of travel. The angle control command sequence refers to the sequence of angle control parameters set for each traveling mechanism at different time points during the execution of the optimal path of travel; the speed control command sequence refers to the sequence of speed control parameters set for each traveling mechanism at different time points during the execution of the optimal path of travel. These two sequences together determine the action state of the traveling mechanism under the optimal path.

[0106] Next, the control layer module obtains the operational logic of each trolley operating mechanism from the initial trolley control command. The operational logic refers to the action coordination rules between each trolley operating mechanism set in the initial trolley control command, such as "the main trolley operating mechanism starts first, and the auxiliary trolley operating mechanism starts after the main trolley operating mechanism reaches a stable speed" and "after the lifting mechanism has finished grabbing the load, it needs to maintain a stable height until it reaches above the target placement position." The logic ensures the coordination of the actions of each operating mechanism and the safety of the operation.

[0107] Then, the control layer module uses the angle control command sequence and speed control command sequence around the optimal path of movement as a benchmark, and combines the operational logic of the initial trolley control command to update the angle control command sequence and speed control command sequence of each trolley operating mechanism in the initial trolley control command. During the update process, it is necessary to ensure that the new command sequence conforms to the action coordination rules of each mechanism in the operational logic. For example, if the operational logic requires the hoisting mechanism to adjust its speed after the trolley has stabilized, then the updated hoisting mechanism speed control command sequence must match the speed stabilization time node of the trolley operating mechanism.

[0108] Finally, the command generated after the update is the first optimized driving control command. This command not only meets the efficiency requirements of the optimal path of the bypass operation, but also meets the coordination requirements of the initial operation logic, and can guide the digital driving to complete the operation task safely and efficiently.

[0109] In one embodiment, the optimal path for the bypass action is path 1, and the instruction sequence corresponding to path 1 is extracted:

[0110] The sequence of angle control commands for the trolley traveling mechanism is as follows: 0 degrees at 0 seconds → 20 degrees at 1 second → 55 degrees at 2.5 seconds → 55 degrees at 3.5 seconds → 55 degrees at 6.92 seconds;

[0111] Speed ​​control command sequence for the trolley traveling mechanism: 0 m / s at 0 seconds → 2 m / s at 1 second → 2 m / s at 2.5 seconds → 2 m / s at 3.5 seconds → 0 m / s at 6.92 seconds;

[0112] Hoisting mechanism angle control command sequence: Maintain 0 degrees throughout (no angle adjustment);

[0113] Hoisting mechanism speed control command sequence: Maintain 0 m / s throughout (no height adjustment).

[0114] The initial driving control command's operating logic is: "After the main trolley traveling mechanism starts and reaches a stable speed (2 m / s), the trolley traveling mechanism starts and adjusts its position."

[0115] The control layer module updates the initial command sequence in conjunction with the operation logic: In the initial command, the trolley running mechanism was originally planned to start at 0 seconds, but after the update, it is adjusted to start at 1 second (the time node when the trolley reaches a stable speed). The speed control command sequence of the trolley running mechanism is updated to 0 m / s at 0 seconds → 1 m / s at 1 second → 1 m / s at 2.5 seconds → 0 m / s at 3.5 seconds → 0 m / s at 6.92 seconds. The angle control command sequence is updated synchronously to 0 degrees at 0 seconds → 10 degrees at 1 second → 40 degrees at 2.5 seconds → 40 degrees at 3.5 seconds → 40 degrees at 6.92 seconds.

[0116] After the update is completed, the first optimized driving control command is generated. This command includes the updated sequence of angle control commands and speed control commands for each operating mechanism, and conforms to the action coordination logic.

[0117] The embodiments of the present invention obtain the first optimized crane control command through a logical link of safety constraint screening, efficiency quantification calculation, optimal path selection, and instruction collaborative update. This not only avoids the risk of mechanical overload but also maximizes the work efficiency, while ensuring the coordination of the actions of each operating mechanism and improving the safe operation level of the digital crane in complex working environments.

[0118] Optionally, the processes of steps 401 to 404 include:

[0119] Step 401: Determine the operating status of each train operation mechanism based on the real-time train operation data.

[0120] Optionally, the control layer module receives real-time travel data from various traveling mechanisms (including the main trolley traveling mechanism, the auxiliary trolley traveling mechanism, and the hoisting mechanism) collected by status sensors. Status sensors are sensing devices deployed at key locations on each traveling mechanism (such as motor outputs, transmission gear sets, and braking systems). Their function is to collect physical quantity data reflecting the working status of the traveling mechanism in real time. Real-time travel data includes, but is not limited to, real-time motor current, real-time voltage, real-time speed, real-time bearing temperature, real-time braking system pressure, real-time position coordinates, and real-time angle parameters of each traveling mechanism. This data directly reflects the current working status of the traveling mechanism.

[0121] Next, the control layer module calls the pre-stored operating status judgment thresholds for each crane traveling mechanism. These operating status judgment thresholds are set based on the design parameters, rated performance, industrial safety operation standards, and long-term operational experience data of each crane traveling mechanism. They are numerical ranges or critical values ​​used to determine whether the traveling mechanism is operating normally. For example, the normal current threshold range for the trolley traveling mechanism motor is 5-15 amps, the normal bearing temperature threshold range is 20-60 degrees Celsius, the normal speed threshold range for the trolley traveling mechanism is 1-5 revolutions per second, and the normal pressure threshold range for the hoisting mechanism's braking system is 0.8-1.2 MPa. If the real-time crane operating data exceeds the corresponding threshold range, it indicates that the traveling mechanism may be malfunctioning.

