Intelligent warehousing system, stacking machine and control method of stacking machine

Abstract

The invention provides an intelligent warehousing system, a stacking machine and a control method of the stacking machine, and relates to the technical field of intelligent warehousing. The stacking machine is suitable for moving goods among the multiple goods grids; the stacker adopts frequency conversion control; the control method comprises the steps of obtaining an operation instruction and analyzing a target position; planning a motion path according to the current position and the target position of the stacker; wherein the path planning is based on a three-axis coupling dynamic model, the shortest total operation time is taken as a target function, and the dynamic parameter difference of each axis is comprehensively considered to optimize the speed curve matching relationship of each axis; generating a smooth speed curve for each motion axis according to the motion path, wherein the speed curve adopts a segmented continuous S-shaped acceleration and deceleration model; wherein if yes, it is judged that the stroke is a long stroke, and a standard seven-section S-shaped speed curve is adopted; if not, the stroke is short, and five or three sections of simplified S-shaped curves are adopted; and the stacking machine is moved to the target position according to the movement path and the speed curve.

Description

Technical Field

[0001] This invention relates to the field of intelligent warehousing technology, and more specifically, to an intelligent warehousing system, a stacker crane, and a control method thereof. Background Technology

[0002] Traditional factories typically store materials and products using low shelving for manual access or floor-level storage, with accounting entirely reliant on manual operations by warehouse staff. This approach results in high floor space requirements, high labor intensity, and data management errors, making it particularly unfriendly to new employees. With the transformation and upgrading towards smart factories, the need to reduce floor space, improve material and product management, reduce labor intensity, and seamlessly integrate with front-end and back-end management systems is becoming increasingly urgent.

[0003] To address these needs, intelligent warehousing systems equipped with automated stacker cranes are widely used. Existing stacker cranes mainly fall into two control categories: servo control, which achieves accurate positioning but is expensive and lacks flexible control over the operational curve, resulting in poor versatility and difficulty in porting between different brands; and simple frequency converter control, which is inexpensive but suffers from inaccurate positioning, unstable operational curves, and is prone to errors. Furthermore, the location calibration of stacker cranes typically relies on manual visual inspection, column by column.

[0004] Existing stacker crane solutions have significant drawbacks. While servo control offers accurate positioning, it is inefficient and lacks flexible control over operating curves. Simple frequency converter control, though low-cost, suffers from insufficient accuracy, impacting warehouse reliability. Furthermore, manual calibration methods are inefficient, error-prone, and increase debugging workload. These issues hinder the performance improvement and widespread application of stacker cranes in smart warehousing. Summary of the Invention

[0005] This invention provides an intelligent warehousing system, a stacker crane, and a control method thereof, aiming to improve at least one of the above-mentioned technical problems.

[0006] In a first aspect, to solve the above-mentioned technical problems, the present invention provides a control method for a stacker crane, comprising: S1. Obtain the job instructions and parse the target location.

[0007] S2. Plan the motion path based on the current position of the stacker crane and the target position. The path planning is based on a three-axis coupled dynamics model, with the shortest total running time as the objective function. It comprehensively considers the differences in the dynamic parameters of each axis to optimize the speed curve coordination relationship of each axis.

[0008] S3. Generate a smooth velocity curve for each motion axis based on the motion path. The velocity curve adopts a piecewise continuous S-shaped acceleration / deceleration model. Wherein, if If the stroke is long, the standard seven-segment S-curve speed curve is used. Otherwise, it is a short stroke, and a simplified five-segment or three-segment S-curve is used. The total length of the motion path, For critical length, For maximum speed, For maximum acceleration, This is the maximum jerk.

[0009] S4. Move the stacker crane to the target position according to the motion path and speed curve.

[0010] In a second aspect, the present invention provides a stacker crane suitable for performing a stacker crane control method as described in the first aspect.

[0011] Thirdly, the present invention provides an intelligent warehousing system comprising multi-level, multi-column racks and a stacker crane as described in the second aspect. The stacker crane is adapted to move goods between the various compartments of the racks.

[0012] By adopting the above technical solution, the present invention can achieve the following technical effects: The stacker crane control method of the present invention decomposes the geometric path into a three-axis position-time sequence, and prioritizes the synchronous execution of the X-axis (horizontal) and Y-axis (vertical) movements, and then executes the Z-axis (fork) movement after they are in place, which effectively avoids the risk of interference during the movement of the forks.

[0013] The X and Y axes are jointly optimized based on a coupled dynamics model, significantly improving the synchronization and smoothness of their coordinated motion. The Z axis is planned independently to ensure precise and reliable telescopic movements. A piecewise continuous S-shaped acceleration and deceleration model is adopted, and the number of curve segments is adaptively selected according to the stroke length, effectively avoiding abrupt acceleration changes and velocity cutoff problems under short strokes. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of the specific embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0015] Figure 1 This is a structural schematic diagram of the stacker crane from a first-person perspective.

[0016] Figure 2 This is a structural schematic diagram of the stacker crane from a second-view perspective.

[0017] Figure 3 This is a structural diagram of the cargo platform assembly from a first-person perspective.

[0018] Figure 4 This is a structural schematic diagram of the cargo platform assembly from a second-view perspective.

