Logistics robot control method and system based on artificial intelligence

Abstract

The invention provides a logistics robot control method and system based on artificial intelligence. After the surface material type, the three-dimensional geometric contour and the surface stress distribution data of a to-be-sorted package are obtained through the logistics robot, the hardness grade and the friction coefficient interval of the package are determined, the geometric model of the package is reconstructed, the minimum external cube size is calculated, and the fragile area is recognized. The adsorption or clamping mode and the unfolding angle are adjusted according to the information, the contact point and track of the grabbing action are optimized, and meanwhile a sorting strategy is generated by considering dynamic factors such as the flow speed of a sorting line. In the execution process, action parameters and decision weights are updated through an artificial intelligence algorithm according to the actual effect, and the sorting strategy is continuously optimized. According to the technical scheme provided by the invention, the control efficiency and precision of the logistics robot can be improved.

Description

Technical Field

[0001] This application relates to the technical field of logistics robot control, and in particular, to a logistics robot control method and system based on artificial intelligence. Background Art

[0002] In modern logistics centers, with the rapid development of e-commerce, the volume of parcel sorting tasks has increased sharply, and the requirements for logistics efficiency and accuracy have also been increasing. To ensure high efficiency and low error rates, the logistics system needs to have the ability to automatically identify the attributes of parcels (such as material, shape, etc.), evaluate their physical characteristics, and dynamically adjust the processing method accordingly. This not only helps to protect the parcels from damage, but also optimizes the operation mode of logistics robots to ensure efficient operation. At the same time, considering the possible fragile areas of different parcels, the system also needs to be able to accurately plan the grasping points and paths to avoid any potential damage.

[0003] Currently, a targeted technical solution is to use a method that combines visual sensors and weight sensors to obtain basic information about parcels, including size, shape, and approximate mass distribution. Through these data, the system can roughly estimate the physical characteristics of the parcels and appropriate processing methods. This solution uses computer vision technology to quickly scan and build a two-dimensional or simple three-dimensional model of the parcel, and then provides basic operation guidance for the robot according to preset rules, such as the selection of adsorption or clamping positions.

[0004] However, this existing solution has obvious limitations. First, due to relying only on visual and weight information, the judgment of key physical properties such as the material type, hardness grade, and friction coefficient of the parcel is not accurate enough, which may lead the robot to adopt inappropriate grasping strategies. Second, this method is difficult to effectively identify the subtle features on the surface of the parcel, especially the boundary of the fragile area, increasing the risk of damage to the parcel during handling. In addition, in the face of complex and changing actual working environments, the decision-making mechanism based solely on fixed rules lacks sufficient flexibility and adaptability, and cannot achieve true dynamic adjustment and optimization. Summary of the Invention

[0005] The embodiments of this application provide a logistics robot control method and system based on artificial intelligence to solve the problems of low control efficiency and poor accuracy of logistics robots in the prior art.

[0006] In a first aspect, the embodiments of this application provide a logistics robot control method based on artificial intelligence, including: Obtaining surface material type data, three-dimensional geometric contour data, and surface stress distribution data of a parcel to be sorted through a logistics robot; Determining the material hardness grade and friction coefficient range of the parcel to be sorted based on the surface material type data; Reconstruct the geometric model of the package to be sorted based on the three-dimensional geometric profile data and calculate the size of the minimum circumscribed cube of the package. Identify the boundary of the fragile area on the surface of the package to be sorted based on the surface stress distribution data; Adjust the adsorption mode or clamping mode of the logistics robot according to the material hardness grade and friction coefficient range. Adjust the deployment angle of the logistics robot in combination with the size of the minimum circumscribed cube of the package. At the same time, constrain the distribution of contact points and the movement trajectory of the grasping action of the logistics robot according to the boundary of the fragile area, and generate a dynamic sorting strategy in combination with the current sorting line flow rate, the capacity of the target sorting area, and the load state of the robot joints; Based on the execution result of the dynamic sorting strategy, update the action parameters of the logistics robot and the priority weight of the dynamic sorting decision through an artificial intelligence algorithm to update the dynamic sorting strategy of the logistics robot.

[0007] Optionally, the step of constraining the distribution of contact points and the movement trajectory of the grasping action of the logistics robot according to the boundary of the fragile area, and generating a dynamic sorting strategy in combination with the current sorting line flow rate, the capacity of the target sorting area, and the load state of the robot joints includes: Based on the boundary of the fragile area, delimit a candidate contact point area on the surface of the geometric model, exclude the coordinate points outside the safety distance of the outer expansion of the boundary of the fragile area, and generate an initial contact point set; Screen candidate contact point pairs that meet the condition of the coincidence degree between the grasping force direction and the projection of the package center of gravity from the initial contact point set, and calculate the force closure stability score of the candidate contact point pairs; Decompose the grasping movement trajectory in the geometric model coordinate system into a horizontal translation component and a longitudinal following component, calculate the speed compensation amount of the horizontal translation component based on the sorting line flow rate, and generate the timing parameters of the longitudinal following component; According to the reciprocal of the ratio of the real-time current data of the logistics robot joints to the historical load, calculate the joint torque balance factor of the candidate contact point pairs, fuse the force closure stability score, timing parameters, and joint torque balance factor to generate a multi-objective optimization weight set, and output a dynamic sorting strategy based on the multi-objective optimization weight set.

[0008] Optionally, the step of fusing the force closure stability score, timing parameters, and joint torque balance factor to generate a multi-objective optimization weight set includes: Determine the initial weight allocation of the force closure stability score, timing parameters, and joint torque balance factor based on the material hardness grade and the sorting line flow rate. Increase the scoring weight when the material hardness grade is higher than the threshold; Convert the force closure stability score into a standardized interval, decompose the timing parameter into a synchronization error tolerance and a time window matching degree, and generate a multi-dimensional parameter vector in combination with the joint torque balance factor; Modify the initial weight assignment according to the change rate of the target sorting area capacity, and superimpose the modified initial weight assignment and the multi-dimensional parameter vector to generate a comprehensive score of the candidate solution; Truncate the joint torque balance factor according to the difference between the real-time current peak value of the joints of the logistics robot and the historical load threshold, and generate a multi-objective optimization weight set based on the comprehensive score of the candidate solution.

