Pressure map based mobile rack adaptive locking feedback control system and method

Abstract

The application provides a mobile shelf adaptive locking feedback control system and method based on a pressure map, and relates to the technical field of intelligent storage and transportation. The system comprises a mobile shelf, a carrying robot and a server. The intelligent tray of the mobile shelf can collect pressure data of placed goods through a pressure sensing array. A shelf controller generates a two-dimensional pressure distribution map according to the pressure data, and controls a limiting device to perform a locking action or an unlocking action according to the two-dimensional pressure distribution map, so as to realize adaptive shape position limitation of the goods at multiple locking points. The server can extract the state characteristics of the goods according to the two-dimensional pressure distribution map, and generate a carrying task containing motion constraint parameters, so as to control the carrying robot to complete the carrying of the goods through the carrying task. The system can maintain the safety constraint of the goods during the carrying and picking process, reduce the risk of sliding and dumping, can realize speed control differentiation according to the risk, and improve the carrying efficiency of the goods.

Description

Technical Field

[0001] This application relates to the field of intelligent storage and transportation technology, and in particular to an adaptive locking feedback control system and method for mobile shelves based on pressure maps. Background Technology

[0002] Robotic Mobile Fulfillment Systems (RMFS) enable robots to transport loaded shelves to the picking area for picking, improving picking efficiency. An RMFS may include a master decision-making unit, transport robots, mobile shelves, and picking equipment. During goods handling, the master decision-making unit issues a transport task to the transport robot. The robot analyzes the task, moves to the shelf to lift it, and then moves to the picking equipment to transport the lifted shelf to the picking area.

[0003] The mobile shelving unit can have multiple storage compartments for holding various goods. These compartments can be created by adding dividers and guardrails to the shelving. During goods handling, a handling robot moves beneath the shelving, lifts and moves the goods from multiple compartments together to the picking area. Because the movement of goods can affect their stability, safety measures are necessary. Safety measures at the shelving level can rely on passive structural constraints on fixed structures, such as partitions, boxes, and baffles.

[0004] However, due to the structural limitations of mobile shelving, the robot's movement system can only be conservatively speed-limited overall, meaning a uniform speed limit, a uniform upper limit for acceleration, and a reduction in speed when turning are applied. Furthermore, because mobile shelving uses simple external protective devices, the position of the goods loaded on the shelving cannot be effectively fixed. Therefore, during goods handling, goods may shift on the shelving, causing a change in the overall center of gravity. If the handling robot then moves according to the initially set conservative speed-limiting strategy, it is prone to causing imbalances in the goods, leading to handling malfunctions and reduced efficiency. Summary of the Invention

[0005] In view of this, embodiments of this application provide an adaptive locking feedback control system and method for mobile shelves based on pressure maps to solve the problem of low cargo handling efficiency.

[0006] According to a first aspect of this application, a pressure map-based adaptive locking feedback control system for a moving shelf is provided, the system comprising:

[0007] A mobile shelving unit includes a shelving body, multiple load-bearing surfaces, and a shelving controller. At least one shelf compartment is provided on each load-bearing surface. Each shelf compartment has a smart pallet and a limiting device at its bottom. The smart pallet includes a pressure sensor array and a data acquisition circuit connected to the pressure sensor array. The data acquisition circuit is configured to receive pressure data collected by the pressure sensor array. The limiting device includes multiple self-locking posts. The multiple self-locking posts form multiple locking points in the shelf compartment. The shelving controller is connected to the data acquisition circuit and the multiple self-locking posts. The shelving controller is configured to generate a two-dimensional pressure distribution map based on the pressure data, and to control the self-locking posts to perform locking or unlocking actions based on the two-dimensional pressure distribution map.

[0008] A transport robot includes a moving mechanism, a lifting mechanism, and a robot controller, wherein the robot controller is connected to the moving mechanism and the lifting mechanism; the robot controller is configured to drive the moving mechanism and the lifting mechanism according to a transport task.

[0009] A server is connected to the shelf controller and the robot controller; the server is configured to:

[0010] Receive a two-dimensional pressure distribution map;

[0011] Cargo status features are extracted from the two-dimensional pressure distribution map. The cargo status features include at least one of cargo weight, center of gravity offset, and shape risk indicators.

[0012] Motion constraint parameters are generated based on the cargo state characteristics. The motion constraint parameters include at least one of the following: maximum speed coefficient, maximum acceleration coefficient, turning coefficient, and emergency stop threshold.

[0013] The system generates a transport task based on the motion constraint parameters and sends the transport task to the robot controller.

[0014] In some embodiments, the shelf controller is further configured to:

[0015] Acquire the pressure data;

[0016] A two-dimensional pressure distribution map is generated based on the pressure data;

[0017] The cargo status parameters are calculated based on the two-dimensional pressure distribution map. The cargo status parameters include the cargo existence status, cargo occupied area, cargo weight, cargo center of gravity position, degree of center of gravity offset, and shape risk index.

[0018] The target locking points are determined based on the cargo status quantity. The target locking points are multiple locking points with the outer contour of the cargo-occupied area as the boundary.

[0019] Generate a locking control signal according to the target locking point, and send the locking control signal to the self-locking column at the target locking point.

[0020] In some embodiments, the self-locking column includes an actuating unit, a pushing elastic element, a reset driving element, and an electromagnetic release mechanism; the pushing elastic element is connected to the actuating unit and is used to push the actuating unit to a locked position; the reset driving element is connected to the actuating unit and is used to reset the actuating unit from the locked position to an initial position; the electromagnetic release mechanism is disposed at the locked position and the initial position and is used to lock or unlock the actuating unit according to a pulse current signal; the shelf controller is further configured to:

[0021] Obtain the current action state of the self-locking column, the current action state including the current action position of the action unit;

[0022] Generate a current set of lock points based on the current action state, and generate a set of target lock points based on the target lock points;

[0023] An action lock point set is determined by comparing the current lock point set with the target lock point set. The action lock point set includes lock points that need to be reclaimed and lock points that need to be locked.

[0024] The pulse current signal is generated based on the set of action lock points, and the pulse current signal is sent to the electromagnetic release mechanism.

[0025] In some embodiments, the self-locking column further includes a positioning detection device; the positioning detection device is disposed at the locked position and the initial position, and is used to detect the positioning state of the action unit; the limiting device further includes a current sensor, the current sensor being connected to the electromagnetic release mechanism, and is used to detect current data; the positioning detection device and the current sensor are connected to the shelf controller; the shelf controller is further configured to:

[0026] Acquire the current data and multiple frames of the pressure data;

[0027] Extract abnormal current features from the current data;

[0028] The cargo sliding distance is obtained by comparing the pressure data from multiple frames.

[0029] Calculate verification information, which includes at least one of the following: the positioning status, the abnormal current characteristics, and the cargo sliding distance;

[0030] Verification output information is generated based on the verification information. The verification output information includes the verification result and the failure reason code when the verification result is a lock failure.

[0031] In some embodiments, the shelf controller is further configured to perform calculations of cargo status quantities based on the two-dimensional pressure distribution map:

[0032] The data points of the two-dimensional pressure distribution map are traversed, and the data points include weight information collected by multiple pressure sensors in the pressure sensing array;

[0033] Extract target data points from the data points, where the target data points are the data points corresponding to the pressure sensors whose weight information is greater than or equal to a weight threshold.