[0019] Figure 5 This is a structural schematic diagram of the upper crossbeam assembly from a first-person perspective.

[0020] Figure 6 This is a structural schematic diagram of the upper crossbeam assembly from a second perspective.

[0021] The markings in the diagram are: 1-Ceiling rail assembly, 2-Upper crossbeam assembly, 3-Column assembly, 4-Fork assembly, 5-Lower crossbeam assembly, 6-Ground rail assembly, 7-Loading platform assembly, 8-Holding rail, 9-Guide wheel assembly, 10-Ultra-high grating, 11-Ultra-wide grating, 12-Vertical frame, 13-Horizontal frame, 14-Safety clamp, 15-Linkage mechanism, 16-Overload slack rope protection mechanism, 17-Elastic element, 18-Fixed seat, 19-Detection unit, 20-Sliding seat, 21-Overload sensor, 22-Slack rope sensor. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0023] like Figures 1 to 6 As shown, the present invention provides a stacker crane, which includes a ground rail assembly 6, a lower crossbeam assembly 5 slidably connected to the ground rail assembly 6, a top rail assembly 1, an upper crossbeam assembly 2 slidably connected to the top rail assembly 1, a column assembly 3 connected to the lower crossbeam assembly 5 and the upper crossbeam assembly 2, a loading platform assembly 7 slidably connected to the column assembly 3, a fork assembly 4 telescopically mounted on the loading platform assembly 7, and an image acquisition module and a distance sensor connected to the loading platform assembly 7.

[0024] The loading platform assembly 7 is raised and lowered via a steel cable traction lifting mechanism. The lifting mechanism drives the loading platform to rise and fall. A variable frequency speed control geared motor is used to achieve stepless speed regulation of the rising and falling speeds, reducing mechanical impact and facilitating accurate positioning of the loading platform. The steel cable traction lifting mechanism significantly reduces the power of the lifting motor and the size of the reducer, lowering operating costs.

[0025] The lower crossbeam assembly 5 is equipped with a running mechanism for horizontal movement along the running track of the ground rail assembly 6. For example... Figure 5 and Figure 6 As shown, the upper crossbeam mainly consists of upper crossbeam welded parts, wire rope wheel mechanism, guide wheel mechanism, and some fasteners.

[0026] like Figure 3 and Figure 4 As shown, the loading platform assembly 7 mainly consists of a guide wheel assembly 9, a vertical frame 12, and a horizontal frame 13. The loading platform assembly 7 is equipped with a telescopic fork assembly 4, a pulley system, an overspeed protection mechanism, a lifting guide wheel mechanism, a lifting positioning mechanism, a cargo position anomaly detection mechanism, and a dual entry detection mechanism. The overspeed protection device includes a brake clamp and / or a rope breakage protection mechanism.

[0027] A rectangular frame is formed by a vertical frame 12 and a horizontal frame 13. Multiple sets of gratings and / or multiple sets of through-beam photoelectric sensors are arranged around the perimeter of the rectangular frame. The multiple sets of gratings include at least an ultra-wide grating 11, an ultra-high grating 10, a collapsible grating, and a protruding grating. The protruding grating is used to detect whether the lowest point of the goods is higher than a preset height.

[0028] When cargo is inspected, if any set of photoelectric sensors is triggered, the stacker crane will alarm and display the corresponding alarm information. Overheight and overwidth checks are performed continuously. Collapse and overhang detections are conducted at the end of the fork retraction process. Collapse detection uses one set of grating and one set of photoelectric sensors on one side, with the grating and sensor illuminating diagonally at an X-shape.

[0029] like Figure 5 and Figure 6 As shown, the upper crossbeam assembly 2 is equipped with an overload slack rope protection mechanism 16. The overload slack rope protection mechanism 16 includes a fixed seat 18 engaged with the upper crossbeam assembly 2, a sliding seat 20 slidably engaged with the fixed seat 18, an elastic element 17 engaged between the fixed seat 18 and the sliding seat 20, a detection part 19 engaged with the movable seat, a slack rope sensor 22 engaged with the upper crossbeam assembly 2, and an overload sensor 21.

[0030] In an optional embodiment, when the loading platform is empty or loaded, the spring compresses by a certain amount, at which point both the slack rope and overload sensors 21 detect it. When the wire rope slacks or breaks, the spring is ejected, at which point the sensing element detaches from the slack rope sensor 22, and the sensor alarms. When the load is overloaded, the spring compresses and increases beyond the design range, at which point the sensing element detaches from the overload sensor 21, and the sensor alarms.

[0031] like Figure 4 As shown, the loading platform assembly 7 is equipped with a fall protection mechanism. This mechanism consists of a speed limiter (not shown), a linkage mechanism 15, a safety clamp 14, and a rail clamp 8. The safety clamp 14 is fixed to the loading platform, and the linkage mechanism 15 is connected to the speed limiter via a steel wire rope. When the loading platform's descent speed exceeds the speed limiter's speed, the speed limiter triggers to engage the steel wire rope, which in turn pulls the linkage mechanism 15. The linkage mechanism 15 then triggers the safety clamp 14 and the fall sensor. The safety clamp 14 engages the rail clamp 8 to prevent the loading platform from falling, and the fall sensor triggers an alarm, preventing further operation of the equipment.