[0009] Optionally, the converting the force closure stability score into a standardized interval, decomposing the timing parameter into a synchronization error tolerance and a time window matching degree, and generating a multi-dimensional parameter vector in combination with the joint torque balance factor includes: Dynamically map the force closure stability score to a standardized interval based on the score distribution range of the historical grasping case library; When decomposing the timing parameter, calculate the synchronization error tolerance according to the ratio of the sorting line flow rate to the size of the circumscribed cube of the package, and generate the time window matching degree based on the theoretical movement time from the center of gravity of the geometric model to the opening of the sorting container; Dynamically adjust the gain of the joint torque balance factor according to the fluctuation variance between the real-time current data of the joints and the historical average current; Align the dimensions of the standardized interval, synchronization error tolerance, time window matching degree, and the adjusted joint torque balance factor according to the logistics scenario priority rule to generate a multi-dimensional parameter vector including mechanical stability, timing synchronization, and mechanical safety.

[0010] Optionally, the adjusting the adsorption mode or clamping mode of the logistics robot according to the material hardness grade and the friction coefficient interval includes: Set an adsorption force safety threshold based on the material hardness grade. When the material hardness grade is lower than the flexible threshold, activate the adsorption mode and dynamically adjust the distribution density of the adsorption hole array according to the friction coefficient interval; When the material hardness grade is higher than the flexible threshold, switch to the clamping mode, calculate the upper limit of the contact pressure of the clamping knuckles based on the friction coefficient interval and constrain the initial closing angle of the knuckles. In the clamping mode, adjust the force closure direction of the knuckles based on the angle between the center of gravity projection of the geometric model and the action line of the clamping force; In the adsorption mode, match the texture of the adsorption contact surface according to the lower limit value of the friction coefficient interval, and increase the activation number of auxiliary adsorption holes in the local low-friction area; When the friction coefficient interval crosses the critical value, start the hybrid mode. The adsorption hole central area remains adsorbed, and the clamping knuckles are deployed in the edge area. The initial closing angle of the knuckles is positively correlated with the diagonal length of the circumscribed cube size.

[0011] Optionally, calculating the upper limit of the contact pressure of the clamping knuckles based on the friction coefficient interval and constraining the initial closing angle of the knuckles includes: Set the maximum contact pressure threshold of the clamping knuckles according to the upper limit value of the friction coefficient interval. When the friction coefficient crosses the high and medium friction domains, adopt a segmented increasing strategy to increase the pressure threshold growth rate; Calculate the safety margin of the initial closing angle of the knuckles based on the distribution density of the area with the maximum curvature in the geometric model. The safety margin is positively correlated with the product of the number of curvature extreme points and the lower limit value of the friction coefficient; Dynamically attenuate the maximum contact pressure threshold based on the shortest distance from the boundary of the fragile area to the clamping area in the surface stress distribution data. The attenuation coefficient is inversely proportional to the square of the shortest distance; Based on the safety margin, when the upper limit of the contact pressure of the clamping knuckles reaches the threshold warning interval, trigger high-frequency micro-amplitude vibration of the contact surface, and real-time monitor the pressure gradient distribution of the clamping knuckles. If it is detected that the pressure concentration area overlaps with the area with the maximum curvature in the geometric model, increase the opening compensation angle of adjacent knuckles to disperse the stress.

[0012] Optionally, obtaining the surface material type data, three-dimensional geometric contour data, and surface stress distribution data of the package to be sorted by the logistics robot includes: Collect the surface reflectance map of the package through the multi-spectral sensor of the logistics robot, extract the multi-band spectral response differences to match the surface material type data, and enable polarization filtering to suppress interference in the highly reflective area; Adopt the fusion of structured light and stereo vision to reconstruct the three-dimensional point cloud, extract the curvature features to generate three-dimensional geometric contour data, and start multi-angle scanning in the blind area to complete the point cloud; Load the pressure gradient through the flexible tactile array, record the strain rate and stress relaxation curve to generate the surface stress distribution data, and trigger high-frequency sampling when stress mutation is detected to obtain high-frequency sampling stress data; Synchronize the multi-band spectral response differences, multi-angle scanning completed point cloud, and high-frequency sampling stress data to the three-dimensional coordinate system to generate a cross-modal correlation mapping table.

[0013] In a second aspect, an embodiment of the present application provides an artificial intelligence-based logistics robot control system, including: An acquisition module, which obtains the surface material type data, three-dimensional geometric contour data, and surface stress distribution data of the package to be sorted through the logistics robot; A determination module, which determines the material hardness grade and the friction coefficient range of the package to be sorted based on the surface material type data; An identification module, which reconstructs the geometric model of the package to be sorted based on the three-dimensional geometric contour data and calculates the size of the minimum circumscribed cube of the package, and identifies the boundary of the fragile area on the surface of the package to be sorted based on the surface stress distribution data; A generation module, which adjusts the adsorption mode or the clamping mode of the logistics robot according to the material hardness grade and the friction coefficient range, adjusts the deployment angle of the logistics robot in combination with the size of the minimum circumscribed cube of the package, and at the same time restricts the distribution of the contact points and the motion trajectory of the grasping action of the logistics robot according to the boundary of the fragile area, and generates a dynamic sorting strategy in combination with the current sorting line flow rate, the capacity of the target sorting area and the load state of the robot joints; An update module, which updates the action parameters of the logistics robot and the priority weight of the dynamic sorting decision through an artificial intelligence algorithm based on the execution result of the dynamic sorting strategy, so as to update the dynamic sorting strategy of the logistics robot.

[0014] In a third aspect, an embodiment of the present application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for controlling a logistics robot based on artificial intelligence as described in the first aspect above.

[0015] In a fourth aspect, an embodiment of the present application provides a computer storage medium, storing a computer program, which when executed by a computer, implements a method for controlling a logistics robot based on artificial intelligence as described in the first aspect.