[0034] The presence status and weight of the goods are obtained by accumulating the weight information of the target data points;

[0035] Based on the weights of the goods and the positions of the shelves on the mobile shelf, calculate the position of the center of gravity of the goods and the degree of center of gravity offset.

[0036] The cargo-occupied area is defined based on the target data points, and the occupancy size of the cargo-occupied area is traversed. The occupancy size includes a first size and a second size. The first size is the distance between edge target data points in a first direction; the second size is the distance between edge target data points in a second direction; the first direction is the direction of maximum distance between edge target data points; the second direction is perpendicular to the first direction.

[0037] The cargo shape index is calculated based on the occupied dimensions, and the cargo shape index includes the ratio of the first dimension to the second dimension;

[0038] The shape risk index is determined based on the shape range to which the cargo shape index belongs.

[0039] In some embodiments, the shelf controller is further configured to:

[0040] The total weight of the goods is calculated by summing the weights of the goods in multiple of the aforementioned compartments;

[0041] Obtain enhanced evaluation parameters, wherein the enhanced evaluation parameters are at least one of the total weight of the cargo, the degree of center of gravity offset, and the cargo shape index;

[0042] When the enhancement evaluation parameters meet the preset enhancement conditions, the support line is determined based on the enhancement evaluation parameters;

[0043] An enhanced locking signal is generated based on the support line, and the enhanced locking signal is sent to the self-locking post corresponding to the support line.

[0044] In some embodiments, the shelf controller is further configured to:

[0045] Acquire resampled data, which is load data obtained by performing weight collection on the remaining goods after any item is picked in the mobile shelf;

[0046] The locked points are calculated and updated based on the resampled data;

[0047] A differential update signal is generated based on the updated lock point and the lock control signal;

[0048] The differential update signal is sent to the self-locking column corresponding to the update lock point.

[0049] In some embodiments, the server is further configured to generate motion constraint parameters based on the cargo state characteristics:

[0050] Call the mapping database, which includes multiple state feature intervals and motion constraint parameters associated with the state feature intervals;

[0051] Obtain the target state feature interval to which the cargo state features belong;

[0052] The motion constraint parameters are determined based on the target state feature interval.

[0053] In some embodiments, the bottom of the shelf body is provided with a lifting and positioning assembly, the lifting and positioning assembly including a positioning groove and an anti-slip stop; the anti-slip stop is disposed on the inner wall of the positioning groove; the top of the lifting mechanism is provided with a positioning protrusion; the robot controller is further configured to:

[0054] Obtain location information, which includes the current location of the handling robot, the shelf location of the mobile shelf, and the picking location;

[0055] The movement path is planned based on the location information, and the movement path includes a pickup path and a handling path; the pickup path is planned based on the current location and the shelf location; the handling path is planned based on the shelf location and the picking location.

[0056] The motion constraint parameters are parsed from the transport task;

[0057] Based on the movement path and the motion constraint parameters, a transport control command is generated, and the transport control command is sent to the moving mechanism and the lifting mechanism.

[0058] According to a second aspect of this application, a moving shelf adaptive locking feedback control method based on a pressure map is provided, applied to the system described in the first aspect; the method includes:

[0059] Obtain stress data;

[0060] A two-dimensional pressure distribution map is generated based on the pressure data, and the self-locking column is controlled to perform locking or unlocking actions based on the two-dimensional pressure distribution map.

[0061] Cargo status features are extracted from the two-dimensional pressure distribution map. The cargo status features include at least one of cargo weight, center of gravity offset, and shape risk indicators.

[0062] Motion constraint parameters are generated based on the cargo state characteristics. The motion constraint parameters include at least one of the following: maximum speed coefficient, maximum acceleration coefficient, turning coefficient, and emergency stop threshold.

[0063] Based on the motion constraint parameters, a transport task is generated, and the transport task is sent to the robot controller so that the robot controller drives the moving mechanism and the lifting mechanism according to the transport task.

[0064] According to a third aspect of this application, a computer device is provided, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor, when executing the program, implements the above-described adaptive locking feedback control method for moving shelves based on pressure maps.

[0065] According to a fourth aspect of this application, a storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the above-described adaptive locking feedback control method for moving shelves based on pressure maps.

[0066] Based on the above technical solution, this application provides a pressure map-based adaptive locking feedback control system and method for mobile shelves. The system includes a mobile shelf, a handling robot, and a server. The intelligent pallets of the mobile shelf can collect pressure data of the placed goods through a pressure sensor array. The shelf controller generates a two-dimensional pressure distribution map based on the pressure data and controls the limit device to perform locking or unlocking actions according to the two-dimensional pressure distribution map, achieving adaptive shape-based positional limitation of the goods at multiple locking points. The server can extract the state characteristics of the goods based on the two-dimensional pressure distribution map and generate handling tasks containing motion constraint parameters, thereby controlling the handling robot to complete the goods handling through the handling tasks. The system can maintain safe constraints on the goods during handling and picking, reduce the risk of slippage and tipping, and achieve risk-differentiated speed control, improving the efficiency of goods handling.

[0067] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0068] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0069] Figure 1 A schematic diagram of the adaptive locking feedback control system for a moving shelf based on a pressure map provided in this application embodiment;

[0070] Figure 2 A schematic diagram of the cargo compartment structure provided in the embodiments of this application;

[0071] Figure 3 This is a schematic diagram of the initial position state of the self-locking pin provided in an embodiment of this application;

[0072] Figure 4 This is a schematic diagram of the self-locking pin's locked position state provided in an embodiment of this application;

[0073] Figure 5 A schematic flowchart of an adaptive locking feedback control method for a moving shelf based on a pressure map, provided in an embodiment of this application.

[0074] Figure 6 A schematic flowchart of the control method corresponding to the handling or lifting action provided in the embodiments of this application;

[0075] Figure 7 A schematic flowchart of a pressure map-based handling control method provided in an embodiment of this application;

[0076] Figure 8 This is a schematic flowchart of the locking feedback control method provided in the embodiments of this application;

[0077] Illustration:

[0078] Among them, 1-mobile shelf; 11-intelligent pallet; 12-limiting device; 13-pressure sensor array; 14-action unit; 15-pushing elastic element; 16-reset drive element; 17-electromagnetic release mechanism; 171-electromagnetic adsorption element; 172-locking element; 173-reset spring; 2-handling robot; 3-server. Detailed Implementation

[0079] The present application will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present application can be combined with each other.

[0080] To address the problem of low cargo handling efficiency, some embodiments of this application provide an adaptive locking feedback control system for a mobile shelving unit based on a pressure map. For example... Figure 1As shown, the adaptive locking feedback control system for the mobile shelf 1 based on the pressure map includes: the mobile shelf 1, the handling robot 2, and the server 3.

[0081] The mobile shelving unit 1 includes a main shelving body, multiple load-bearing surfaces, and a shelving controller. The main shelving body is a multi-layered frame structure assembled from multiple uprights and beams. Each layer of the multi-layered frame structure can form a load-bearing surface, and each load-bearing surface has at least one storage compartment for placing goods.

[0082] For example, the main body of the shelf is a metal frame structure assembled from 4 uprights and 20 beams. Every 4 uprights are measured as a group and fixed at the same height as a rectangular frame. Hollow metal plates are set in the rectangular frame to form a load-bearing surface. That is, 4 uprights and 20 beams can form 5 load-bearing surfaces, so 5 storage compartments can be formed on the mobile shelf 1.