[0032] The present invention also provides a rack for use with the aforementioned stacker crane. The rack and the stacker crane together form an intelligent warehousing system. The rack has multiple rows and columns of storage compartments. Specific markers for identifying coordinates are provided on the rack.

[0033] Preferably, the specific marking includes a QR code, a barcode, or a geometric pattern. The QR code adopts the QR code standard and stores the location code information of the storage compartment. The barcode adopts the Code128 format. The geometric pattern includes a circle, a rectangle, or a triangle, with a pattern size of 10 mm by 10 mm and a pattern contrast of not less than 70% to ensure reliable identification under complex lighting conditions.

[0034] This invention also provides a control method for a stacker crane, which can be executed by the stacker crane described above. Specifically, it is executed by one or more processors in the stacker crane to implement steps S1 to S4.

[0035] S1. Obtain the job instructions and parse the target location.

[0036] Specifically, the stacker crane control system receives operation instructions from the upper-level management system. These instructions include target storage compartment coordinate information, and the specific coordinates of the target location in three-dimensional space are determined based on a preset storage compartment coordinate mapping table.

[0037] Preferably, the parsing process of the work instruction includes: establishing a mapping relationship database between the coordinates of the cargo cell and the three-dimensional spatial position. The database stores the specific values ​​of the horizontal coordinate X, vertical coordinate Y, and depth coordinate Z corresponding to each cargo cell. When a work instruction is received, the target cargo cell number is converted into the corresponding three-dimensional coordinate value by querying the mapping relationship database, and the coordinate parsing accuracy reaches 0.1 mm.

[0038] S2. Plan the motion path based on the current position of the stacker crane and the target position. The path planning is based on a three-axis coupled dynamics model, with the shortest total running time as the objective function, comprehensively considering the differences in the dynamic parameters of each axis to optimize the speed curve coordination relationship of each axis. Preferably, step S2 includes steps S21 to S25.

[0039] S21. Based on the current position and target position of the stacker crane, and using a pre-built 3D grid map, a modified A* search algorithm is used to globally plan the motion path.

[0040] The cost function of the improved A* search algorithm Represented as: .

[0041] In the formula, From the starting point to the node The actual path cost (calculated in terms of Euclidean or Manhattan distance). For heuristic cost estimation (using diagonal distance or Euclidean distance). This is the dynamic weight adjustment coefficient (range 0.1~0.5). For nodes The Euclidean distance to the nearest obstacle. The preset safety distance threshold (usually set to 1.5 times the maximum outline radius of the stacker crane). It is a natural exponential function.

[0042] The steps for constructing a 3D raster map model of an automated warehouse are as follows: divide the warehouse space into discrete voxel units, and label each unit as free space, obstacle, or storage area.

[0043] The preferred, improved A* algorithm introduces a dynamic pruning strategy during the search process: when a node's... When the cost of the current known optimal path exceeds 1.2 times, the expansion of that branch is terminated early to improve search efficiency. Simultaneously, the algorithm supports parallel search for multiple target points, making it suitable for batch job scenarios.

[0044] S22. Smooth the discrete path point sequence output by the A* algorithm. Specifically, use cubic spline interpolation or B-spline curve fitting to generate a continuous and differentiable geometric path, ensuring continuous curvature and avoiding sharp turns.

[0045] Preferably, the path smoothing process uses a fifth-order B-spline curve, and the control points are generated by simplifying the path points output by A* using the Douglas-Peucker algorithm to ensure that the second derivative of the curve is continuous and the maximum curvature does not exceed the turning capability limit of the stacker crane.

[0046] S23. Parameterize the smoothed geometric path according to time, decomposing it into position-time series of the three motion axes (X, Y, Z) of the stacker crane. .

[0047] S24. Based on the pre-constructed three-axis coupled dynamic model, use sequential quadratic programming (SQP) or model predictive control (MPC) methods to solve for the optimal velocity curve parameters, so that the X and Y axes, under the premise of satisfying dynamic and synchronization constraints, achieve the optimal velocity curve parameters in the shortest time. Upon reaching the target position synchronously, the Z-axis then independently performs a telescopic motion.

[0048] The three-axis coupled dynamics model takes the shortest total running time as the objective function and comprehensively considers the differences in dynamic parameters of each axis (including motor inertia, transmission ratio, friction coefficient), path length, obstacle avoidance margin, and cooperative motion constraints (such as synchronous start and stop, maximum relative deviation limit) to optimize the speed curve coordination relationship of each axis.

[0049] In the three-axis coupled dynamics model, the X-axis (horizontal movement) and Y-axis (vertical lifting) move synchronously first. After the X and Y axes reach the target position, the Z-axis (fork extension and retraction) movement is started to avoid interference or collision of the forks during the movement.

[0050] The objective function and constraints of the triaxial coupled dynamics model are as follows: Objective function: .

[0051] In the formula, , , These represent the time required for each of the X, Y, and Z axes to complete its respective motion.

[0052] Dynamic constraints: .