[0122] Then, the control layer module compares and analyzes the real-time train operation data of each train operation mechanism with the corresponding operation status judgment threshold one by one: for each real-time data of each operation mechanism, it determines whether it is within the corresponding normal threshold range. If all real-time train operation data of a certain operation mechanism are within the corresponding normal threshold range, the operation mechanism is determined to be in normal operation; if at least one real-time train operation data of a certain operation mechanism exceeds the corresponding normal threshold range, the operation mechanism is determined to be in abnormal operation, thus determining the operation status of each train operation mechanism.

[0123] Step 402: If each traveling mechanism is in normal operation, then drive control is executed based on the first optimized traveling control command to operate each traveling mechanism.

[0124] Optionally, the control layer module confirms whether the operating status of each crane operating mechanism is normal based on the analysis results of the operating status of each crane operating mechanism.

[0125] If it is confirmed that all crane operating mechanisms (trolley operating mechanism, trolley operating mechanism, and hoisting mechanism) are in normal operating condition, it indicates that the current working status of each operating mechanism is stable and has the conditions to execute the operation task according to the optimized control instructions, and there is no operation risk caused by mechanism abnormality.

[0126] Next, the control layer module extracts the first optimized train control command determined in step 30454. The first optimized train control command includes the angle control command sequence and speed control command sequence of each train operating mechanism, as well as the operational logic for the coordinated action of each mechanism. This command is the optimal command after path optimization, efficiency screening and coordination adjustment, which can guide the digital train to complete the operation task safely and efficiently.

[0127] Then, the control layer module decomposes the first optimized trolley control command into specific drive commands for each trolley operating mechanism. The specific drive commands are converted from the angle control command sequence and speed control command sequence into electrical or hydraulic signal commands that can directly drive the actuators (such as motors, hydraulic valves, brake solenoids, etc.) according to the control interface protocol and action execution requirements of each operating mechanism. For example, the command "the trolley operating mechanism moves eastward at a speed of 2 m / s and adjusts the angle to 55 degrees" is converted into a motor drive voltage signal, a speed control signal, and an angle adjustment servo drive signal.

[0128] Finally, the control layer module sends the decomposed specific drive instructions to the execution drive modules of each traveling mechanism. After receiving the instructions, the execution drive modules drive the corresponding traveling mechanisms to perform actions according to the angle and speed parameters set in the instructions, thereby realizing the drive control of each traveling mechanism and promoting the work tasks to be executed according to the optimized path and efficiency.

[0129] Step 403: If there is a first target operating mechanism in an abnormal operating state, the abnormal time node is determined, and based on the first optimized train control command, the first optimized train control command is optimized for the target action command sequence of the first target operating mechanism at the abnormal time node to obtain the second optimized train control command.

[0130] Optionally, if there is an abnormal operating state, the operating mechanism in the abnormal operating state shall be identified as the first target operating mechanism (if there are multiple abnormal operating mechanisms, each abnormal mechanism shall be an independent first target operating mechanism).

[0131] Next, the control layer module uses the real-time data timestamps collected by the status sensors to determine the abnormal time point when the first target operating mechanism enters an abnormal operating state. The abnormal time point refers to the specific time when one or more real-time driving operation data of the first target operating mechanism first exceeds the corresponding operating state judgment threshold. For example, if the status sensor collects that the hoisting mechanism motor current reaches 16 amps at 3 seconds (exceeding the normal threshold of 6A-14A), then the abnormal time point is 3 seconds, which marks the beginning of the abnormality of the first target operating mechanism.

[0132] Furthermore, the control layer module optimizes the first optimized driving control command based on the target action command sequence of the first target operating mechanism at an abnormal time point, to obtain the second optimized driving control command, as described in steps 4041 to 4045.

[0133] Step 404: Drive the second target operating mechanism to operate based on the second optimized driving control command.

[0134] Optionally, the second target operating mechanism refers to the operating mechanism that is in normal operation except for the first target operating mechanism. For example, if the first target operating mechanism is the hoisting mechanism, then the second target operating mechanism is the trolley operating mechanism and the trolley operating mechanism. The real-time operating data of these mechanisms are all within the normal threshold range and have the conditions to execute actions according to instructions.

[0135] Next, the control layer module has coordinated the action command sequence (angle control command sequence and speed control command sequence) of the second target operating mechanism according to the abnormal situation of the first target operating mechanism to ensure that the action of the second target operating mechanism will not cause logical conflicts or safety risks due to the abnormality of the first target operating mechanism. For example, if the hoisting mechanism of the first target operating mechanism is suspended, the speed command of the trolley operating mechanism in the second optimized trolley control command will be adjusted to ensure that the trolley can still run stably during the suspension of the hoisting mechanism and avoid load swaying.

[0136] Then, the control layer module decomposes the sequence of action instructions for the second target operating mechanism in the second optimized vehicle control instructions into specific drive instructions that can directly drive the actuators of the second target operating mechanism. The decomposition process must follow the control protocol and execution characteristics of the second target operating mechanism. For example, the instruction of the trolley operating mechanism "3 seconds - 6.92 seconds, angle maintained at 55 degrees, speed maintained at 2 m / s" is converted into an electrical signal instruction with a motor drive voltage of 220 volts and an angle servo maintained at 55 degrees.