[0016] In the embodiments of the present application, a logistics robot is used to obtain data on the surface material type, three-dimensional geometric contour, and surface stress distribution of the package to be sorted; based on the surface material type data, the material hardness grade and friction coefficient range of the package to be sorted are determined; based on the three-dimensional geometric contour data, the geometric model of the package to be sorted is reconstructed and the size of the minimum circumscribed cube of the package is calculated, and based on the surface stress distribution data, the boundary of the fragile area on the surface of the package to be sorted is identified. According to the material hardness grade and friction coefficient range, the adsorption mode or clamping mode of the logistics robot is adjusted, the unfolding angle of the logistics robot is adjusted in combination with the size of the minimum circumscribed cube of the package, and at the same time, the distribution of contact points and the movement trajectory of the grasping action of the logistics robot are restricted according to the boundary of the fragile area, and in combination with the current sorting line flow rate, the capacity of the target sorting area, and the load state of the robot joints, a dynamic sorting strategy is generated; based on the execution result of the dynamic sorting strategy, the action parameters of the logistics robot and the priority weight of the dynamic sorting decision are updated through an artificial intelligence algorithm to update the dynamic sorting strategy of the logistics robot.

[0017] The technical solution of the present application has the following beneficial effects: In the present application, a logistics robot is used to obtain data on the surface material type, three-dimensional geometric contour, and surface stress distribution of the package, ensuring a comprehensive understanding of the physical characteristics of the package. Based on the surface material type data, the hardness grade and friction coefficient range of the package are accurately determined, providing basic parameters for subsequent operations. The geometric model of the package is reconstructed using the three-dimensional geometric contour data and the size of the minimum circumscribed cube is calculated, and at the same time, the boundary of the fragile area is accurately identified based on the surface stress distribution data, which helps to protect the integrity of the package. Combining information such as the material hardness grade, friction coefficient range, minimum circumscribed cube size, and fragile area boundary, the adsorption or clamping mode, unfolding angle, and the distribution of contact points and movement trajectory of the grasping action of the robot are dynamically adjusted, improving the safety and efficiency of handling the package. Considering factors such as the current sorting line flow rate, the capacity of the target sorting area, and the load state of the robot joints, a dynamic sorting strategy is generated, and the strategy is continuously optimized through the feedback of the execution result.

[0018] Further, the method first delimits a candidate area of contact points for safe grasping on the geometric model based on the boundary of the fragile area, screens out the contact points that meet the condition of the coincidence degree of the force direction and the projection of the center of gravity, and calculates their stability scores; then decomposes the grasping movement trajectory into horizontal and vertical components, adjusts the horizontal speed compensation according to the sorting line flow rate, and plans the longitudinal following time sequence parameters; finally, fuses the joint torque balance factor, the force closure stability score, and the time sequence parameters to generate a multi-objective optimization weight set to output the final dynamic sorting strategy. This process effectively improves the safety, stability, and adaptability of the grasping action, significantly reduces the risk of damage to the package during handling, and at the same time improves the working efficiency and flexibility of the logistics robot.

[0019] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0021] Figure 1 A flowchart showing a method for controlling a logistics robot based on artificial intelligence provided by the present application is shown; Figure 2 A schematic structural diagram showing a control system for a logistics robot based on artificial intelligence provided by the present application is shown; Figure 3 A schematic structural diagram showing a computing device provided by the present application is shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] In order to enable those skilled in the art to better understand the solutions of the present application, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application.

[0023] In some processes described in the specification and claims of the present application and the above-mentioned drawings, a plurality of operations that appear in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first" and "second" in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit that "first" and "second" are of different types.

[0024] This solution obtains data on the surface material type, three-dimensional geometric contour, and surface stress distribution of the package to be sorted through a logistics robot, aiming to collect the surface reflectance map of the package using a multispectral sensor and reconstruct the three-dimensional model of the package by combining structured light and stereovision technology. At the same time, a flexible tactile array is used to record the pressure response characteristics of the package surface, so as to realize a comprehensive perception of the physical characteristics of the package. This method emphasizes the cross-modal fusion of data to improve the accuracy and integrity of information collection.

[0025] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0026] Figure 1 The flowchart of a logistics robot control method based on artificial intelligence provided by an embodiment of the present application is as Figure 1 shown, and the method includes: 101. Obtain the surface material type data, three-dimensional geometric contour data, and surface stress distribution data of the package to be sorted through the logistics robot; In this step, the surface material type data refers to the data for determining its physical properties by analyzing the external material components of the package, including paper, plastic, or fabric, etc. This information is used to select appropriate processing methods later.

[0027] The three-dimensional geometric contour data refers to all the dimensional information and its changes describing the shape of the package, which is used to reconstruct the digital model of the package to help identify the shape and size of the package.

[0028] The surface stress distribution data refers to the pressure values at each point on the surface of the package, which helps to identify the vulnerable areas that may exist on the package and ensure that excessive force is not applied to these fragile parts during handling.

[0029] In the embodiments of the present application, first, spectral analysis technology is used to identify the surface material type of the package. The material type is determined by irradiating with light of a specific wavelength and analyzing the reflection spectrum. Then, a laser scanner is used to capture the external shape characteristics of the package to generate a high-precision digital model. Finally, an embedded pressure sensing device is used to measure the pressure values at different positions on the surface of the package to obtain the surface stress distribution map. By integrating the above three types of data, the logistics robot can comprehensively understand the specific characteristics of each package, so as to formulate corresponding processing strategies.

[0030] In a busy logistics center, an intelligent logistics robot first performs spectral analysis on a cardboard box and quickly determines that its material is corrugated cardboard. Subsequently, the robot uses a laser scanner to construct a detailed model of the cardboard box, including parameters such as its length, width, and height. At the same time, the pressure distribution on the surface of the cardboard box is recorded through the built-in pressure sensing patches, especially paying attention to those areas where higher stress may exist, to prepare for the next operation.

[0031] 102. Determine the material hardness grade and friction coefficient range of the package to be sorted based on the surface material type data; In this step, the material hardness grade is an indicator for measuring the ability of the package to resist deformation and is determined based on the surface material type data. This grade helps to select a suitable adsorption or clamping mode to prevent the package from being damaged due to excessive extrusion.

[0032] The friction coefficient range refers to the degree reflecting the possible slippage between the package and the clamping device. It is calculated based on the surface material type data and is used to guide the selection of the correct operation mode to ensure that the package does not slip due to insufficient friction.