[0053] In the formula, , and These are the equivalent moments of inertia for the X, Y, and Z axes, respectively. , and These are the coefficients of viscous friction for the X, Y, and Z axes, respectively. , and Nonlinear friction models (including Coulomb friction and static friction) are provided for the X, Y, and Z axes, respectively. , and These are the motor output torques for the X, Y, and Z axes, respectively. , and These represent the motor rotation angles along the X, Y, and Z axes, respectively.

[0054] Kinematic constraints: .

[0055] In the formula, , and These are the X, Y, and Z coordinates of the stacker crane, respectively. For time. , and These are the transmission ratio conversion coefficients for the X, Y, and Z axes, respectively. , and These represent the maximum movement speeds along the X, Y, and Z axes, respectively. , and These represent the maximum accelerations along the X, Y, and Z axes, respectively.

[0056] Cooperative motion constraints: .

[0057] In the formula, and These are the jointly optimized reference trajectories for the X and Y axes, respectively. The maximum allowable synchronization deviation (usually set to 1~2mm). This represents the maximum travel time in the X and Y axis directions. This represents the total time. , and These represent the travel time in the X, Y, and Z axes, respectively.

[0058] S3. Generate a smooth velocity curve for each motion axis based on the motion path. The velocity curve adopts a piecewise continuous S-shaped acceleration / deceleration model. Wherein, if If the stroke is long, the standard seven-segment S-curve speed curve is used. Otherwise, it is a short stroke, and a simplified five-segment or three-segment S-curve is used.

[0059] .

[0060] In the formula, The total length of the motion path, For critical length, For maximum speed, For maximum acceleration, This is the maximum jerk.

[0061] Specifically, a smooth velocity curve ensures continuous variation in acceleration and jerk, avoiding mechanical shock. Preferably, step 3 includes the following sub-steps: S31. Based on the total length and direction of the motion path and the dynamic parameters of each axis of the stacker crane, determine whether the motion type is long-stroke or short-stroke. The dynamic parameters include maximum speed, maximum acceleration, and maximum jerk.

[0062] Critical length Determined by the following formula: .

[0063] S32, if If the distance is within a certain range, it is considered a long distance. Otherwise, it is considered a short distance.

[0064] S33. For long strokes, a standard seven-segment S-shaped velocity curve is adopted. The standard seven-segment S-shaped velocity curve includes an acceleration rising segment, a constant acceleration segment, an acceleration falling segment, a constant velocity segment, a deceleration rising segment, a deceleration constant segment, and a deceleration falling segment.

[0065] Preferably, the S-shaped velocity curve is generated based on analytical expressions or lookup table interpolation to ensure rapid calculation within the control cycle. For the seven-segment S-shaped curve, the time for each stage is determined by solving the following constraints: Total displacement Equal to path length: .

[0066] speed Not exceeding the maximum speed threshold : .

[0067] acceleration Not exceeding the maximum acceleration threshold : .

[0068] accelerometer Not exceeding the maximum jerk threshold : .

[0069] In the formula, This represents the total time. For time. It is a differential.

[0070] velocity function The seven segments are defined as follows: when hour: .

[0071] when hour: .

[0072] when hour: .

[0073] when hour: .

[0074] when hour: .

[0075] when hour: .

[0076] when hour: .

[0077] In the formula, , , , , , and These represent the durations of the first to seventh segments of the S-shaped velocity curve. , , , , , and These represent the velocities of the first to seventh segments of the S-shaped velocity curve. This is the first substitution symbol, used to shorten the length of formulas. .

[0078] S34. For short-distance travel, if the path length is insufficient to complete the full seven-segment acceleration process, the system will automatically switch to a five-segment or three-segment S-shaped speed curve. The five-segment curve omits the constant speed segment and the constant acceleration segment, while the three-segment curve retains only the acceleration increase, constant speed (or peak speed maintenance), and acceleration decrease segments.

[0079] The five-segment S-shaped velocity curves are as follows: when hour: .

[0080] when hour: .

[0081] when hour: .

[0082] when hour: .

[0083] when hour: .

[0084] The three S-shaped velocity curves are as follows: .

[0085] In the formula, The maximum speed for the shortest distance is determined by the path length. The solution is obtained by combining the dynamic constraints and the solution, ensuring that the conditions are met. , , , , These represent the durations of the first to fifth segments of the S-shaped velocity curve. , , , , These represent the velocities of the first to fifth segments of the S-shaped velocity curve. This is the maximum jerk threshold. This is the maximum acceleration threshold.

[0086] Preferably, step 34 introduces a path length adaptive speed limiting mechanism: when the single-axis motion path length is detected... Less than the critical length (Depend on , and When jointly decided, the maximum operating speed of the axis is automatically reduced to ensure sufficient distance to complete the acceleration and deceleration process, and to avoid impact caused by the speed curve being cut off due to too short a stroke.

[0087] In the formula For speed coefficient. Preferably, .