[0083] like Figure 2 As shown, each compartment is equipped with a smart pallet 11 and a limiting device 12 at its bottom. The smart pallet 11 is used to collect pressure data of the placed goods, therefore the smart pallet 11 includes a pressure sensor array 13 and a data acquisition circuit connected to the pressure sensor array 13. The pressure sensor array 13 consists of multiple pressure sensors arranged according to an array rule. The pressure sensors can be one of the following: resistive pressure sensors, piezoelectric pressure sensors, capacitive pressure sensors, and optical pressure sensors.

[0084] For example, each shelf compartment can be equipped with a smart pallet 11 (tile). By arranging 35 pressure sensors in a 7×5 array, a rectangular pressure sensor array 13 can be formed. This array is compatible with partition structures and supports the control target of relocking individual shelves. Each smart pallet 11 (tile) uses a thin-film pressure sensor to collect pressure signals based on the weight and position of the goods placed on the shelf. The pressure signals can be used to generate a two-dimensional pressure or load sensor array (Pressure Grid) to output a two-dimensional load map or pressure map P[u, v]. The two-dimensional pressure map is used to determine whether the corresponding shelf has goods, the occupied area (footprint), and to estimate the quality of the goods (total load), the center of gravity position (offset), and as a signal source for feedback or verification before and after locking.

[0085] The data acquisition circuit can convert and summarize the pressure signals sensed by the pressure sensor array 13 to form pressure data that the controller can recognize and process. In other words, the data acquisition circuit is configured to receive the pressure data acquired by the pressure sensor array 13. Therefore, the data acquisition circuit can incorporate multiple signal processing components, such as interface circuits, analog-to-digital converters, digital signal processors, power management modules, and communication interface modules.

[0086] The limiting device 12 is used to position goods in a shape-adaptive manner. The limiting device 12 includes multiple self-locking posts. The multiple self-locking posts form multiple locking points in the storage compartment; the rack controller is connected to the data acquisition circuit and the multiple self-locking posts.

[0087] In some embodiments, such as Figure 3 As shown, the self-locking pin includes an actuation unit 14, a pushing elastic element 15, a reset driving element 16, and an electromagnetic release mechanism 17. The pushing elastic element 15 is connected to the actuation unit 14 and is used to push the actuation unit 14 to the locked position. The reset driving element 16 is connected to the actuation unit 14 and is used to reset the actuation unit 14 from the locked position to the initial position. The electromagnetic release mechanism 17 is disposed in the locked position and the initial position and is used to lock or unlock the actuation unit 14 according to a pulse current signal.

[0088] In some embodiments, the actuation unit 14 is a cylindrical structure with slots at both ends, and the ends of the actuation unit 14 are connected to the pushing elastic member 15 and are close to the reset drive member 16. Figure 4 As shown, the electromagnetic release mechanism 17 includes an electromagnetic adsorption component 171, a locking component 172, and a return spring 173; the return spring 173 is disposed between the locking component 172 and the electromagnetic adsorption component 171. Limiting grooves are provided in both the locked position and the initial position, and the locking component 172 is movably disposed within the limiting grooves and moves along the limiting grooves.

[0089] When the electromagnetic adsorption component 171 is not energized, it does not exert a magnetic attraction force on the locking component 172. The locking component 172, under the force of the return spring 173, moves away from the electromagnetic adsorption component 171, thus entering the slot at the end of the actuating unit 14 and limiting the position of the actuating unit 14, achieving a locking action. When the electromagnetic adsorption component 171 is energized, it exerts a magnetic attraction force on the locking component 172. At this time, because the magnetic attraction force is greater than the force of the return spring 173, the locking component 172, under the influence of the magnetic attraction force, moves towards the electromagnetic adsorption component 171, causing it to move out of the slot at the end of the actuating unit 14, ceasing to limit the position of the actuating unit 14, thus achieving an unlocking action.

[0090] By setting electromagnetic release mechanisms 17 in the locked and initial positions, locking and unlocking actions can be performed in the locked and initial positions respectively, controlling the position of the action unit 14. For example, the action unit 14 in the self-locking pin is an electromagnetic release pin unit with a bistable structure and upper and lower locking. The bistable structure is used to form a stable state in the upper locked position (UP) and the lower initial position (DOWN). Each pin unit includes a guide hole or guide sleeve to ensure that the pin moves in a vertical direction. The push elastic member 15 can push the pin to the UP position after the lower lock is released by compressing the spring, and is therefore also called the push elastic member. The locking member 172 of the upper electromagnetic release mechanism 17 is an upper locking member (UP latch), which is used to automatically lock and hold after the pin unit moves to the UP position; the locking member 172 of the lower electromagnetic release mechanism 17 is a lower locking member (DOWN latch), which is used to automatically lock and hold after the pin unit moves to the DOWN position. The electromagnetic release mechanism 17 can release the UP latch or DOWN latch according to the received pulse current.

[0091] The reset drive 16 can be a fallback electromagnetic attraction device, a small pull-down mechanism, or a gravity-assisted reset mechanism, used to return the pin cell to the DOWN position after the UP latch is released. For better reset performance, a fallback electromagnetic attraction device 171 or an equivalent mechanism can be used in engineering to move the pin cell from the UP position to the DOWN position through the combined action of electromagnetic force and gravity.

[0092] Based on the structure of the self-locking column described above, the operation logic of the self-locking column is as follows: when performing the RAISE action, the electromagnetic release mechanism 17 releases the DOWN latch, causing the push elastic element to push the pin unit to the UP position. Then, through the pulse control signal, the UP latch is driven to insert into the slot of the pin unit and lock the pin unit in the UP position.

[0093] Similarly, during the RETRACT action, the electromagnetic release mechanism 17 first releases the UP latch, and the reset drive 16 returns the pin unit to the DOWN position and re-compresses the push-up elastic element. Then, the DOWN latch is locked by a pulse control signal.

[0094] The rack controller can be used to perform locking or unlocking actions by controlling the self-locking columns at different locking points to accommodate goods of different shapes. Specifically, the rack controller is configured to generate a two-dimensional pressure distribution map based on pressure data, and to control the self-locking columns to perform locking or unlocking actions based on the two-dimensional pressure distribution map.

[0095] During locking control, the rack controller can use a row / column addressing drive network (Row / Col Bus) to bus address the pin unit array, reducing wiring and supporting the erection of a pin unit (RAISE(i,j)) and the retraction of a pin unit (RETRACT(i,j)). To avoid ghost points, each pin unit can also incorporate anti-crosstalk structures, such as diodes or equivalent isolation elements.

[0096] To achieve motion control of the self-locking columns, in some embodiments, the rack controller is also configured to execute a pressure map-based adaptive locking feedback control method for moving racks. This method can be applied to the rack controller to achieve adaptive control of the shape of placed goods. Figure 5 As shown, the method includes:

[0097] S101, Obtain pressure data.

[0098] During goods handling, the rack controller can establish adaptive locking feedback control for the moving rack 1 at one end, thus acquiring pressure data through the pressure sensor array 13. In some embodiments, when acquiring pressure data, the rack controller can send a data acquisition request to the data acquisition circuit. After receiving the data acquisition request, the data acquisition circuit can summarize and convert the pressure signals collected by multiple pressure sensors in the pressure sensor array 13 into digital pressure data, and then feed the pressure data back to the rack controller.