[0033] In the embodiment of the present application, using a pre-trained machine learning model, the surface material type data obtained in step 101 is used as input, and by comparing with the standard samples in the database, the material hardness grade and the friction coefficient range of the package are calculated. This process involves using classification algorithms such as support vector machine or decision tree algorithms, which are achieved by training on a large dataset of known materials. The final result is based on the closest match and provides an estimated value of the hardness and friction characteristics to guide the selection of subsequent operations.

[0034] According to the previous analysis, the logistics robot queries the internal database and analyzes through the machine learning model to obtain that the corrugated cardboard has a moderate hardness grade and a relatively high friction coefficient range, indicating that it is neither easily damaged by external pressure nor easily slips from the gripper, providing an important reference for the next operation.

[0035] 103. Reconstruct the geometric model of the package to be sorted based on the three-dimensional geometric contour data and calculate the size of the minimum circumscribed cube of the package, and identify the boundary of the fragile area on the surface of the package to be sorted based on the surface stress distribution data; In this step, the geometric model is a digital representation of the package reconstructed using the three-dimensional geometric contour data, which is used to accurately describe the shape and size of the package.

[0036] The size of the minimum circumscribed cube of the package refers to the size of the space required to completely contain the package, which is calculated based on the geometric model and helps to optimize the utilization rate of storage space.

[0037] The boundary of the fragile area is the area on the surface of the package with relatively high stress identified based on the surface stress distribution data. These areas are considered to be relatively fragile parts and need special attention to protection during handling.

[0038] In the embodiment of the present application, the geometric model of the package is reconstructed by computer-aided design software according to the three-dimensional geometric contour data, and then an algorithm is applied to calculate the size of the minimum circumscribed cube of the package that can completely contain the model. For the identification of the boundary of the fragile area, a threshold is set for the surface stress distribution data for analysis, and the areas with abnormally high stress are marked as potential vulnerable parts. This process combines image processing technology and mathematical modeling methods to ensure accuracy and reliability.

[0039] Continuing from the previous step, the logistics robot not only determines the optimal storage size of the carton but also discovers that there is a relatively high stress level in the part with a printed pattern on one side, which may be a relatively fragile area. Therefore, when planning the handling path, this area is deliberately avoided to prevent accidental damage and ensure that the package reaches the destination safely.

[0040] 104. Adjust the adsorption mode or clamping mode of the logistics robot according to the material hardness grade and friction coefficient range, adjust the deployment angle of the logistics robot in combination with the minimum circumscribed cube size of the package, and at the same time constrain the contact point distribution and movement trajectory of the grasping action of the logistics robot according to the fragile area boundary, and generate a dynamic sorting strategy in combination with the current sorting line flow rate, the capacity of the target sorting area, and the load state of the robot joints; In this step, the adsorption mode or clamping mode refers to the way the logistics robot grasps the package, including using a vacuum suction cup for adsorption or a mechanical claw for clamping. Select the most suitable mode according to the material hardness grade and friction coefficient range of the package to ensure that the package is neither damaged nor slips during handling.

[0041] The sorting line flow rate is the speed at which packages flow on the sorting line, which affects the operation rhythm and efficiency of the logistics robot.

[0042] The capacity of the target sorting area refers to the maximum number or volume of packages that the target sorting area can accommodate, which is crucial for planning the placement position of the logistics robot.

[0043] The load state of the robot joints refers to the description of the current load borne by each joint of the robot, which helps to optimize the grasping and moving strategies and prevent equipment damage caused by overloading.

[0044] In the embodiment of the present application, first, select the most suitable adsorption or clamping method according to the material hardness grade and friction coefficient range, such as using a vacuum suction cup or a mechanical claw. Then, calculate the appropriate deployment angle of the robotic arm using the minimum circumscribed cube size of the package to ensure that the package can be firmly grasped without causing damage. For the fragile area, determine the optimal grasping path through a path planning algorithm to avoid directly acting on these areas. In addition, comprehensively consider factors such as the sorting line flow rate, the capacity of the target sorting area, and the load state of the robot joints, and use an optimization algorithm to formulate the most effective sorting strategy.

[0045] Based on the previous data analysis, the logistics robot selected a clamping method suitable for the characteristics of the corrugated cardboard and adjusted the angle of the robotic arm for safe grasping. When planning the handling path, special attention was paid to avoiding the identified fragile areas to ensure that the packages would not be damaged due to improper operation. At the same time, considering the speed of the current sorting line and the space limitations of the target area, the robot also optimized the action sequence to improve the overall efficiency.

[0046] 105. Based on the execution results of the dynamic sorting strategy, update the action parameters of the logistics robot and the priority weights of the dynamic sorting decision through an artificial intelligence algorithm to update the dynamic sorting strategy of the logistics robot.

[0047] In this step, the execution results of the dynamic sorting strategy refer to the actual effects after the logistics robot executes the sorting task according to the predetermined strategy, such as the completion time, the package damage rate, etc., which are used to evaluate the effectiveness of the strategy.

[0048] The artificial intelligence algorithm is a series of algorithms used to analyze the execution results of the dynamic sorting strategy and accordingly adjust the action parameters of the logistics robot and the decision priority weights, such as the reinforcement learning algorithm.

[0049] The action parameters refer to specific settings including the adsorption force magnitude, clamping force, moving speed, etc., which directly affect the performance of the logistics robot when executing tasks.

[0050] The decision priority weight is the standard weight allocation for making the optimal choice among various possible operations. For example, while ensuring speed, maximizing safety, and by adjusting these weights, the sorting strategy can be optimized.

[0051] In the embodiment of the present application, the reinforcement learning algorithm is used to evaluate the effect of each operation, collect feedback information and accordingly adjust the strategy parameters of the next action. Specifically, the result of each task execution is used as input, evaluated through a series of predefined evaluation criteria (such as the time to successfully complete the task, the potential damage to the package, etc.), and then the corresponding parameter values are adjusted through the algorithm. This process is a closed-loop control mechanism, and through continuous iterative learning, the performance of the system is gradually improved.