[0137] Finally, the control layer module sends the decomposed specific drive instructions to the execution drive module of the second target running mechanism. After receiving the instructions, the execution drive module drives the second target running mechanism to perform actions according to the angle and speed parameters set in the instructions, thereby realizing the drive control of the second target running mechanism. This ensures that even if the first target running mechanism is abnormal, the normally operating second target running mechanism can still continue to perform the key links of the operation task, reducing the impact of the abnormality on the overall operation.

[0138] The embodiments of the present invention achieve dynamic and precise control of the digital crane operating mechanism through a logical chain of state perception, normal driving, abnormal optimization, and partial execution. This ensures efficient execution of operations under normal working conditions and enables rapid response and instruction optimization when mechanism abnormalities occur, thereby improving the operational stability of the digital crane in complex operating environments.

[0139] Optionally, the process of steps 4041 to 4045 includes:

[0140] Step 4041: Based on the operational logic relationship between various train operation mechanisms, determine the associated operation mechanism that has a direct logical relationship with the first target operation mechanism, and based on the first optimized train control command, determine the associated action command sequence for the associated operation mechanism at the abnormal time point and the time node after the abnormal time point.

[0141] Optionally, the control layer module invokes the pre-stored operational logic relationships between various crane operating mechanisms. These operational logic relationships refer to the logical relationships set according to the digital crane's operational process, the functional division of each operating mechanism, and the requirements for action coordination. They reflect the dependencies or cooperation between different operating mechanisms. For example, "the crane operating mechanism can only start its horizontal movement after the lifting mechanism completes the load grabbing action (lifting the load to a preset height)," and "the lifting mechanism can only perform its lowering action after the trolley operating mechanism is adjusted to a position directly above the load." This relationship ensures that the actions of each mechanism are orderly and coordinated, avoiding action conflicts that could lead to operational failures or safety accidents.

[0142] Next, based on the aforementioned operational logic relationships, the control layer module filters out operational mechanisms that have a direct logical relationship with the first target operational mechanism and identifies them as associated operational mechanisms. A direct logical relationship means that the initiation or execution of an operational mechanism's action must be premised on a specific action of the first target operational mechanism, or the action of the first target operational mechanism must depend on the action of the first operational mechanism. For example, if the first target operational mechanism is a hoisting mechanism, and the operational logic relationship states that "the start of the trolley operational mechanism requires the hoisting mechanism to raise its load to a safe height," then the trolley operational mechanism is an associated operational mechanism; if the relationship states that "the descent of the hoisting mechanism requires the trolley operational mechanism to be positioned at the target location," then the trolley operational mechanism is also an associated operational mechanism.

[0143] Then, the control layer module extracts the first optimized driving control command determined in step 30454, filters out the time nodes at and after the abnormal time node, and identifies the action command sequence of the associated operating mechanism as the associated action command sequence. The associated action command sequence is a set of angle control commands and speed control commands that the associated operating mechanism should execute at and after the abnormal time node in the first optimized driving control command. For example, if the abnormal time node is 3 seconds, the action command sequence of the trolley operating mechanism (associated operating mechanism) in the first optimized driving control command from 3 seconds to 6.92 seconds is "keep the angle at 55 degrees and the speed at 2 meters per second". This sequence is the associated action command sequence.

[0144] Step 4042: Based on the triggering relationship between the target action instruction sequence and the associated action instruction sequence, construct an initial logical triggering dependency chain, and determine all affected first action instruction sequences in the initial logical triggering dependency chain with the target action instruction sequence as the termination execution constraint.

[0145] Optionally, the control layer module determines the triggering relationship between the target action instruction sequence and the associated action instruction sequence. The triggering relationship refers to the influence of the execution status (such as execution, pause, or termination) of the target action instruction sequence on the execution of the associated action instruction sequence based on the operational logic association. This includes types such as "the associated action instruction sequence can only start after the target action instruction sequence is executed", "the associated action instruction sequence needs to be paused synchronously when the target action instruction sequence is paused", and "the associated action instruction sequence needs to adjust its execution parameters when the target action instruction sequence is terminated". For example, if the operational logic association is "the start of the trolley traveling mechanism requires the hoisting mechanism load to be raised to a safe height as a prerequisite", then the triggering relationship is "the associated action instruction sequence (horizontal movement) of the trolley traveling mechanism can only start after the target action instruction sequence (raised to a safe height) of the hoisting mechanism is completed".

[0146] Next, the control layer module constructs an initial logical trigger dependency chain based on the aforementioned triggering relationships. This initial logical trigger dependency chain is a logical chain of related action command sequences arranged hierarchically according to the triggering relationship, with the target action command sequence as its core. Each node in the chain represents an action command sequence (either the target action command sequence or a related action command sequence). The connections between nodes represent the triggering relationship. For example, the target action command sequence (lifting mechanism) is the upstream node, and related action command sequence 1 (moving trolley operating mechanism) is the downstream node, connected by a line of "upstream execution completed → downstream start". Similarly, related action command sequence 1 (moving trolley operating mechanism) is the upstream node, and related action command sequence 2 (fine-tuning trolley operating mechanism) is the downstream node, connected by a line of "upstream start → downstream synchronous start", forming the initial logical trigger dependency chain of "target action command sequence → related action command sequence 1 → related action command sequence 2".