[0099] Since the pressure data is acquired by multiple pressure sensors in the pressure sensor array 13, each pressure sensor is distributed at a different detection position. Therefore, in the acquired pressure data, each pressure signal can also be associated with the detection position information of the corresponding pressure sensor. For example, the pressure data can include multiple pressure data points, each pressure data point including pressure value and position coordinate information, that is, the pressure data can be represented as P={P i (F) i x i y i ); i = 1, 2, ..., N}. Where, F i This represents the pressure value collected by the i-th pressure sensor; (x i y i ) represents the position coordinates corresponding to the i-th pressure sensor.

[0100] In some embodiments, after acquiring the pressure data, the rack controller can also preprocess the pressure data. The preprocessing process may include necessary preprocessing items such as zero-point calibration, filtering, and outlier removal to ensure that the pressure data is usable.

[0101] S102. Generate a two-dimensional pressure distribution map based on the pressure data.

[0102] After acquiring the pressure data, a two-dimensional pressure distribution map can be generated based on the pressure data. That is, the rack controller can read two-dimensional load or pressure data for each storage compartment or each smart pallet 11 at the rack end to obtain the load distribution map of the current goods, i.e., the two-dimensional pressure distribution map.

[0103] A two-dimensional pressure distribution map can include multiple data points, each corresponding to a pressure sensor and the pressure value collected by that sensor. Therefore, when generating the two-dimensional pressure distribution map, the rack controller can read the pressure values ​​and corresponding location coordinates from the pressure data. Then, according to the location coordinates, the pressure values ​​are arranged and mapped into image data to form the two-dimensional pressure distribution map.

[0104] S103. Calculate cargo state quantities based on two-dimensional pressure distribution map.

[0105] After generating the two-dimensional pressure distribution map, the shelf controller can estimate the cargo status based on the pressure values ​​of each data point in the two-dimensional pressure distribution map. That is, it calculates the cargo status quantities based on the two-dimensional pressure distribution map. The cargo status quantities include the cargo existence status, cargo occupancy area, cargo weight, cargo center of gravity position, center of gravity offset degree, and shape risk index.

[0106] For example, when estimating the condition of goods at the shelf end, the shelf controller can calculate the presence status of goods based on a two-dimensional pressure distribution map, i.e., whether goods are placed in the storage compartment; the area occupied by the goods, i.e., the approximate range occupied by the goods, equivalent to the goods' footprints; the total weight of the goods, which can be estimated by load integration; and the position and degree of the goods' center of gravity, with higher risk corresponding to a greater distance between the center of gravity and the geometric center of the storage compartment. The position and degree of the center of gravity can also include the position and degree of the center of gravity of the entire shelf, which can be calculated based on the position and degree of the center of gravity of each item, as well as the physical structural parameters of the shelf. The risk index for the shape of the goods can indicate whether the area occupied by the goods is elongated or narrow, or whether it is fragmented into multiple small clusters, etc.

[0107] In some embodiments, when the shelf controller performs the calculation of cargo status based on the two-dimensional pressure distribution map, it can also obtain weight information collected by multiple pressure sensors in the pressure sensor array 13 by traversing the data points of the two-dimensional pressure distribution map. Then, target data points are extracted from the data points, wherein the target data points are the data points corresponding to pressure sensors within the two-dimensional pressure distribution map whose weight information is greater than or equal to a weight threshold. Finally, by accumulating the weight information of the target data points, the cargo presence status and cargo weight are obtained.

[0108] For example, after generating a two-dimensional pressure distribution map, the pressure sensor that collected the cargo weight information can be determined by reading the pressure value corresponding to each data point in the two-dimensional pressure distribution map. For example, for the pressure value F of a single data point... i This can be compared with a weight threshold F0 set based on the sensor's zero-degree drift, when the pressure value F i When the weight is greater than or equal to the weight threshold F0, it indicates that the sensor has collected the weight information of the goods. Therefore, this data point can be marked as the target data point P. i '.

[0109] When a target data point exists or the number of target data points is greater than or equal to a quantity threshold, it can be determined that goods have been placed in the current storage cell, thus generating a goods presence status. After determining that goods exist, the shelf controller then targets the target data point P. i Perform load integration to estimate the weight of the cargo by summing.

[0110] Furthermore, the shelf controller can calculate the center of gravity position and degree of center of gravity shift of the goods based on the weight of multiple goods and the position of the shelves in the moving shelf 1. For example, for a pressure sensor array 13 composed of m×n pressure sensors, by evenly distributing multiple pressure sensors on a plane, the position of each pressure sensor can be represented by coordinates. And, from the pressure data collected by each sensor, a set of pressure values ​​P can be obtained. i,j Where i and j represent the row and column positions of the sensor in the array, respectively. The x-coordinate of the centroid is then calculated using the following formula. center :

[0111]

[0112] And calculate the ordinate y of the centroid according to the following formula. center :

[0113]

[0114] Where m and n are the number of rows and columns of the sensor array, respectively; P i,j It is the pressure value of the sensor in the i-th row and j-th column; x i,j and y i,j This represents the coordinates of the sensor in the i-th row and j-th column. This is determined by the x-coordinate of the centroid. center The y-coordinate of the centroid center By combining them, the center of gravity of the cargo can be obtained.

[0115] After calculating the position of the cargo's center of gravity, the degree of center of gravity offset can be obtained by comparing the coordinates of the cargo's center of gravity with the position of the geometric center point of the cargo compartment and calculating the distance between the coordinates of the cargo's center of gravity and the geometric center point of the cargo compartment.

[0116] The shelf controller can also define the goods occupation area based on target data points, and traverse the occupation dimensions of the goods occupation area. The occupation dimensions include a first dimension and a second dimension. In some embodiments, the first dimension of the locking member 172 is the distance between the edge target data points in a first direction; the second dimension of the locking member 172 is the distance between the edge target data points in a second direction; the first direction of the locking member 172 is the direction of maximum distance between the edge target data points; the second direction of the locking member 172 is perpendicular to the first direction of the locking member 172. Then, a goods shape index, including the ratio of the first dimension to the second dimension, is calculated based on the occupation dimensions, thereby determining a shape risk index according to the shape range to which the goods shape index belongs.

[0117] For example, by extracting target data points located at the edge contour, the area occupied by the goods can be determined. Then, by calculating the distance between two edge target data points one by one, the two target data points with the farthest distance can be determined. The direction of the line connecting the two target data points is the first direction. Then, based on the first direction, a perpendicular second direction is determined. Finally, based on the cumulative number or distance of target data points in the two directions, the first dimension and the second dimension are obtained.

[0118] In some embodiments, the first dimension is the cumulative number of target data points in the first direction; the second dimension is the cumulative number of target data points in the second direction; the first direction is the direction with the largest cumulative number of target data points, and similarly, the second direction is perpendicular to the first direction.

[0119] After obtaining the first and second dimensions, the ratio of the first to the second dimension is calculated to indicate whether the occupied area is elongated, thereby obtaining a cargo shape index. In some embodiments, the cargo shape index may further include a clustering index to indicate whether the cargo occupied area is divided into multiple small clusters. The clustering index can be obtained by comparing target data points to determine the continuity of the occupied area based on the interval distance between the target data points.