[0088] S35. In multi-axis cooperative motion, the X-axis and Y-axis jointly optimize the velocity curve parameters to ensure they remain within a unified time interval. The motion is completed synchronously within the internal axis. The Z-axis motion begins after the X and Y axes have reached their positions, and its motion time interval is [duration missing]. In the formula, This represents the maximum travel time in the X and Y axis directions. The Z-axis travel time is used. The velocity curves of the X and Y axes are jointly optimized using a coupled dynamics model to ensure strict synchronous start and stop. The Z-axis generates its own velocity curve independently based on its path length and dynamic parameters, and is triggered to execute after the X and Y axes have reached their positions. At the same time, a coupled compensation algorithm corrects dynamic deviations caused by load changes and differences in inter-axis dynamics, ensuring smooth movement of the X and Y axes and precise movement of the Z-axis.

[0089] Preferably, in step 35, the coordinated motion of the X-axis and Y-axis is jointly optimized based on a coupled dynamics model. The goal is to minimize the total running time of both axes, dynamically adjusting their respective speed curve shapes and time allocation to ensure strictly synchronized start and stop. The Z-axis motion is planned independently and starts after the X and Y axes are in position, avoiding vibration and positioning errors caused by the coupling of fork movement and overall machine movement.

[0090] S4. Move the stacker crane to the target position according to the motion path and speed curve.

[0091] Specifically, the calculated motion path and speed curve are sent to the stacker crane's variable frequency drive. Through real-time position feedback and speed closed-loop control, the stacker crane's various motion axes are driven to run along the predetermined trajectory and speed curve until they accurately reach the target position.

[0092] Preferably, in step 4, the position control employs a piecewise PID control strategy, using different PID parameters at different motion stages to control the position error. With control output The relationship is represented as: .

[0093] In the formula This is the first adaptive gain function. This is the second adaptive gain function. This is the third adaptive gain function. is the time constant. It is a differential. For time.

[0094] The stacker crane control method of the present invention decomposes the geometric path into a three-axis position-time sequence, and prioritizes the synchronous execution of the X-axis (horizontal) and Y-axis (vertical) movements, and then executes the Z-axis (fork) movement after they are in place, which effectively avoids the risk of interference during the movement of the forks.

[0095] The X and Y axes are jointly optimized based on a coupled dynamics model, significantly improving the synchronization and smoothness of their coordinated motion. The Z axis is planned independently to ensure precise and reliable telescopic movements. A piecewise continuous S-shaped acceleration and deceleration model is adopted, and the number of curve segments is adaptively selected according to the stroke length, effectively avoiding abrupt acceleration changes and velocity cutoff problems under short strokes.

[0096] Preferably, the stacker crane adopts a variable frequency control system, including a main controller, a variable frequency drive, an encoder feedback system, and a communication module. The main controller is responsible for executing all the calculation tasks of the control method, the variable frequency drive drives the motor to run according to the speed curve command, the encoder feedback system provides real-time position information, and the communication module realizes data exchange with the upper-level system.

[0097] Under the variable frequency drive architecture, positioning performance close to that of a servo system is achieved (measured accuracy better than 1 mm), while the cost is reduced by about 50% compared to a full servo solution.

[0098] Based on the above embodiments, in an optional embodiment of the present invention, a stacker crane control method further includes an automatic grid calibration function. This function uses a laser rangefinder sensor installed on the stacker crane to measure the distance to the rack reference surface in real time, and combines this with position information fed back from the encoder to automatically correct the grid coordinate mapping relationship, achieving a calibration accuracy of 0.5 mm. This online adaptive calibration function reduces reliance on manual adjustments and improves the long-term accuracy, stability, and robustness of the system.

[0099] The automatic compartment calibration function includes the following specific steps A1 to A8.

[0100] A1. The stacker crane moves to the origin of the track along the longitudinal and transverse directions of the rack, respectively.

[0101] A2. The stacker crane moves sequentially according to the coordinates of the storage compartments stored in the system. The image acquisition module acquires images of each storage compartment, while the distance sensor collects rack feature data. The storage compartment images include specific markings set on the rack.

[0102] A3. Fit the shelf plane and calculate the flatness based on the shelf feature data, and determine whether the shelf plane exceeds the tilt threshold and whether the flatness exceeds the flatness threshold. Issue an alarm when either the tilt threshold or the flatness threshold is exceeded.

[0103] Specifically, the flatness monitoring submodule monitors the flatness information of the shelf surface in real time, using distance sensors equipped on the stacker crane. After acquiring the flatness information, the rule acquisition submodule retrieves the corresponding calibration rules based on the current shelf type and environmental conditions. The calibration rules include key parameters such as the maximum allowable tilt angle of the shelf surface and the flatness error range. The flatness judgment submodule compares the monitoring data with the calibration rules. If the rules are met, subsequent steps are continued; otherwise, an alarm is issued.

[0104] The distance sensor includes a laser rangefinder and / or an ultrasonic sensor. The laser rangefinder has a measurement accuracy of ±0.1 mm. The ultrasonic sensor has a measurement accuracy of ±1 mm.

[0105] The shelving feature data includes distance data from at least 100 sampling points on the shelving surface, and the plane equation of the shelving surface is fitted using the least squares method. The fitting formula for the plane equation is as follows: ,in Indicates the surface height of the shelf. and Represents the horizontal and vertical coordinates. For the horizontal fitting coefficient, These are the longitudinal fitting coefficients. The initial value is used. The flatness is set to the standard deviation of the fitted residuals.