[0052] After multiple actual operations, the logistics robot adjusted the adsorption force and clamping force settings based on the collected data and feedback information, and reallocated the decision priorities. For example, when handling similar types of packages, it learned to be more cautious with the printed parts, reducing the error rate. Over time, this optimization ability made the entire logistics sorting system more efficient and reliable.

[0053] In summary, the system integrating advanced sensing technology and intelligent algorithms in steps 101 to 105 realizes highly personalized processing of the parcels to be sorted, greatly improving the efficiency and safety of logistics sorting. By accurately identifying the characteristics of parcels and adjusting the operation mode accordingly, combined with real-time data analysis and self-optimization mechanisms, the logistics robot can flexibly handle various challenges in a complex and changing working environment. This not only increases the success rate of a single task but also promotes the continuous improvement and development of the entire logistics process.

[0054] To solve the problem of optimizing the grasping strategy according to the vulnerable area of the parcel, this solution determines the material hardness grade and friction coefficient range of the parcel to be sorted based on the surface material type data. It aims to identify the specific material properties of the parcel by analyzing the surface reflectivity spectrum of the parcel and evaluate its hardness and friction characteristics accordingly. This process utilizes the correlation between the material and the optical response to provide the necessary physical parameter support for subsequent operations. This step is a key link in realizing automated processing, ensuring the effective adjustment of subsequent grasping and handling strategies. In some embodiments, in step 104, constraining the contact point distribution and movement trajectory of the grasping action of the logistics robot according to the fragile area boundary, and combining the current sorting line flow rate, the capacity of the target sorting area, and the robot joint load status, to generate a dynamic sorting strategy, including: 201. Based on the fragile area boundary, delimit a candidate contact point area on the surface of the geometric model, exclude the coordinate points outside the safety distance of the outer expansion of the fragile area boundary, and generate an initial contact point set; In step 201, the fragile area boundary refers to the area with higher stress on the surface of the parcel. An additional safety interval area is set to avoid applying pressure to the vulnerable area, ensuring that the grasping action does not directly act on the fragile part. The safety distance is an additional range set to protect the fragile area from direct contact. The initial contact point set is a series of position points that can be used for grasping operations screened according to these conditions. The candidate contact point area is the possible position for the robot to grasp except for these fragile areas.

[0055] In the embodiments of the present application, first, identify the fragile area on the parcel by analyzing the stress distribution in the three-dimensional geometric contour data, and set an appropriate safety distance around it as a buffer zone. Use image processing technology to mark the coordinate points of all non-vulnerable areas, and then mark all possible safe grasping points on the remaining surface to form an initial contact point set. This process combines computer vision algorithms and mathematical modeling to ensure accuracy.

[0056] 202. Screen candidate contact point pairs that meet the condition of the coincidence degree between the grasping force direction and the projection of the parcel's center of gravity from the initial contact point set, and calculate the force closure stability score of the candidate contact point pairs; In step 202, the condition of the coincidence degree between the grasping force direction and the projection of the center of gravity of the package means that when selecting the grasping point, it is desired that the direction of the grasping force passes through the center of gravity of the package as much as possible to reduce the rotational effect. Considering measuring whether the direction of the grasping force is close to the projection of the center of gravity of the package on the horizontal plane, a high coincidence degree means better stability. The force-closure stability score evaluates the possibility of the package remaining stable during the grasping process, and a high score means higher stability.

[0057] In the embodiments of the present application, a physical simulation software is used to calculate the distance and the included angle between each pair of contact points and the projection of the center of gravity of the package, and select those pairs of contact points whose force directions coincide with the center of gravity projection as much as possible. For each selected pair of contact points, the force-closure stability score is calculated based on their geometric relationship and mechanical principles. Finally, the most suitable pair of contact points is selected according to the scoring results to ensure the best grasping effect.

[0058] 203. Decompose the grasping motion trajectory in the geometric model coordinate system into a lateral translation component and a longitudinal following component, calculate the velocity compensation amount of the lateral translation component based on the flow rate of the sorting line, and generate the timing parameters of the longitudinal following component; In step 203, the lateral translation component and the longitudinal following component respectively refer to the two main directions of the movement of the robot arm during the grasping process. The former is to match the speed of the sorting line and adjust the speed compensation amount of the robot. It is the change amount required to adjust the lateral movement speed of the robot according to the flow rate of the sorting line to ensure that the package can be accurately placed. The latter describes the time series parameters of the robot following the target path in the vertical direction to ensure a smooth transition.

[0059] In the embodiments of the present application, a path planning algorithm is used to decompose the grasping motion trajectory into actions in two dimensions, namely, the lateral and longitudinal dimensions. According to the current speed of the sorting line, the lateral speed compensation amount that needs to be increased or decreased is calculated to synchronize the movement speed of the robot with the sorting line. At the same time, the time series parameters of the longitudinal movement are formulated to ensure that the package can reach the target position smoothly and quickly.

[0060] 204. Calculate the joint torque balance factor of the candidate contact point pair according to the reciprocal of the ratio of the real-time current data of the joints of the logistics robot to the historical load, fuse the force-closure stability score, the timing parameters and the joint torque balance factor to generate a multi-objective optimization weight set, and output a dynamic sorting strategy based on the multi-objective optimization weight set.

[0061] In step 204, the joint torque balance factor reflects the load balance state of each joint of the robot during task execution and is calculated by real-time monitoring of joint current data. The force closure stability score is a quantitative criterion for evaluating whether each candidate grasping point pair can stably hold an object without dropping. The multi-objective optimization weight set is the optimal parameter combination determined by comprehensively considering multiple factors such as force closure stability, motion trajectory, and timing parameters, and is used to guide the robot to complete the optimal operation strategy.

[0062] In the embodiment of the present application, the joint torque balance factor is calculated by real-time monitoring of the robot joint current data and combining with the historical load ratio. This factor is combined with the previously obtained force closure stability score, timing parameters, etc. to form a multi-objective optimization weight set. This process involves data analysis, application of optimization algorithms, etc. Based on this weight set, a dynamic sorting strategy most suitable for the current environment is formulated.