[0120] S104. Determine the target locking point based on the cargo status quantity.

[0121] After calculating the cargo status quantity based on the two-dimensional pressure distribution map, the rack controller can determine the cargo edge contour according to the cargo occupancy area in the cargo status quantity, and then determine the target locking points that can adaptively limit the shape of the cargo based on the cargo edge contour. The target locking points are multiple locking points with the outer contour of the cargo occupancy area as the boundary.

[0122] When determining the target locking point, the rack controller can extract the outer contour of the goods based on the area occupied by the goods in the goods status data. The outer contour can be determined based on the position of the outermost pressure sensor in the pressure sensor array 13 that can detect the pressure generated by the goods. Then, the target locking point is determined based on the position of the outermost pressure sensor, that is, the target locking point is the locking point on the outside that is closest to the outermost pressure sensor.

[0123] S105. Generate a locking control signal according to the target locking point, and send the locking control signal to the self-locking column at the target locking point.

[0124] After determining the target locking point, the rack controller can generate a locking control signal according to the target locking point, and then send the locking control signal to the self-locking column at the target locking point so that the self-locking column can perform a locking action according to the locking control signal.

[0125] For example, at the shelf end, a locking control signal can be generated to determine the set of locking points that should be locked this time. That is, the shelf end automatically generates the locking points that need to be erected this time based on the estimated area occupied by the goods. When locking, outer contour locking can be performed based on basic rules. Using the outer contour of the area occupied by the goods as the boundary, self-locking posts are erected at the bottom points corresponding to the contour to form a fence-like barrier, restricting the goods from sliding in the storage compartment.

[0126] When adaptively limiting the position of goods based on their shape, in addition to the basic rules for locking according to the outer contour, enhanced rules can be used for special goods, such as those with significant eccentricity, elongated shapes, or large weights, to further improve the effectiveness of limiting the position of the goods. Therefore, in some embodiments, the shelf controller can also calculate the total weight of the goods by accumulating the weights of the goods in multiple shelves, and then obtain enhanced evaluation parameters. These enhanced evaluation parameters are at least one of the following: total weight of the goods, degree of center of gravity offset, and goods shape index.

[0127] The enhancement evaluation parameters are then compared with preset enhancement conditions. Corresponding to the indicators included in the enhancement evaluation parameters, the preset enhancement conditions may include at least one of the following: total weight threshold, center of gravity offset threshold, and shape indicator threshold. Based on the comparison results, when the enhancement evaluation parameters meet the preset enhancement conditions, a support line is determined according to the enhancement evaluation parameters. At this point, an enhancement locking signal can be generated based on the support line, and the enhancement locking signal can be sent to the self-locking post corresponding to the support line.

[0128] For example, by setting enhanced rules and performing cargo status detection based on these rules, when the total weight of the cargo exceeds a total weight threshold, or the center of gravity shift exceeds a center of gravity shift threshold, or the cargo shape index exceeds a shape index threshold, it can be detected that the current cargo has significant eccentricity, a slender shape, and a large weight. In this case, one or more locking points of support lines can be added within the cargo's occupied area. Then, enhanced locking signals are generated based on the support lines and sent to the corresponding self-locking posts on the support lines, causing all self-locking posts on the support lines to stand up, thereby improving the anti-tipping or anti-shaking capability.

[0129] Considering that locking points within the storage compartment may already exist, such as some points remaining after the previous cycle or round of handling, the shelf end may not employ a complete retraction and re-erectation locking method. Instead, a strategy of locking or unlocking only the changed parts can be adopted. Therefore, in some embodiments, the shelf controller can obtain the current action state of the self-locking column, wherein the current action state includes the current action position of the locking member 172. Then, a current set of locking points is generated based on the current action state, and a target set of locking points is generated based on the target locking point.

[0130] Then, the action lock point set is determined by comparing the current lock point set and the target lock point set. This action lock point set includes lock points that need to be retracted and lock points that need to be locked. A pulse current signal is then generated based on the action lock point set and sent to the electromagnetic release mechanism 17.

[0131] For example, by obtaining the current action state of the self-locking pins, if some pins are currently in an upright state, but this calculation determines that maintaining the upright state is unnecessary, then a retraction action is performed on these pins. Similarly, if some pins are determined to be upright in this calculation, but are not currently in an upright state, then an upright action is performed on these pins. For pins that still need to be upright and are already upright, the action is not repeated. This method of only modifying differences can reduce the number of actions, reduce energy consumption and wear, and avoid unnecessary changes to constraints, thus improving picking friendliness.

[0132] Therefore, the aforementioned mobile shelving 1 can achieve closed-loop control with adaptive locking at the shelf level. During this adaptive locking closed-loop control process, the input data includes the current mode, the current lock point set C, the pressure map acquisition interface, threshold, batch size, pulse duration, waiting time, and other parameters. The output data includes the updated lock point set C; whether handling is permitted; the risk level; speed / acceleration constraints (vmax, amax); and fault codes.

[0133] When performing compartment-level adaptive locking control, a two-dimensional pressure distribution map P can be acquired first and preprocessed to obtain a preprocessed pressure map. Then, from the pre-processed pressure map... An occupancy map (Occ) is generated to determine which areas contain goods. Goods clusters are then identified from the Occ map, and their status characteristics, including weight, center of gravity shift, and shape risk, are estimated. Based on the goods status, the target lock point set (D) is calculated, including outer contour locking and, if necessary, adding support lines. Next, the set of lock points (R) to be retracted (points currently erected but not needed this time) and the set of lock points (U) to be erected (points needed this time but not currently erected) are calculated. Then, the retraction action is executed in batches on R, and the erection action is executed in batches on U, erecting the corresponding points.

[0134] In some embodiments, the shelf controller can also verify the execution result of the self-locking column. Therefore, the shelf controller can also acquire current data and multi-frame pressure data, extract abnormal current features from the current data, and obtain the cargo sliding distance by comparing the multi-frame pressure data. Verification information is then calculated, and verification output information is generated based on the verification information. The verification information includes at least one of the following: positioning status, abnormal current features, and cargo sliding distance; the verification output information includes the verification result and a failure reason code when the verification result is a locking failure.

[0135] To verify the execution result of the self-locking column, the self-locking column also includes a position detection device. The position detection device is set in the locked position and the initial position to detect the position status of the action unit 14 or the locking member 172. For example, the position detection device can be a proximity switch or Hall element set near the locked position or the initial position of the self-locking column to detect whether the action unit 14 or the locking member 172 has been moved to the correct position and obtain the position status.

[0136] The limiting device 12 also includes a current sensor connected to the electromagnetic release mechanism 17 for detecting current data. The current data may include abnormal current characteristics, such as those occurring during open circuits, short circuits, or jamming. Therefore, by detecting whether the current data contains abnormal current characteristics, it is possible to indirectly detect whether the self-locking pin is jammed or exhibiting abnormal phenomena.