[0106] The flatness threshold is adaptively adjusted based on ambient temperature. The adjustment formula is: .

[0107] In the formula This indicates the adjusted error threshold. This represents the basic error threshold. This is the temperature coefficient, with a value of 0.01 per degree Celsius. It indicates the amount of change in ambient temperature.

[0108] Preferably, the flatness threshold does not exceed 0.5 mm. The tilt threshold is dynamically adjusted according to the type of shelving. For heavy-duty shelving, the maximum allowable tilt angle is set to 0.5 degrees, and for light-duty shelving, the maximum allowable tilt angle is set to 1 degree.

[0109] A4. Based on image segmentation algorithms, each shelf cell image is processed to extract feature point information for each cell. Each shelf cell image contains multiple cell labeling regions, each composed of specific markings on the shelf.

[0110] The image segmentation algorithm employs edge detection technology combined with morphological processing to accurately segment the warehouse marker area. The edge detection technology uses the Canny edge detection algorithm, and the morphological processing includes erosion and dilation operations. Feature point information includes the boundary points, geometric center points, and corner key data of the warehouse marker area.

[0111] The edge detection technique of the image segmentation algorithm adopts the Canny edge detection algorithm, with a Gaussian filter kernel size of 5 x 5, a high threshold to low threshold ratio of 2:1, and a 3 x 3 rectangular structuring element for morphological processing erosion operation and the same structuring element for dilation operation. The number of iterations is 2 to effectively remove noise and enhance the clarity of feature region contours.

[0112] A5. Calculate the actual position coordinates of each compartment marking area based on the feature point information, and identify the geometric center point of the compartment marking area.

[0113] The formula for calculating the geometric center point is: .

[0114] .

[0115] In the formula and These represent the x and y coordinates of the boundary points of the cargo compartment identification area, respectively. Indicates the number of boundary points.

[0116] A6. The actual coordinates of the center point are calculated by combining the relative distance between the geometric center point and the origin with the kinematic model formula of the stacker crane.

[0117] The kinematic model formula is: .

[0118] .

[0119] In the formula This represents the actual x-axis. This represents the actual vertical coordinate. The x-coordinate of the origin is given. The ordinate is the y-coordinate of the origin. This indicates the relative distance between the geometric center point and the reference point. This indicates the angle between the stacker crane's direction of movement and the reference axis. This is the error compensation term.

[0120] Error compensation item Dynamic estimation is performed using the Kalman filter algorithm.

[0121] The state equation and observation equation for the Kalman filter are as follows: .

[0122] .

[0123] In the formula This represents the system state vector, including position and velocity. Indicates control input, and This represents process noise and observation noise. , and For the system matrix, the error compensation term is... The estimated value is obtained through the posterior estimation of the state vector, with an estimation accuracy of ±0.05 mm.

[0124] A7. Compare the actual location coordinates with the grid coordinates to generate a grid deviation matrix. The deviation matrix is ​​used to identify systematic deviation patterns.

[0125] The formula for calculating the price deviation matrix is: .

[0126] In the formula Indicates the first Line number Deviation values ​​of the listed goods. The actual x-axis, The actual vertical axis, The x-axis of the grid is , y is the vertical axis of the grid.

[0127] The generation of the compartment deviation matrix includes statistical analysis of the deviation values ​​and calculation of the mean of all compartment deviation values. and standard deviation .

[0128] The formula for calculating the mean is: .

[0129] The formula for calculating standard deviation is: .

[0130] In the formula, This represents the number of rows on the shelf. This refers to the number of columns on the shelf.

[0131] A8. If the maximum deviation value in the cargo grid deviation matrix exceeds the preset threshold, the cargo grid coordinates in the system database will be dynamically adjusted according to the cargo grid deviation matrix.

[0132] The method for dynamically adjusting the coordinates of the storage compartments is as follows: .

[0133] .

[0134] In the formula The adjusted x-axis, The adjusted ordinate. This is for adjusting the coefficient. The value range is between 0 and 1.

[0135] Adjustment coefficient The value is determined adaptively based on the magnitude of the deviation.

[0136] The adaptive formula is: .

[0137] In the formula This is the sensitivity coefficient, with a value of 10. This represents the deviation threshold; when the deviation value... When approaching the threshold, The values ​​are smoothly transitioned to avoid oscillations during the adjustment process. It is the natural base.

[0138] Based on the above embodiments, in an optional embodiment of the present invention, the automatic calibration function of the cargo compartment further includes step A8.

[0139] A8. After the dynamic adjustment is completed, the system verifies the updated grid coordinates. The verification method is to re-collect the actual position coordinates of some grids, calculate the verification deviation value, and the verification deviation value must be less than 50% of the preset threshold. Otherwise, the calibration process is re-executed.

[0140] The formula for calculating the verification deviation value is: .

[0141] In the formula To verify the deviation value. To verify the actual x-coordinates collected. To verify the actual ordinate collected. This is the updated x-coordinate. This is the updated ordinate.

[0142] The automatic grid calibration function of this invention achieves high-precision automatic calibration of grid positions by integrating a multi-level technical solution of flatness monitoring, image processing, coordinate calculation, deviation analysis and dynamic adjustment. It effectively overcomes the problems of insufficient accuracy and poor adaptability caused by the reliance on manual visual inspection and fixed parameters in traditional methods, and improves the positioning stability and operating efficiency of stacker cranes in complex warehousing environments.