[0147] Then, the control layer module uses the target action command sequence as the termination constraint. That is, if the target action command sequence cannot be executed normally due to an abnormality in the first target operating mechanism, it needs to be terminated or paused. Based on the triggering relationship of the initial logic triggering dependency chain, it deduces all action command sequences affected by this termination constraint in reverse or forward, and determines them as the first action command sequence. For example, if the target action command sequence (lifting mechanism) needs to be terminated due to an abnormality, according to the triggering relationship "target action command sequence execution completed → associated action command sequence 1 started", associated action command sequence 1 (trolley operating mechanism movement) cannot be executed normally due to the lack of a starting prerequisite, and is an affected sequence; further, according to the triggering relationship "associated action command sequence 1 started → associated action command sequence 2 started synchronously", associated action command sequence 2 (trolley operating mechanism fine adjustment) is also affected because the upstream sequence cannot be started. Therefore, both associated action command sequence 1 and associated action command sequence 2 are first action command sequences.

[0148] Step 4043: Generate an initial standby command sequence for the associated operating mechanism based on the first action command sequence, and generate an emergency braking command sequence based on the abnormal state characteristics of the first target operating mechanism.

[0149] Optionally, for each first action command sequence, the control layer module, in conjunction with the real-time status data of the associated operating mechanism at the abnormal time node, generates an initial standby command sequence for the associated operating mechanism. The initial standby command sequence refers to the angle and speed control command sequence set to ensure that the associated operating mechanism remains in a safe and stable state after the first action command sequence is affected. Its core requirements are: the angle is maintained at the current value at the abnormal time node, that is, the actual angle parameter of the associated operating mechanism at the abnormal time node (e.g., 3 seconds), which is obtained through the real-time angle data collected by the status sensor at the abnormal time node; the speed is reduced to zero, that is, the speed of the associated operating mechanism is gradually reduced from the real-time speed at the abnormal time node to 0 m / s, to avoid load shaking or mechanism damage due to sudden stop braking.

[0150] For example, at the abnormal time point of 3 seconds, the real-time angle of the trolley traveling mechanism is 55 degrees and the real-time speed is 2 meters per second. Then its initial standby command sequence is "3 seconds to 3.5 seconds, the angle remains at 55 degrees and the speed drops from 2 meters per second to 0 meters per second; after 3.5 seconds, the angle remains at 55 degrees and the speed remains at 0 meters per second".

[0151] Next, the control layer module analyzes the abnormal state characteristics of the first target operating mechanism. The abnormal state characteristics refer to the specific manifestation attributes of the abnormality of the first target operating mechanism, including the abnormality type (such as abnormal motor current, abnormal bearing temperature, abnormal braking pressure), the degree of abnormality (such as exceeding the moderate threshold, exceeding the severe threshold), and the abnormal trend (such as abnormal data continuously rising, remaining stable, or slowly decreasing). For example, the abnormal state characteristics of the hoisting mechanism are "abnormal motor current, exceeding the slight threshold, and the current value remaining stable".

[0152] Then, an emergency braking command sequence for the first target operating mechanism is generated based on the abnormal state characteristics. The emergency braking command sequence refers to the sequence of angle and speed control commands set to prevent the first target operating mechanism from aggravating the abnormality and to ensure that it quickly enters a safe state. Its core requirements are: the angle is locked at the current value at the abnormal time node, that is, the real-time angle parameter of the first target operating mechanism at the abnormal time node. After locking, no angle adjustment is allowed to prevent angle changes from aggravating the load on the mechanism; the speed linearly drops to zero within a preset safe time. The preset safe time is a fixed time set according to the braking system performance, load weight and abnormality degree of the first target operating mechanism.

[0153] For example, when the anomaly is minor, the preset safety time is 1 second; when the anomaly is severe, the preset safety time is 0.5 seconds. A linear decrease to zero indicates that the speed decreases uniformly over time, avoiding braking impact. For example, if the hoisting mechanism has an abnormal time point of 3 seconds, with a real-time angle of 0 degrees and a real-time speed of 0 m / s (lifting not started), and the abnormal state characteristic is a slight current anomaly, and the preset safety time is 1 second, then the emergency braking command sequence is "3 seconds - 4 seconds, angle locked at 0 degrees, speed maintained at 0 m / s (no need to reduce speed since the initial speed is 0); after 4 seconds, angle locked at 0 degrees, speed maintained at 0 m / s"; if the hoisting mechanism's real-time speed is 0.3 m / s, then the emergency braking command sequence is "3 seconds - 4 seconds, angle locked at 0 degrees, speed linearly decreases from 0.3 m / s to 0 m / s; after 4 seconds, angle locked at 0 degrees, speed maintained at 0 m / s".

[0154] Step 4044: Based on the braking execution time interval of the emergency braking command sequence, the standby execution time range of the initial standby command sequence is calibrated to obtain the target standby command sequence.

[0155] Optionally, the control layer module determines the braking execution time interval of the emergency braking command sequence. The braking execution time interval refers to the time range during which the first target operating mechanism performs braking action (mainly the process of linearly reducing the speed from the real-time value to 0 m / s) in the emergency braking command sequence. If the initial speed of the first target operating mechanism is 0 m / s, the braking execution time interval is the interval from the abnormal time node to the end of the preset safe time; if the initial speed is not 0, it is the time interval from the abnormal time node to the speed reducing to 0 m / s. For example, if the emergency braking command sequence is "3 seconds - 3.8 seconds, angle locked at 0 degrees, speed linearly reduced from 0.3 m / s to 0 m / s", then the braking execution time interval is 3 seconds - 3.8 seconds.