[0137] Position detection devices and current sensors can be connected to the rack controller, enabling the rack controller to perform execution result verification at the rack end after the locking action is completed. This verification determines whether the locking was successful, whether there is any jamming or abnormality, etc. The execution result verification can employ one or more of the following methods in combination: self-locking post position detection, drive current characteristic detection, and secondary load distribution consistency detection. During secondary load distribution consistency detection, the stability characteristics of the occupied boundary before and after execution can be compared to detect suspected slippage. By verifying the execution result, verification output can be obtained, including whether the locking was successful or failed, and the failure reason code, for server-side decision-making.

[0138] The handling robot 2 is used to lift and move the mobile shelf 1 as a whole. Therefore, the handling robot 2 can include a moving mechanism, a lifting mechanism, and a robot controller. The moving mechanism can move autonomously along a set path and can use wheeled, tracked, articulated, or rail-driven mechanisms, depending on the specific application. The lifting mechanism can be used to lift the mobile shelf 1 as a whole and can also use electric screw, scissor, rack and pinion, hydraulic, or pneumatic mechanisms, depending on the specific application.

[0139] In some embodiments, the bottom of the shelf body is provided with a lifting and positioning component, which includes a positioning groove and an anti-slip stop; the anti-slip stop is provided on the inner wall of the positioning groove; the top of the lifting mechanism is provided with a positioning protrusion, which can be embedded in the positioning groove.

[0140] For example, the handling robot 2 can be equipped with a docking or lifting interface, and the bottom of the mobile shelf 1 is reserved with an entry space for the handling robot 2, and is equipped with positioning holes or positioning grooves and anti-slip stops, so that the handling robot 2 is more stable when it is lifted, and the mobile shelf 1 will not tip over as a whole due to the speed of the handling robot 2.

[0141] The robot controller connects the moving mechanism and the lifting mechanism, and is configured to drive the moving mechanism and the lifting mechanism according to the handling task. Furthermore, the robot controller is also configured to execute the control methods corresponding to the handling or lifting actions, such as... Figure 6 As shown, the method includes:

[0142] S201. Obtain location information;

[0143] S202. Plan the movement path based on location information;

[0144] S203. Analyze motion constraint parameters from the handling task;

[0145] S204. Generate transport control commands based on the movement path and motion constraint parameters, and send the transport control commands to the moving mechanism and the lifting mechanism.

[0146] When performing goods handling, the handling robot 2 can first obtain location information. This location information includes the current location of the handling robot 2, the shelf position of the mobile shelf 1, and the picking location. The current location of the handling robot 2 can be obtained through its built-in positioning module. The shelf position of the mobile shelf 1 can be obtained through its built-in positioning module. The picking location can be pre-set according to the application scenario layout.

[0147] After acquiring location information, the robot controller can plan a movement path based on a preset scene path planning algorithm. This movement path includes a pickup path and a transport path. The pickup path is planned based on the current location and the shelf location; the transport path is planned based on the shelf location and the picking location.

[0148] After planning the movement path, motion constraint parameters are parsed from the handling task issued by the server. Then, handling control instructions are generated based on the movement path and motion constraint parameters, and sent to the moving mechanism and lifting mechanism so that the moving mechanism and lifting mechanism can perform goods handling and goods lifting according to the handling control instructions, thereby moving the mobile rack 1 from the rack position to the picking position.

[0149] The server connects to the shelf controller and the robot controller. Furthermore, the server is configured to execute a stress map-based handling control method, such as... Figure 7 As shown, the method includes:

[0150] S301, Receive two-dimensional pressure distribution map;

[0151] S302. Extract cargo status features from the two-dimensional pressure distribution map;

[0152] S303. Generate motion constraint parameters based on cargo status characteristics;

[0153] S304. Generate a transport task based on motion constraint parameters and send the transport task to the robot controller.

[0154] When performing handling control, the server can first receive a two-dimensional pressure distribution map and extract cargo status characteristics from the two-dimensional pressure distribution map. The cargo status characteristics include at least one of the following: cargo weight, degree of center of gravity offset, and shape risk indicators that can affect the stability of the cargo handling process.

[0155] Then, motion constraint parameters are generated based on the cargo's state characteristics. These parameters include at least one of the following: maximum speed coefficient, maximum acceleration coefficient, turning coefficient, and emergency stop threshold. For example, the shelf-side or server-side can assess handling risks and generate robot motion constraints. This involves evaluating the risk level based on information such as weight, center of gravity shift, shape risk indicators, locking success, and locking quality (e.g., lock point coverage, number of anomalies), and generating robot motion constraint parameters accordingly. These parameters include at least the maximum speed coefficient and the maximum acceleration coefficient. The maximum speed coefficient is the ratio of the maximum handling speed to the standard speed; the maximum acceleration coefficient is the ratio of the maximum handling acceleration to the standard acceleration. If necessary, other robot motion constraint parameters such as turning coefficients or emergency stop thresholds may also be included.

[0156] To generate motion constraint parameters, in some embodiments, when the server generates motion constraint parameters based on cargo state characteristics, it may first call a mapping database. The mapping database may include multiple state feature intervals and motion constraint parameters associated with those intervals. Then, the target state feature interval to which the cargo state characteristics belong is obtained, and the motion constraint parameters are determined based on the target state feature interval.

[0157] For example, after verifying the locking result, the shelf controller can obtain verification results such as whether the locking was successful (ok) and the locking quality (q). Then, it assesses the risk level (risk) based on parameters such as the cargo status characteristics and locking quality. The risk level is then mapped to robot motion constraints (vmax, amax). A parameter (permit) is generated to characterize whether handling is permitted. If it's OK and the risk does not exceed the threshold, handling is allowed; otherwise, handling is prohibited. Finally, {permit, risk, vmax, amax, fault_code} is reported to the server, and the target lock point set D is updated to the current lock point set C.

[0158] Then, based on the reported content, the server and robot engage in collaborative interaction. The input data for this collaborative interaction consists of the shelf-side report of {permit, risk, vmax, amax} and a queue of tasks to be executed. The output data is the robot's handling task or action carrying constraint parameters. During the collaborative interaction, the server can first receive the shelf-side report results. By reading the report results, if permit=false, the shelf handling task is canceled or postponed, the reason is recorded, and manual verification or replanning is triggered. If permit=true, a handling task is generated, and (vmax, amax) is written to the task constraint field.

[0159] After generating motion constraint parameters, the server generates a handling task based on these parameters and sends the task to the robot controller. This allows the robot to perform the handling operation according to the task. In other words, the robot performs the handling and injects constraints. Upon receiving a task, the robot can inject the received velocity or acceleration constraints into its local motion controller for amplitude limiting, trajectory planning, and execution of lifting and handling according to the constraints, ensuring safety and stability. For example, a task can be sent to the robot, which, upon receiving the task, injects (vmax, amax) into the motion controller for amplitude limiting and then performs the handling. Furthermore, the robot can report the execution status of the handling task, allowing the server to log the process for strategy optimization.

[0160] In some embodiments, as a refinement and extension of the specific implementation of the above embodiments, and in order to fully illustrate the specific implementation process of this embodiment, some embodiments of this application also provide an adaptive locking feedback control method for a moving shelf based on a pressure map, such as... Figure 8 As shown, the method includes:

[0161] S401, Obtain resampled data;

[0162] S402. Calculate and update the locked points based on the resampled data;

[0163] S403. Generate a differential update signal based on the updated lock point and the lock control signal;

[0164] S404. Send the differential update signal to the self-locking column corresponding to the update lock point.