[0143] Preferably, the above control method is implemented on the Siemens TIA Portal platform. The control logic is written in a structured text language, and the motion control library functions provided by the platform are used to perform complex mathematical operations and real-time control tasks, ensuring that the control cycle is less than 10 milliseconds.

[0144] Obviously, the above detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to describe preferred embodiments, not all embodiments, and is not intended to limit the invention. Various modifications and variations can be made to the invention by those skilled in the art. Based on the embodiments of the invention, any modifications, equivalent substitutions, improvements, etc., made by those skilled in the art to all other embodiments obtained without inventive effort are within the scope of protection of the invention.

Claims

1. A control method for a stacker crane, characterized in that, Stacker cranes are suitable for moving goods between multiple storage compartments; The stacker crane uses frequency conversion control; the control method includes: S1. Obtain the work instruction and parse the target location; S2. Plan the motion path based on the current position of the stacker crane and the target position; wherein, the path planning is based on a three-axis coupled dynamics model, with the shortest total running time as the objective function, and comprehensively considers the differences in the dynamic parameters of each axis to optimize the speed curve coordination relationship of each axis; S3. Generate a smooth velocity curve for each motion axis according to the motion path. The velocity curve adopts a piecewise continuous S-shaped acceleration / deceleration model; wherein, if If the stroke is long, the standard seven-segment S-curve speed curve is used; otherwise, the stroke is short, and a simplified five-segment or three-segment S-curve is used. ; The total length of the motion path, For critical length, For maximum speed, For maximum acceleration, This is the maximum jerk. S4. Move the stacker crane to the target position according to the motion path and speed curve.

2. The control method for a stacker crane according to claim 1, characterized in that, S1 specifically includes: Based on the current and target positions of the stacker crane, and using a pre-built 3D grid map, an improved A* search algorithm is employed to globally plan the motion path; the cost function of the improved A* search algorithm is... Represented as: ； In the formula, From the starting point to the node The actual path cost; For heuristic cost estimation; This is the dynamic weight adjustment coefficient; For nodes Euclidean distance to the nearest obstacle; The preset safe distance threshold; It is a natural exponential function; Smooth the discrete path point sequence output by the A* algorithm; The smoothed geometric path is parameterized by time and decomposed into the position-time series of the three motion axes of the stacker crane. Based on a pre-constructed triaxial coupled dynamics model, the optimal velocity curve parameters are solved using sequential quadratic programming (SQP) or model predictive control (MPC) methods, so that the X and Y axes achieve the desired velocity curve in the shortest possible time while satisfying dynamic and synchronization constraints. Upon reaching the target position synchronously, the Z-axis then independently performs a telescopic motion.

3. The control method for a stacker crane according to claim 1, characterized in that, The three-axis coupled dynamics model takes the shortest total running time as the objective function, and comprehensively considers the differences in dynamic parameters of each axis, path length, obstacle avoidance margin, and cooperative motion constraints. The objective function and constraints of the triaxial coupled dynamics model are as follows: Objective function: ； In the formula, , , These represent the time required for each axis (X, Y, and Z) to complete its respective motion. Dynamic constraints: ； In the formula, , and These are the equivalent moments of inertia for the X, Y, and Z axes, respectively. , and These are the coefficients of viscous friction for the X, Y, and Z axes, respectively. , and Nonlinear friction models for the X, Y, and Z axes, respectively; , and The motor output torques for the X, Y, and Z axes are respectively; , and These represent the motor rotation angles along the X, Y, and Z axes, respectively. Kinematic constraints: ； In the formula, , and These are the X, Y, and Z axis coordinates of the stacker crane, respectively. For time; , and These are the transmission ratio conversion factors for the X, Y, and Z axes, respectively; , and These represent the maximum movement speeds along the X, Y, and Z axes, respectively. , and These are the maximum accelerations along the X, Y, and Z axes, respectively. Cooperative motion constraints: ； In the formula, and These are the jointly optimized reference trajectories for the X and Y axes, respectively. The maximum allowable synchronization deviation; The maximum travel time in the X and Y axis directions; Total time; , and These represent the travel time in the X, Y, and Z axes, respectively.