[0156] Next, the control layer module analyzes the standby execution time range of the initial standby command sequence. The standby execution time range refers to the time range within the initial standby command sequence during which the associated operating mechanism, starting from the abnormal time node, reduces its speed to 0 m / s and maintains a standby state. This includes the speed reduction phase and the stable standby phase. For example, if the initial standby command sequence for the trolley operating mechanism is "3 seconds - 3.5 seconds, speed reduces from 2 m / s to 0 m / s; 3.5 seconds - 6.92 seconds, speed remains at 0 m / s", then the standby execution time range is 3 seconds - 6.92 seconds, where 3 seconds - 3.5 seconds is the speed reduction phase and 3.5 seconds - 6.92 seconds is the stable standby phase.

[0157] Then, the control layer module calibrates the standby execution time range of the initial standby command sequence based on the braking execution time interval. The core principle of calibration is that the speed descent phase of the associated operating mechanism must end synchronously with the braking execution time interval of the first target operating mechanism. This ensures that when the first target operating mechanism completes braking and enters a safe state, the associated operating mechanism also enters a stable standby state synchronously, avoiding the risk of uncoordinated actions caused by the associated operating mechanism stopping braking prematurely or delayed. For example, if the braking execution time interval is 3 seconds to 3.8 seconds (duration 0.8 seconds), and the speed descent phase of the initial standby command sequence of the trolley operating mechanism is 3 seconds to 3.5 seconds (duration 0.5 seconds), the speed descent phase needs to be extended to 3 seconds to 3.8 seconds, and the speed descent rate needs to be adjusted so that the speed drops to 0 m / s at 3.8 seconds. The stable standby phase starts from 3.8 seconds. The calibrated standby execution time range is 3 seconds to 6.92 seconds, where 3 seconds to 3.8 seconds is the speed descent phase, and 3.8 seconds to 6.92 seconds is the stable standby phase. The corresponding command sequence is the target standby command sequence.

[0158] Step 4045: Based on the first action instruction sequence, the initial logic trigger dependency chain, the emergency braking instruction sequence, and the target standby instruction sequence, optimize the first optimized driving control instruction to obtain the second optimized driving control instruction.

[0159] Optionally, the control layer module optimizes the first optimized driving control command based on the first action command sequence, the initial logic trigger dependency chain, the emergency braking command sequence, and the target standby command sequence to obtain the second optimized driving control command, as described in steps 40451 to 40453.

[0160] This invention achieves precise control of digital cranes when local mechanisms malfunction by identifying associated mechanisms, sorting out logical dependencies, generating safety instructions, calibrating time synchronization, and optimizing the overall progressive logic of instructions. This not only prevents the risk of abnormal mechanisms from spreading but also ensures the stable state of associated mechanisms, improves the safety and controllability of digital cranes under fault conditions, and guarantees a smooth transition in the operation process.

[0161] Optionally, the processes of steps 40451 to 40453 include:

[0162] Step 40451: Take the action command sequence other than the first action command sequence in the first optimized driving control command as the second action command sequence, and merge the second action command sequence, the emergency braking command sequence and the target standby command sequence according to the time axis to obtain the intermediate optimized command sequence.

[0163] Optionally, the control layer module selects the remaining action command sequences (excluding the first action command sequence) from all action command sequences included in the first optimized train control command and identifies them as the second action command sequence. The second action command sequence is the command sequence in the first optimized train control command that is not affected by the abnormality of the first target operating mechanism and can continue to be executed as originally planned. For example, if the first action command sequence is the command sequence of the trolley operating mechanism and the trolley operating mechanism, and the first optimized train control command only includes the command sequences of the hoisting mechanism, the trolley operating mechanism, and the trolley operating mechanism, then there is no second action command sequence; if there are command sequences of other auxiliary mechanisms (such as clamping mechanisms) that are not affected, then the command sequence of the auxiliary mechanism is the second action command sequence.

[0164] Next, the control layer module merges the second action command sequence, the emergency braking command sequence, and the target standby command sequence using the time axis as a reference. The time axis is a coordinate axis that arranges all command execution periods in chronological order, with abnormal time nodes as key reference points. The merging operation must adhere to the principles of "non-overlapping time intervals and non-conflicting command parameters": for different mechanism commands within the same time interval, ensure that the execution parameters (angle, speed) of each command do not cause conflict in mechanism actions; for commands from different time intervals, arrange them sequentially on the time axis to form a continuous sequence containing all commands to be executed; this sequence is the intermediate optimized command sequence. The intermediate optimized command sequence initially integrates the commands from unaffected mechanisms, abnormal mechanisms, and related mechanisms.

[0165] Step 40452: After the braking execution time interval ends, delete the logical trigger node corresponding to the target action instruction sequence in the initial logical trigger dependency chain, and adjust the trigger condition of the first action instruction sequence in the initial logical trigger dependency chain to take the completion of the emergency braking instruction sequence as a prerequisite, and generate the target trigger dependency chain.

[0166] Optionally, the control layer module determines the end time of the braking execution time interval. For example, if the braking execution time interval is 3 seconds to 3.8 seconds, then the end time is 3.8 seconds.