[0165] The mobile shelf 1 can be dynamically relocked at the picking station. During dynamic relocking, resampling data can be acquired first. The resampling data is the load data obtained by weighing the remaining goods after any item in the mobile shelf 1 has been picked.

[0166] Then, the updated locking point is calculated based on the resampled data, and a differential update signal is generated based on the updated locking point and the locking control signal. The differential update signal is then sent to the self-locking column corresponding to the updated locking point. For example, when the handling robot 2 moves the mobile shelf 1 to the picking station and some goods are taken away, causing a change in the state of the remaining items in the shelf, the shelf end can re-collect and estimate the state of the remaining goods. That is, by rereading the load distribution, the occupied area, weight, center of gravity shift, and other states of the remaining goods can be estimated.

[0167] Then, the new locking points are recalculated, following the same contour locking and necessary support line rules as before the item was retrieved, resulting in a new target locking point distribution. A differential update is then performed, using a process of first canceling and then re-locking to control the locking and unlocking of the self-locking posts. Furthermore, to avoid disrupting manual operations, obstructions or unnecessary locking points are first canceled at the shelf end before new locking points are added, making manual picking smoother while ensuring that remaining items remain restrained during subsequent handling.

[0168] Similarly, for re-implemented constraints, the execution results can be verified and necessary constraints can be updated. For example, if the risk level changes significantly after relocking, especially if the risk increases, the shelf end can issue new constraints to the server or robot end to ensure the safety of the next handling segment.

[0169] Furthermore, the system can also perform anomaly monitoring and emergency coordination during the handling process. Specifically, when the rack is in the handling state, it can selectively monitor for abnormal changes in load distribution at low frequency. If a significant anomaly occurs, such as suspected slippage or tipping, the rack can issue an emergency deceleration or stop command to the robot or server. The robot will then immediately execute the deceleration or stop action, while the server records the event and triggers manual intervention or replanning.

[0170] By applying the technical solutions of the above embodiments, the adaptive locking feedback control system for the mobile shelving 1 based on the pressure map described in the above embodiments can maintain safety constraints even after picking, reducing the risk of slippage and tipping. When the picking station supports the re-sensing and locking strategy reconstruction of remaining goods in the compartment, and when the removal of some goods causes changes in the center of gravity and shape, it can automatically cancel unnecessary locking points and fill in new locking points, so that the remaining goods are once again in an effective restraint state, thereby significantly reducing the risk of slippage, scattering, and tipping during subsequent handling.

[0171] The system enables risk-differentiated speed control, improving system throughput. By combining state information such as rack load distribution, weight, and center of gravity shift with the locking execution results, it outputs speed and acceleration constraints that can be directly used by the robot. Higher motion parameters are allowed for racks with lower risk and good locking performance, while automatic speed limits or prohibition of handling are implemented for racks with higher risk or insufficient locking. This improves overall handling efficiency and system throughput while ensuring safety.

[0172] The system can perform closed-loop verification and anomaly diagnosis of the locking process, improving reliability. After the locking action is completed, the result is verified, and traceable cause information is provided for abnormal situations, such as incomplete locking, abnormal drive, abnormal sensor, or excessive risk. This allows the system to detect potential faults in advance before handling and take measures such as restricting handling, slowing down, or manual intervention, reducing secondary losses caused by faults during handling and improving the system's operational reliability and maintainability.

[0173] The system is compatible with the RMFS architecture, making it easy to integrate and deploy in engineering projects. Using storage racks or pallets as the basic units for control and decision-making, it does not require structural modifications to existing warehouse management systems or the robot itself. It only needs to provide information exchange at the communication layer, such as handling permissions, risk levels, motion constraints, and alarms, to work collaboratively with existing scheduling systems and robot control systems, demonstrating strong engineering feasibility.

[0174] The system can also be naturally extended to multi-source sensing and multi-strategy control to adapt to complex scenarios. The system's control chain is based on the framework of sensing, decision-making, execution, verification, and constraint output. It can be extended to include more sensor information such as RFID, vision, inertial measurement, and historical statistics, as well as strategy rules such as priority protection of fragile items and adjustment of abnormal self-learning thresholds, without changing the overall process. This allows it to adapt to multi-category, multi-compartment, and highly dynamic handling scenarios.

[0175] In summary, this application provides an adaptive locking feedback control method for a moving shelf 1 based on a pressure map, the method comprising:

[0176] Obtain stress data;

[0177] A two-dimensional pressure distribution map is generated based on the pressure data, and the self-locking column is controlled to perform locking or unlocking actions based on the two-dimensional pressure distribution map.

[0178] Cargo status features are extracted from the two-dimensional pressure distribution map. The cargo status features include at least one of cargo weight, center of gravity offset, and shape risk indicators.

[0179] Motion constraint parameters are generated based on the cargo state characteristics. The motion constraint parameters include at least one of the following: maximum speed coefficient, maximum acceleration coefficient, turning coefficient, and emergency stop threshold.

[0180] Based on the motion constraint parameters, a transport task is generated, and the transport task is sent to the robot controller so that the robot controller drives the moving mechanism and the lifting mechanism according to the transport task.

[0181] By applying the technical solutions of the above embodiments, the adaptive locking feedback control method for the mobile shelf 1 based on pressure maps described in the above embodiments can collect pressure data of placed goods through the pressure sensor array 13. The shelf controller generates a two-dimensional pressure distribution map based on the pressure data, and controls the limit device 12 to perform locking or unlocking actions according to the two-dimensional pressure distribution map, thereby achieving adaptive shape position limitation of goods at multiple locking points. The server can extract the state characteristics of the goods based on the two-dimensional pressure distribution map and generate a handling task containing motion constraint parameters, so as to control the handling robot 2 to complete the handling of goods through the handling task. The method can maintain the safety constraint of goods during handling and picking, reduce the risk of slippage and tipping, realize speed adjustment control based on risk differentiation, and improve the efficiency of goods handling.

[0182] This application also provides a computer device, specifically a personal computer, server, network device, etc. The computer device includes a bus, processor, memory, and communication interface, and may also include input / output interfaces and a display device. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer device stores location information. The network interface of the computer device is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the steps in the various method embodiments.

[0183] Those skilled in the art will understand that the structure of the computer device described above is only a partial structure related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. A specific computer device may include more or fewer components, or combine certain components, or have different component arrangements.

[0184] In one embodiment, a computer-readable storage medium is also provided, which may be non-volatile or volatile, having stored thereon a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0185] In one embodiment, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0186] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data that have been authorized by the user or fully authorized by all parties.

[0187] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods.

[0188] Any references to memory, database, or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc.

[0189] Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can take many forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0190] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, graphics processors, digital signal processors, programmable logic devices, etc., and are not limited to these.

[0191] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0192] The embodiments described above are merely examples of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the protection scope of this application.