[0063] The following is a specific example: In a busy logistics center, an intelligent logistics robot first processes a cardboard box. It identifies that there is a high stress level in the part with a printed pattern on one side of the cardboard box, marks it as a fragile area and sets a safety distance around it. Then, the robot finds multiple potential safe grasping points on the remaining surface and selects several points closest to the projection of the package's center of gravity. After that, the robot calculates the force closure stability scores of these points and adjusts the speed compensation amount of the grasping and placing actions according to the flow rate of the sorting line. Finally, considering the load conditions of the joints, the robot optimizes the grasping force distribution to ensure the safety and efficiency of the operation.

[0064] In summary, steps 201 to 204 significantly improve the accuracy and efficiency in the logistics sorting process, while reducing the risk of package damage. By accurately identifying the package characteristics and adjusting the operation mode accordingly, combined with real-time data analysis and self-optimization mechanism, the logistics robot can flexibly handle various challenges in a complex working environment, thereby enhancing the performance and reliability of the entire logistics process.

[0065] To solve the sorting efficiency and safety problems of intelligent logistics robots in the logistics center when handling fragile items, the solution reconstructs the geometric model of the package to be sorted based on three-dimensional geometric contour data and calculates the size of the minimum circumscribed cube of the package. At the same time, it identifies the boundary of the fragile area according to the surface stress distribution data. The main goal is to accurately reconstruct the package's shape and conduct a detailed safety assessment on this basis. Through high-precision three-dimensional modeling technology, not only can the size and shape of the package be obtained, but also the possible vulnerable parts can be accurately located, laying a foundation for formulating a safe handling plan. In some embodiments, generating the multi-objective optimization weight set by fusing the force closure stability score, timing parameters, and joint torque balance factor in step 204 includes: 301. Determine the initial weight distribution of the force closure stability score, timing parameters, and joint torque balance factor based on the material hardness level and the sorting line flow rate. Increase the scoring weight when the material hardness level is higher than the threshold. In step 301, the material hardness level represents the ability level of the packaging material to resist deformation or breakage, which is obtained through physical tests or packaging information. The initial weight distribution of the force closure stability score is to adjust the emphasis on grasping stability according to the material hardness. For harder materials, the requirement for stability can be appropriately reduced. The initial weight distribution of the timing parameters is a process of determining the action time arrangement based on the sorting line flow rate. A fast assembly line requires more precise time control. The initial weight distribution of the joint torque balance factor needs to consider the load balance of the robot joints to avoid mechanical damage caused by uneven loads.

[0066] In the embodiment of the present application, first analyze the material hardness of the cardboard box and set the initial weights of various indicators in combination with the real-time flow rate of the sorting line. If it is detected that the material hardness of the cardboard box is higher than the preset threshold, the weight of the force closure stability score is correspondingly increased, and the requirement for the tolerance of the synchronization error is reduced. This step uses the sensor data to compare and analyze with the historical database to obtain the corresponding weight value, thereby providing a basic reference for the subsequent steps.

[0067] 302. Convert the force closure stability score into a standardized interval, decompose the timing parameters into the synchronization error tolerance and the time window matching degree, and generate a multi-dimensional parameter vector in combination with the joint torque balance factor. In step 302, the standardized interval is to convert the force closure stability score to a common comparison scale, which is convenient for comparison between different schemes. The synchronization error tolerance refers to the range of time deviation allowed during the execution process to ensure that even a slight time difference will not affect the completion of the overall task. The time window matching degree is a measure of whether the actual operation conforms to the predetermined time plan to ensure the close connection of each link. The multi-dimensional parameter vector contains all the key parameters after standardization processing and is used to comprehensively evaluate the overall performance of the candidate scheme.

[0068] In the embodiment of the present application, a mathematical transformation method is used to map the force closure stability score to the standardized interval, and the timing parameters are decomposed into two dimensions: the synchronization error tolerance and the time window matching degree. Then, a multi-dimensional parameter vector is formed in combination with the joint torque balance factor. This process involves statistical methods and data analysis techniques, and finally generates a quantitative index reflecting the advantages and disadvantages of different schemes, providing a basis for the next adjustment.

[0069] 303. Modify the initial weight distribution according to the change rate of the target sorting area capacity, and superimpose the modified initial weight distribution and the multi-dimensional parameter vector to generate the comprehensive score of the candidate scheme. In step 303, the change rate of the target sorting area capacity describes the change speed of the space utilization rate of the sorting area and affects the policy adjustment frequency. The adjusted initial weight allocation dynamically adjusts the importance ratio of each index according to the status of the sorting area. The comprehensive score of the candidate solutions refers to obtaining the overall evaluation score of each candidate solution by superimposing the adjusted weights and the multi-dimensional parameter vector, which guides the selection of the optimal path.

[0070] In the embodiment of the present application, the change situation of the target sorting area capacity is monitored in real time, and the previously set weight allocation is adjusted accordingly. The adjusted weights are combined with the previously constructed multi-dimensional parameter vector to calculate the comprehensive score of each candidate solution. Here, an adaptive algorithm is used to respond to environmental changes to ensure that the policy is always in the optimal state to maximize efficiency and minimize risks.

[0071] 304. Truncate the joint torque balance factor according to the difference between the real-time current peak value of the joints of the logistics robot and the historical load threshold, and generate a multi-objective optimization weight set based on the comprehensive score of the candidate solutions.

[0072] In step 304, the difference between the real-time current peak value of the joints and the historical load threshold is used to judge whether the current load is close to the limit to protect the machine from overload damage. Truncating the joint torque balance factor means reducing the influence of this factor when approaching the limit to prevent over-reliance on operations that may cause mechanical failures. The multi-objective optimization weight set is a set of optimization strategies that generate the final guidance for the robot's behavior based on all the above information to ensure efficient and safe operations.

[0073] In the embodiment of the present application, the real-time current peak value of the robot joints is monitored and compared with the historical load threshold. If it is found that overload is about to occur, the influence of the joint torque balance factor is correspondingly reduced. Based on this adjustment and the comprehensive score of the candidate solutions, a new set of multi-objective optimization weight sets is generated. This process uses a prediction model and a feedback mechanism to ensure that the robot can operate efficiently within a safe range and achieve the safety and accuracy of automated operations.