[0167] Next, the control layer module locates the corresponding logical trigger node in the chain based on the initial logical trigger dependency chain. A logical trigger node is the node representing the target action command sequence in the initial logical trigger dependency chain; this node is the key link in the chain where upstream commands trigger downstream commands. For example, if the initial logical trigger dependency chain is "hoisting mechanism target action command node → trolley traveling mechanism first action command node → trolley traveling mechanism first action command node," then the hoisting mechanism target action command node is the logical trigger node corresponding to the target action command sequence. Because the first target traveling mechanism malfunctioned, the target action command sequence has been replaced by the emergency braking command sequence, and the original triggering function can no longer be executed.

[0168] Therefore, after the braking execution time interval ends, the control layer module removes the logical trigger node from the initial logical trigger dependency chain, eliminating invalid trigger relationships.

[0169] Then, the control layer module adjusts the triggering conditions corresponding to the first action instruction sequence in the initial logical triggering dependency chain. The original triggering conditions refer to the preconditions required to start the first action instruction sequence in the initial logical triggering dependency chain (such as the completion of the target action instruction sequence). Since the target action instruction node has been deleted, the original triggering conditions need to be adjusted to "completion of the emergency braking instruction sequence." That is, the first action instruction sequence (actually replaced by the target standby instruction sequence) can only execute subsequent actions according to the target standby instruction sequence after the entire emergency braking instruction sequence has been executed. The adjusted logical triggering dependency chain is the target triggering dependency chain. This chain updates the triggering relationships to ensure that the execution order of each mechanism's instructions conforms to safety logic.

[0170] Step 40453: Based on the intermediate optimized instruction sequence and the target trigger dependency chain update instruction execution logic, the second optimized driving control instruction is obtained.

[0171] Optionally, the control layer module updates the execution logic of the intermediate optimized instruction sequence based on the target trigger dependency chain. The instruction execution logic refers to the start order, execution priority, and parameter coordination rules of each mechanism's instructions on the timeline. The update process must ensure that: the start order of all instructions strictly follows the triggering relationship of the target trigger dependency chain; for example, "the target standby instruction of the trolley operating mechanism can only be started after the emergency braking instruction sequence is completed"; the execution priority of instructions within the same time interval is set according to "safety instructions take precedence over action instructions," meaning that the execution priority of the emergency braking instruction and the target standby instruction is higher than the second action instruction (if there is a parameter conflict, the execution of the safety instruction is prioritized); simultaneously, it checks whether the parameters (angle, speed, execution duration) of each instruction match the triggering relationship, for example, whether the start time of the target standby instruction of the trolley operating mechanism is consistent with the completion time of the emergency braking instruction. If there is a deviation, the time parameters of the intermediate optimized instruction sequence are fine-tuned to ensure that the execution logic is completely consistent with the target trigger dependency chain.

[0172] Finally, the instruction sequence after the execution logic update is the second optimized driving control instruction, which includes all the mechanism instructions to be executed and clarifies the instruction execution order through the target trigger dependency chain.

[0173] This invention achieves a safe upgrade of the first optimized driving control command by integrating the instruction sequence, updating the trigger logic, and matching the execution rules. This not only ensures the safe braking of the first target operating mechanism, but also ensures the orderly operation of the associated and unaffected mechanisms, thereby improving the accuracy and safety of the command control of the digital driving system under abnormal operating conditions.

[0174] Furthermore, the digital driving control command optimization system based on multi-sensor fusion provided by the present invention will be described below. The digital driving control command optimization system based on multi-sensor fusion described below corresponds to the digital driving control command optimization method based on multi-sensor fusion described above.

[0175] Reference Figure 2 , Figure 2 This is a schematic diagram of the structure of the digital driving control command optimization system based on multi-sensor fusion provided by the present invention. The digital driving control command optimization system based on multi-sensor fusion includes:

[0176] The operation response module 210 is used to respond to the operation digital signals of each traveling mechanism based on the magnetic encoder, and to generate initial traveling control commands based on the operation digital signals.

[0177] The object recognition module 220 is used to identify obstacles based on the three-dimensional point cloud data of the working environment collected by the lidar, and to obtain the obstacle location information in the working area. It also performs target recognition based on the image data of the working environment collected by the vision sensor, and obtains the load placement location information and target placement location information in the working area.

[0178] The instruction optimization module 230 is used to optimize the path planning of the initial driving control instruction based on obstacle location information, load placement location information and target placement location information to obtain the first optimized driving control instruction.

[0179] The control module 240 is used to control each traveling mechanism based on the first optimized traveling control command and the real-time traveling operation data of each traveling mechanism collected by the status sensor.

[0180] The embodiments of the present invention achieve precise positioning and smooth transfer of vehicle loads, reduce collision risks and operational errors, and improve the safety and accuracy of vehicle operations.

[0181] Please see Figure 3 , Figure 3 An embodiment diagram of an electronic device provided in accordance with the present invention. For example... Figure 3 As shown, an embodiment of the present invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, it implements the processes of steps 10 to 40.

[0182] Please see Figure 4 , Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with an embodiment of the present invention is shown. Figure 4 As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, it implements the processes of steps 10 to 40.

[0183] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the digital driving control instruction optimization method based on multi-sensor fusion provided by the above methods, which includes steps 10 to 40.