Claims

1. A pressure-map-based adaptive locking feedback control system for a moving shelf, characterized in that, The system includes: A mobile shelving unit includes a shelving body, multiple load-bearing surfaces, and a shelving controller. At least one shelf compartment is provided on each load-bearing surface. Each shelf compartment has a smart pallet and a limiting device at its bottom. The smart pallet includes a pressure sensor array and a data acquisition circuit connected to the pressure sensor array. The data acquisition circuit is configured to receive pressure data collected by the pressure sensor array. The limiting device includes multiple self-locking posts. The multiple self-locking posts form multiple locking points in the shelf compartment. The shelving controller is connected to the data acquisition circuit and the multiple self-locking posts. The shelving controller is configured to generate a two-dimensional pressure distribution map based on the pressure data, and to control the self-locking posts to perform locking or unlocking actions based on the two-dimensional pressure distribution map. The self-locking column includes an actuating unit, a pushing elastic element, a resetting drive element, and an electromagnetic release mechanism. The pushing elastic element is connected to the actuating unit and is used to push the actuating unit to the locked position. The resetting drive element is connected to the actuating unit and is used to reset the actuating unit from the locked position to the initial position. The electromagnetic release mechanism is disposed at the locked position and the initial position and is used to lock or unlock the actuating unit according to a pulse current signal. The shelf controller is further configured to: acquire the current actuating state of the self-locking column, the current actuating state including the current actuating position of the actuating unit; generate a current locking point set according to the current actuating state, and generate a target locking point set according to the target locking point; determine an actuating locking point set by comparing the current locking point set and the target locking point set, the actuating locking point set including locking points that need to be retracted and locking points that need to be locked; generate the pulse current signal based on the actuating locking point set, and send the pulse current signal to the electromagnetic release mechanism. The self-locking column further includes a positioning detection device; the positioning detection device is disposed at the locked position and the initial position, and is used to detect the positioning state of the action unit; the limiting device further includes a current sensor, the current sensor is connected to the electromagnetic release mechanism, and is used to detect current data; the positioning detection device and the current sensor are connected to the shelf controller; the shelf controller is further configured to: acquire the current data and multiple frames of pressure data; extract abnormal current features from the current data; obtain the cargo sliding distance by comparing the multiple frames of pressure data; calculate verification information, the verification information including at least one of the positioning state, the abnormal current features, and the cargo sliding distance; generate verification output information based on the verification information, the verification output information including a verification result and a failure reason code when the verification result is a locking failure; A transport robot includes a moving mechanism, a lifting mechanism, and a robot controller, wherein the robot controller is connected to the moving mechanism and the lifting mechanism; the robot controller is configured to drive the moving mechanism and the lifting mechanism according to a transport task. A server is connected to the shelf controller and the robot controller; the server is configured to: Receive a two-dimensional pressure distribution map; Cargo status features are extracted from the two-dimensional pressure distribution map. The cargo status features include at least one of cargo weight, center of gravity offset, and shape risk indicators. Motion constraint parameters are generated based on the cargo state characteristics. The motion constraint parameters include at least one of the following: maximum speed coefficient, maximum acceleration coefficient, turning coefficient, and emergency stop threshold. The system generates a transport task based on the motion constraint parameters and sends the transport task to the robot controller.

2. The system according to claim 1, characterized in that, The shelf controller is also configured to: Acquire the pressure data; A two-dimensional pressure distribution map is generated based on the pressure data; The cargo status parameters are calculated based on the two-dimensional pressure distribution map. The cargo status parameters include the cargo existence status, cargo occupied area, cargo weight, cargo center of gravity position, degree of center of gravity offset, and shape risk index. The target locking points are determined based on the cargo status quantity. The target locking points are multiple locking points with the outer contour of the cargo-occupied area as the boundary. Generate a locking control signal according to the target locking point, and send the locking control signal to the self-locking column at the target locking point.

3. The system according to claim 2, characterized in that, The rack controller performs calculations of cargo status quantities based on the two-dimensional pressure distribution map and is also configured to: The data points of the two-dimensional pressure distribution map are traversed, and the data points include weight information collected by multiple pressure sensors in the pressure sensing array; Extract target data points from the data points, where the target data points are the data points corresponding to the pressure sensors whose weight information is greater than or equal to a weight threshold. The presence status and weight of the goods are obtained by accumulating the weight information of the target data points; Based on the weights of the goods and the positions of the shelves on the mobile shelf, calculate the position of the center of gravity of the goods and the degree of center of gravity offset. The cargo-occupied area is defined based on the target data points, and the occupancy size of the cargo-occupied area is traversed. The occupancy size includes a first size and a second size. The first size is the distance between edge target data points in a first direction; the second size is the distance between edge target data points in a second direction; the first direction is the direction of maximum distance between edge target data points; the second direction is perpendicular to the first direction. The cargo shape index is calculated based on the occupied dimensions, and the cargo shape index includes the ratio of the first dimension to the second dimension; The shape risk index is determined based on the shape range to which the cargo shape index belongs.

4. The system according to claim 3, characterized in that, The shelf controller is also configured to: The total weight of the goods is calculated by summing the weights of the goods in multiple of the aforementioned compartments; Obtain enhanced evaluation parameters, wherein the enhanced evaluation parameters are at least one of the total weight of the cargo, the degree of center of gravity offset, and the cargo shape index; When the enhancement evaluation parameters meet the preset enhancement conditions, the support line is determined based on the enhancement evaluation parameters; An enhanced locking signal is generated based on the support line, and the enhanced locking signal is sent to the self-locking post corresponding to the support line.

5. The system according to claim 2, characterized in that, The shelf controller is also configured to: Acquire resampled data, which is load data obtained by performing weight collection on the remaining goods after any item is picked in the mobile shelf; The locked points are calculated and updated based on the resampled data; A differential update signal is generated based on the updated lock point and the lock control signal; The differential update signal is sent to the self-locking column corresponding to the update lock point.

6. The system according to claim 1, characterized in that, The server generates motion constraint parameters based on the cargo state characteristics and is further configured to: Call the mapping database, which includes multiple state feature intervals and motion constraint parameters associated with the state feature intervals; Obtain the target state feature interval to which the cargo state features belong; The motion constraint parameters are determined based on the target state feature interval.

7. The system according to claim 1, characterized in that, The bottom of the main body of the shelf is provided with a lifting and positioning component, which includes a positioning groove and an anti-slip stop; the anti-slip stop is disposed on the inner wall of the positioning groove; the top of the lifting mechanism is provided with a positioning protrusion; the robot controller is further configured to: Obtain location information, which includes the current location of the handling robot, the shelf location of the mobile shelf, and the picking location; The movement path is planned based on the location information, and the movement path includes a pickup path and a handling path; the pickup path is planned based on the current location and the shelf location; the handling path is planned based on the shelf location and the picking location. The motion constraint parameters are parsed from the transport task; Based on the movement path and the motion constraint parameters, a transport control command is generated, and the transport control command is sent to the moving mechanism and the lifting mechanism.

8. A method for adaptive locking feedback control of a moving shelf based on a pressure map, characterized in that, Applied to the system according to any one of claims 1-7; the method comprises: Obtain stress data; A two-dimensional pressure distribution map is generated based on the pressure data, and the self-locking column is controlled to perform locking or unlocking actions based on the two-dimensional pressure distribution map. Cargo status features are extracted from the two-dimensional pressure distribution map. The cargo status features include at least one of cargo weight, center of gravity offset, and shape risk indicators. Motion constraint parameters are generated based on the cargo state characteristics. The motion constraint parameters include at least one of the following: maximum speed coefficient, maximum acceleration coefficient, turning coefficient, and emergency stop threshold. Based on the motion constraint parameters, a transport task is generated, and the transport task is sent to the robot controller so that the robot controller drives the moving mechanism and the lifting mechanism according to the transport task.

Images