4. The control method for a stacker crane according to claim 1, characterized in that, S3 specifically includes: Based on the total length and direction of the motion path and the dynamic parameters of each axis of the stacker crane, determine whether the motion type is long stroke or short stroke; like If the distance is long, it is considered a long journey; otherwise, it is considered a short journey. For long strokes, a standard seven-segment S-shaped speed curve is used; For short distances, if the path length is insufficient to complete the full seven-segment acceleration process, it will automatically switch to a five-segment or three-segment S-shaped velocity curve; the five-segment curve omits the constant speed segment and the constant acceleration segment, while the three-segment curve only retains the acceleration increase, constant speed, and acceleration decrease segments; The five-segment S-shaped velocity curves are as follows: when hour: ; when hour: ; when hour: ; when hour: ； when hour: ； The three S-shaped velocity curves are as follows: ； In the formula, This is the maximum speed for the shortest stroke. , , , , These represent the durations of the first to fifth segments of the S-shaped velocity curve; , , , , These represent the velocities of the first to fifth segments of the S-shaped velocity curve; The maximum jerk threshold; The maximum acceleration threshold; In multi-axis coordinated motion, the velocity curve parameters of the X-axis and Y-axis are jointly optimized to ensure they remain within a unified time interval. The motion is completed synchronously within the internal axis; the Z-axis motion starts after the X and Y axes have reached their positions, and its motion time interval is... In the formula, The maximum travel time in the X and Y axis directions; The Z-axis travel time is used; the velocity curves of the X and Y axes are jointly optimized using a coupled dynamics model; the Z-axis generates its own velocity curve independently based on its path length and dynamic parameters, and is triggered to execute after the X and Y axes are in position; at the same time, a coupled compensation algorithm is used to correct dynamic deviations caused by load changes and differences in dynamics between axes.

5. The control method for a stacker crane according to claim 1, characterized in that, S4 specifically involves sending the calculated motion path and speed curve to the stacker crane's frequency converter driver. Through real-time position feedback and speed closed-loop control, the stacker crane's motion axes are driven to run along the predetermined trajectory and speed curve until they accurately reach the target position. The drive employs a PID control strategy; Position error With control output The relationship is represented as: ； In the formula, This is the first adaptive gain function; This is the second adaptive gain function; This is the third adaptive gain function; It is a time constant; It is a differential; For time.

6. A control method for a stacker crane according to any one of claims 1 to 5, characterized in that, It also includes an automatic storage compartment calibration function; the automatic storage compartment calibration function includes the following specific steps: The stacker crane moves to the track origin along the longitudinal and transverse directions of the rack, respectively, and the image acquisition module acquires the origin image containing the rack origin; The stacker crane moves sequentially according to the coordinates of the storage compartments stored in the system. The image acquisition module acquires images of each storage compartment, while the distance sensor acquires rack feature data. The storage compartment images include specific markings set on the rack. The shelf plane is fitted and the flatness is calculated based on the shelf feature data, and it is determined whether the shelf plane exceeds the tilt threshold and whether the flatness exceeds the flatness threshold. An alarm is triggered when the tilt or flatness threshold is exceeded. The images of each cargo compartment are processed separately based on the image segmentation algorithm to extract the feature point information of each cargo compartment; Calculate the actual position coordinates of each compartment marking area based on the feature point information, and identify the geometric center point of the compartment marking area; The actual coordinates of the center point are calculated by combining the relative distance between the geometric center point and the origin with the kinematic model formula of the stacker crane. The actual location coordinates are compared with the coordinates of the cargo cells to generate a cargo cell deviation matrix; If the maximum deviation value in the cargo grid deviation matrix exceeds the preset threshold, the cargo grid coordinates in the system database will be dynamically adjusted based on the cargo grid deviation matrix.

7. The control method for a stacker crane according to claim 6, characterized in that, The shelving feature data includes distance data from at least 100 sampling points on the shelving surface, and the plane equation of the shelving surface is fitted using the least squares method. The fitting formula for the plane equation is as follows: ,in Indicates the surface height of the shelf. and Represents the horizontal and vertical coordinates. For the horizontal fitting coefficient, These are the longitudinal fitting coefficients. The initial value is used; the flatness is set to the standard deviation of the fitted residuals. The flatness threshold is adaptively adjusted based on ambient temperature; the adjustment formula is: ； In the formula This indicates the adjusted error threshold; Indicates the basic error threshold; This is the temperature coefficient, with a value of 0.01 per degree Celsius. It indicates the amount of change in ambient temperature.

8. The control method for a stacker crane according to claim 6, characterized in that, The image segmentation algorithm uses edge detection technology combined with morphological processing to accurately segment the warehouse marker area; the edge detection technology uses the Canny edge detection algorithm, and the morphological processing includes erosion and dilation operations; the feature point information includes the boundary points, geometric center points, and corner key data of the warehouse marker area; The formula for calculating the geometric center point is: ； ； In the formula and These represent the x and y coordinates of the boundary points of the cargo compartment identification area, respectively. Indicates the number of boundary points; The kinematic model formula is: ； ； In the formula The actual x-axis; The actual vertical coordinate; The x-coordinate of the origin; The ordinate of the origin; Indicates the relative distance between the geometric center point and the reference point; Indicates the angle between the stacker crane's direction of movement and the reference axis; This is the error compensation term; The formula for calculating the price deviation matrix is: ； In the formula Indicates the first Line number Deviation values ​​of the listed goods. The actual x-axis, The actual vertical axis, The x-axis of the grid is , The vertical axis represents the grid cell; The method for dynamically adjusting the coordinates of the storage compartments is as follows: ； ； In the formula The adjusted x-axis, The adjusted ordinate. This is for adjusting the coefficient.

9. A stacker crane, characterized in that, Suitable for performing a stacker crane control method as described in any one of claims 1 to 8.

10. An intelligent warehousing system, characterized in that, The system includes a multi-layer, multi-column rack and a stacker crane as described in claim 9; the stacker crane is adapted to move goods between the various compartments of the rack.

Images