[0184] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing digital vehicle control commands based on multi-sensor fusion, characterized in that, This method is applied to digital driving systems, which integrate multiple sensors, including magnetic encoders, lidar, vision sensors, and status sensors. The system responds to the operation digital signals of each traveling mechanism based on the magnetic encoder, and generates initial traveling control commands based on the operation digital signals. Obstacle identification is performed based on the 3D point cloud data of the working environment collected by LiDAR to obtain the location information of obstacles in the working area. Target identification is performed based on the image data of the working environment collected by the vision sensor to obtain the load placement location information and target placement location information in the working area. Based on obstacle location information, load placement information and target placement information, the initial driving control command is optimized by path planning to obtain the first optimized driving control command. Based on the first optimized driving control command combined with the real-time driving operation data of each driving mechanism collected by the status sensor, control each driving mechanism. include: The operating status of each train operation mechanism is determined based on real-time train operation data. If there is a first target operating mechanism in an abnormal operating state, then the abnormal time node is determined, and based on the first optimized train control command, the first optimized train control command is optimized for the target action command sequence of the first target operating mechanism at the abnormal time node to obtain the second optimized train control command. The second target operating mechanism is driven and controlled to operate based on the second optimized driving control command; the second target operating mechanism is the remaining operating mechanism among all driving operating mechanisms except for the first target operating mechanism. The steps to obtain the second optimized vehicle control command include: Based on the operational logic relationship between various train operation mechanisms, identify the associated operation mechanisms that have a direct logical relationship with the first target operation mechanism, and based on the first optimized train control instructions, determine the sequence of associated action instructions for the associated operation mechanisms at abnormal time nodes and time nodes after abnormal time nodes. Based on the triggering relationship between the target action instruction sequence and the associated action instruction sequence, an initial logical triggering dependency chain is constructed, and all affected first action instruction sequences in the initial logical triggering dependency chain are determined with the target action instruction sequence as the termination execution constraint. An initial standby command sequence for the associated operating mechanism is generated based on the first action command sequence, and an emergency braking command sequence is generated based on the abnormal state characteristics of the first target operating mechanism. The target standby command sequence is obtained by calibrating the standby execution time range of the initial standby command sequence based on the braking execution time range of the emergency braking command sequence. Based on the first action command sequence, the initial logic trigger dependency chain, the emergency braking command sequence, and the target standby command sequence, the first optimized driving control command is optimized to obtain the second optimized driving control command.

2. The method for optimizing digital vehicle control commands based on multi-sensor fusion according to claim 1, characterized in that, The control of each train operating mechanism based on the first optimized train control command combined with real-time train operation data collected by the status sensors includes: If each traveling mechanism is in normal operation, then the driving control of each traveling mechanism is executed based on the first optimized traveling control command.

3. The method for optimizing digital vehicle control commands based on multi-sensor fusion according to claim 1, characterized in that, The action command sequence includes an angle control command sequence and a speed control command sequence; the initial standby command sequence indicates that the angle remains at the current value at the abnormal time node, and the speed drops to zero; the emergency braking command sequence indicates that the angle is locked at the current value at the abnormal time node, and the speed linearly drops to zero within a preset safe time.

4. The method for optimizing digital vehicle control commands based on multi-sensor fusion according to claim 1, characterized in that, The first optimized driving control command is further optimized based on the first action command sequence, the initial logic trigger dependency chain, the emergency braking command sequence, and the target standby command sequence to obtain the second optimized driving control command, including: The action command sequence other than the first action command sequence in the first optimized driving control command is taken as the second action command sequence, and the second action command sequence, the emergency braking command sequence and the target standby command sequence are merged according to the time axis to obtain the intermediate optimized command sequence; After the braking execution time interval ends, delete the logical trigger node corresponding to the target action command sequence in the initial logical trigger dependency chain, and adjust the trigger condition of the first action command sequence in the initial logical trigger dependency chain to take the completion of the emergency braking command sequence as a prerequisite, and generate the target trigger dependency chain. Based on the intermediate optimized instruction sequence and the target triggered dependency chain update instruction execution logic, the second optimized driving control instruction is obtained.

5. The method for optimizing digital vehicle control commands based on multi-sensor fusion according to claim 1, characterized in that, The steps for obtaining the first optimized driving control command through path planning optimization include: An obstacle constraint boundary is constructed using the obstacle location information of each obstacle. An initial straight path is constructed using the load placement location information and the target placement location information. Spatial interference prediction is performed based on the obstacle constraint boundary of each obstacle and the initial straight path to determine the target obstacle that has spatial interference with at least one path point in the initial straight path. For each target obstacle, the path points that have spatial interference in the initial straight path are connected to obtain the initial path intersection line segment. Then, with the initial path intersection line segment as the central axis, a preset driving safety distance is extended in a direction perpendicular to the initial path intersection line segment to obtain the spatial safety buffer zone. Based on the load placement location information and the target placement location information, the initial path direction is determined, and with the spatial safety buffer as the constraint boundary and the starting point of the initial path intersection segment as the center, the candidate detour path area is determined based on the initial path direction and the preset length parameter along the intersection segment. The target detour path region is obtained by intersecting the regional spatial boundary of the work area with the candidate detour path region of each target obstacle. The initial driving control command is then optimized based on the target detour path region of each target obstacle to obtain the first optimized driving control command.