[0074] The following is a specific example: In a busy logistics center scenario, an intelligent logistics robot first scans the cartons entering the sorting line in detail, identifies their material hardness and avoids fragile areas. Then, based on the carton material and the sorting line flow rate, the initial weight allocation is determined, and a multi-dimensional parameter vector is formed through a series of mathematical transformations of key parameters. At the same time, the system dynamically adjusts these weights according to the status of the sorting area and calculates the comprehensive score of each candidate solution. Finally, considering the safe load of the robot joints, the overall strategy is optimized to ensure that each package can be processed quickly and safely.

[0075] In summary, steps 301 to 304 significantly enhance the ability of intelligent logistics robots in the logistics center to handle complex tasks, not only improving the sorting efficiency, but also effectively reducing the risk of item damage, and achieving the safety and accuracy of automated operations.

[0076] To solve the problem of optimizing the grasping strategy of intelligent logistics robots in a dynamic environment, this solution adjusts the adsorption mode or clamping mode of the logistics robot according to the material hardness grade and friction coefficient range, adjusts the unfolding angle in combination with the minimum circumscribed cube size of the package, and constrains the contact point distribution and movement trajectory of the grasping action according to the boundary of the fragile area. The key lies in dynamically optimizing the operation mode of the robot based on the physical characteristics of the package. By intelligently adjusting the adsorption or clamping force and position, it ensures that the package is stable and safe during handling, minimizing the risk of damage to the greatest extent. In some embodiments, in step 302, converting the force closure stability score into a standardized interval, decomposing the timing parameters into a synchronization error tolerance and a time window matching degree, and generating a multi-dimensional parameter vector in combination with the joint torque balance factor includes: 401. Dynamically map the force closure stability score to a standardized interval based on the score distribution range of the historical grasping case library; In step 401, the score distribution range of the historical grasping case library is based on the set of force closure stability score data in previous operations, and is used to determine the boundaries of the standardized interval. The force closure stability score is a quantitative criterion for evaluating whether each candidate grasping point pair can stably hold an object without dropping. The standardized interval is to convert the force closure stability scores under different conditions to a common comparison scale, facilitating comparative analysis between different solutions.

[0077] In the embodiments of this application, first, collect and analyze the force closure stability score data in the historical grasping case library to determine its distribution range. Then, use linear transformation or other mathematical methods (such as min-max normalization) to dynamically adjust the current score to a preset standardized interval according to this distribution range. This step ensures that even scores obtained under different conditions can be compared on the same basis, thus providing a reliable foundation for subsequent steps. The final result is a list of force closure stability scores after standardization.

[0078] 402. When decomposing the timing parameters, calculate the synchronization error tolerance according to the ratio of the sorting line flow rate to the circumscribed cube size of the package, and generate the time window matching degree based on the theoretical movement time from the center of gravity of the geometric model to the opening of the sorting container; In step 402, the synchronization error tolerance allows for a range of time deviations that exist during the execution process to accommodate minor time differences in actual operations without affecting task completion. The time window matching degree measures whether the actual operations conform to the predetermined time plan, ensuring the tight connection of each link and improving the coordination of the overall process. The size of the circumscribed cube of the package refers to the maximum outer dimension of the package, which is used to calculate the required operating space and time, and helps to determine the optimal grasping timing.

[0079] In the embodiment of the present application, the synchronization error tolerance is determined by calculating the ratio of the sorting line flow rate to the size of the circumscribed cube of the package, which reflects the system's adaptability to speed changes. At the same time, the time window matching degree is generated based on the theoretical movement time from the centroid of the geometric model to the opening of the sorting container, involving the application of path planning algorithms. This process ensures the precise adjustment of the robot's action rhythm, thereby achieving an efficient and accurate operation process. The specific values of the synchronization error tolerance and the time window matching degree are finally obtained, providing a basis for the optimization strategy.

[0080] 403. Dynamically adjust the joint torque balance factor according to the variance of the fluctuations between the real-time joint current data and the historical average current; In step 403, the variance of the fluctuations between the real-time joint current data and the historical average current reflects the change in the load of the robot joints and evaluates the health status of the mechanical system. The joint torque balance factor reflects whether the loads borne by each joint of the robot are balanced during task execution and is calculated by monitoring the joint current data in real time. The dynamic gain adjustment can timely adjust the joint torque balance factor according to the joint load condition, optimizing the safety and efficiency of mechanical operations.

[0081] In the embodiment of the present application, the real-time current of the robot joints is continuously monitored, compared with the historical average current, and the variance of the fluctuations is calculated. Based on this data, a proportional-integral-derivative control or other adaptive control algorithms are used to dynamically adjust the joint torque balance factor. This process can not only effectively prevent the risk of overload but also improve the flexibility and response speed of operations. The final result is the adjusted joint torque balance factor, ensuring the stable operation of the robot while improving the operation efficiency.

[0082] 404. Align the standardized interval, synchronization error tolerance, time window matching degree, and the adjusted joint torque balance factor according to the logistics scenario priority rules to generate a multi-dimensional parameter vector including mechanical stability, timing synchronization, and mechanical safety.

[0083] In step 404, the logistics scenario priority rule is used to determine the importance ranking of each parameter according to the requirements of specific application scenarios. For example, when handling fragile items, the force closure stability score may be more important than the synchronization error tolerance; while on a high-speed sorting line, the time window matching degree may take a higher priority. The multi-dimensional parameter vector is a vector containing quantitative indicators such as mechanical stability, timing synchronization, and mechanical safety, and is used to comprehensively evaluate the overall performance of candidate solutions.

[0084] In the embodiments of the present application, according to the logistics scenario priority rule, the standardization interval, synchronization error tolerance, time window matching degree, and the adjusted joint torque balance factor are dimensionally aligned. This step involves complex weight allocation strategies and data analysis techniques, such as the analytic hierarchy process or entropy weight method, to determine the weight of each parameter. Finally, a multi-dimensional parameter vector is formed, which comprehensively reflects the status of each key indicator and provides a scientific basis for formulating the optimal strategy. The output is a multi-dimensional parameter vector containing all necessary information, guiding the robot to efficiently and safely complete the sorting task.