6. The method for optimizing digital vehicle control commands based on multi-sensor fusion according to claim 5, characterized in that, The initial driving control command is optimized by path planning based on the target detour path region for each target obstacle to obtain a first optimized driving control command, including: For each target obstacle, a path segment is constructed based on the order of the load placement location information, detour start point, detour end point, and target placement location information to obtain a detour candidate path; the detour start point is the starting point of the intersection line segment of the initial path, the detour end point is any point within the target detour path area, and the detour end point satisfies the condition that it is outside the safety buffer zone and the direction of the line connecting it to the detour start point is perpendicular to the direction of the initial path. Based on the discrete points on each candidate path of detour, all obstacles are traversed to determine whether any discrete point belongs to its obstacle constraint boundary; the distance between two adjacent discrete points is less than 1 / 2 of the set safe driving distance. For each candidate path for bypassing, if none of its discrete points belong to the obstacle constraint boundary of any obstacle, it is determined as the first target path for bypassing. Based on the maximum and minimum angle changes of the angle control parameters of each train operating mechanism at all time points during the execution of each first bypass target path, the total angle change of each train operating mechanism in each first bypass target path is determined. Based on the total angle change of each traveling mechanism in each first orbital action target path, the initial traveling control command is optimized by path planning to obtain the first optimized traveling control command.

7. The method for optimizing digital vehicle control commands based on multi-sensor fusion according to claim 6, characterized in that, The initial train control command is optimized by path planning based on the total angle change of each train operating mechanism in each first bypass target path to obtain the first optimized train control command, including: For each first orbital movement target path, if the total angle change of all corresponding train operating mechanisms is less than or equal to their respective angle change limit value, then it is determined as the second orbital movement target path. Based on the line segment length of each path segment in each second bypass target path, combined with the maximum speed, deceleration coefficient and deceleration time of each train operating mechanism in different path segments in the initial train control command, the running time is obtained, and the total running time of each second bypass target path is determined based on the running time of each path segment. The second orbital target path with the shortest total running time is determined as the optimal orbital path; Based on the angle control command sequence and speed control command sequence of each traveling mechanism corresponding to the optimal path of the bypass operation, and combined with the operation logic of each traveling mechanism in the initial traveling control command, the angle control command sequence and speed control command sequence of each traveling mechanism in the initial traveling control command are updated to obtain the first optimized traveling control command.

8. A digital vehicle control command optimization system based on multi-sensor fusion, characterized in that, For implementing the digital driving control command optimization method based on multi-sensor fusion as described in any one of claims 1 to 7; This system is applied to digital driving systems, which integrate various sensors, including magnetic encoders, LiDAR, vision sensors, and status sensors. The system includes: The operation response module is used to respond to the operation digital signals of each crane operating mechanism based on the magnetic encoder, and to generate initial crane control commands based on the operation digital signals; The object recognition module is used to identify obstacles based on the 3D point cloud data of the working environment collected by the LiDAR, and to obtain the location information of obstacles in the working area. It also performs target recognition based on the image data of the working environment collected by the vision sensor, and obtains the load placement location information and target placement location information in the working area. The instruction optimization module is used to optimize the path planning of the initial driving control instruction based on obstacle location information, load placement information and target placement information to obtain the first optimized driving control instruction. The control module is used to control each traveling mechanism based on the first optimized traveling control command and the real-time traveling operation data of each traveling mechanism collected by the status sensor. include: The operating status of each train operation mechanism is determined based on real-time train operation data. If there is a first target operating mechanism in an abnormal operating state, then the abnormal time node is determined, and based on the first optimized train control command, the first optimized train control command is optimized for the target action command sequence of the first target operating mechanism at the abnormal time node to obtain the second optimized train control command. The second target operating mechanism is driven and controlled to operate based on the second optimized driving control command; the second target operating mechanism is the remaining operating mechanism among all driving operating mechanisms except for the first target operating mechanism. The steps to obtain the second optimized vehicle control command include: Based on the operational logic relationship between various train operation mechanisms, identify the associated operation mechanisms that have a direct logical relationship with the first target operation mechanism, and based on the first optimized train control instructions, determine the sequence of associated action instructions for the associated operation mechanisms at abnormal time nodes and time nodes after abnormal time nodes. Based on the triggering relationship between the target action instruction sequence and the associated action instruction sequence, an initial logical triggering dependency chain is constructed, and all affected first action instruction sequences in the initial logical triggering dependency chain are determined with the target action instruction sequence as the termination execution constraint. An initial standby command sequence for the associated operating mechanism is generated based on the first action command sequence, and an emergency braking command sequence is generated based on the abnormal state characteristics of the first target operating mechanism. The target standby command sequence is obtained by calibrating the standby execution time range of the initial standby command sequence based on the braking execution time range of the emergency braking command sequence. Based on the first action command sequence, the initial logic trigger dependency chain, the emergency braking command sequence, and the target standby command sequence, the first optimized driving control command is optimized to obtain the second optimized driving control command.

9. An electronic device, comprising: Memory, used to store computer software programs; A processor for reading and executing the computer software program, characterized in that, when the processor executes the computer software program, it implements the digital driving control instruction optimization method based on multi-sensor fusion as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium, wherein a computer software program is stored therein, characterized in that, When the computer software program is executed by the processor, it implements the digital driving control command optimization method based on multi-sensor fusion as described in any one of claims 1 to 7.

Images