[0085] The following is a specific example: In a busy logistics center scenario, when an intelligent logistics robot starts working, it first scans the cardboard boxes entering the sorting line in detail, identifies their material hardness, and avoids fragile areas. Then, based on the data in the historical grasping case library, the force closure stability score is mapped to the standardization interval. At the same time, the synchronization error tolerance is calculated according to the flow rate of the sorting line and the size of the package, and the time window matching degree is generated based on the distance from the center of gravity of the geometric model to the opening of the sorting container. In addition, the real-time current of the robot joints is monitored, and the joint torque balance factor is adjusted according to historical data. Finally, a multi-dimensional parameter vector is constructed according to the priority rule by combining all the information, guiding the robot to efficiently and safely complete the sorting task.

[0086] In summary, steps 401 to 404 significantly improve the ability of intelligent logistics robots in the logistics center to handle complex tasks, not only improving the sorting efficiency, but also effectively reducing the risk of item damage, achieving the safety and accuracy of automated operations. Through dynamic adjustment of strategies, the robot can flexibly respond to various situations, thus greatly enhancing the flexibility and reliability of the entire logistics system.

[0087] To solve the problem of intelligent logistics robots selecting appropriate grasping modes when handling packages with different materials and friction coefficients, when generating a dynamic sorting strategy, factors such as the current flow rate of the sorting line, the capacity of the target sorting area, and the load status of the robot joints are comprehensively considered, aiming to construct a highly adaptable and efficient sorting system. By real-time monitoring and analyzing changes in the logistics environment, the sorting strategy is adjusted in a timely manner to ensure the smooth operation of the entire system. This flexibility enables logistics robots to maintain high efficiency in complex and changing working environments. In some embodiments, adjusting the adsorption mode or clamping mode of the logistics robot according to the material hardness level and friction coefficient range in step 104 includes: 501. Set an adsorption force safety threshold based on the material hardness level. When the material hardness level is lower than the flexibility threshold, activate the adsorption mode and dynamically adjust the distribution density of the adsorption hole array according to the friction coefficient range; In step 501, the material hardness level represents the ability level of the package material to resist deformation or damage, which is obtained through physical tests or packaging information. The adsorption force safety threshold is the maximum allowable adsorption force set based on the material hardness to avoid damaging the package. The flexibility threshold is a preset hardness value. When the value is lower than this value, the adsorption mode is adopted, and when it is higher than this value, the clamping mode is switched. The distribution density of the adsorption hole array refers to the arrangement density of the suction holes on the adsorption device, which is dynamically adjusted according to the friction coefficient.

[0088] In the embodiments of the present application, first, the material hardness level of the cardboard box is measured by a sensor, and the adsorption force safety threshold is set accordingly. If the detected material hardness is lower than the flexibility threshold, the adsorption mode is activated. Next, image recognition technology is used to analyze the friction coefficient range on the surface of the package, and an algorithm (such as a linear regression model) is applied to dynamically calculate and adjust the distribution density of the adsorption hole array. For example, the number of adsorption holes is reduced in high-friction areas and increased in low-friction areas to adapt to different friction characteristics. The final result is an optimized adsorption hole layout plan for the current package characteristics.

[0089] 502. When the material hardness level is higher than the flexibility threshold, switch to the clamping mode, calculate the upper limit of the contact pressure of the clamping knuckles based on the friction coefficient range and constrain the initial closing angle of the knuckles. In the clamping mode, adjust the closing direction of the knuckle force based on the angle between the geometric model centroid projection and the clamping force action line; In step 502, the clamping mode is a grasping method enabled when the material hardness exceeds the flexibility threshold, suitable for relatively hard or regularly shaped objects. This mode uses mechanical knuckles for firm grasping. The upper limit of the contact pressure is the maximum allowable clamping pressure determined based on the friction coefficient, preventing damage caused by excessive squeezing. A higher friction coefficient allows a greater clamping pressure. The initial closing angle of the knuckle refers to the angle setting when the gripper starts to act, associated with the size of the circumscribed cube, ensuring the optimal grasping position. The force closing direction refers to the direction adjusted according to the angle between the projection of the center of gravity of the geometric model and the line of action of the clamping force, ensuring stable clamping.

[0090] In the embodiment of the present application, when the material hardness exceeds the flexibility threshold, the system switches to the clamping mode. Using the friction coefficient interval data, the upper limit of the contact pressure of the clamping knuckle is calculated through mechanical analysis, and the initial closing angle of the knuckle is set to be proportional to the diagonal length of the circumscribed cube size of the package. At the same time, the force closing direction of the knuckle is adjusted according to the angle between the projection of the center of gravity of the geometric model and the line of action of the clamping force to ensure the optimal grasping stability. This step involves the application of complex mechanical simulation and path planning algorithms. The final output is an optimized clamping strategy configuration to ensure efficient and safe grasping operations.

[0091] 503. In the adsorption mode, match the adsorption contact surface texture according to the lower limit value of the friction coefficient interval, and increase the activation quantity of auxiliary adsorption holes in the local low-friction area; In step 503, the lower limit value of the friction coefficient interval refers to the lowest friction coefficient value in the friction coefficient distribution on the surface of the package. This value is used to determine the specific area where the adsorption force needs to be enhanced to ensure firm adsorption even under low-friction conditions. The activation quantity of the auxiliary adsorption holes refers to the number of additional adsorption holes added in the local low-friction area. These holes can be automatically opened or closed under specific conditions to improve the overall adsorption effect. The adsorption contact surface texture refers to the microscopic structural characteristics of the package surface, and different textures may affect the adsorption effect. By analyzing these texture characteristics, the design and configuration of the adsorption device can be optimized.

[0092] In the embodiment of the present application, in the adsorption mode, according to the lower limit value of the friction coefficient interval, the texture characteristics of the adsorption contact surface are matched through image recognition technology. For the detected local low-friction area, the number of auxiliary adsorption holes is increased to improve the adsorption force. This step involves the application of computer vision technology and data analysis methods, such as convolutional neural networks for texture recognition. Finally, a comprehensive adsorption layout plan including main adsorption holes and auxiliary adsorption holes is formed to ensure reliable adsorption even on complex surfaces.

[0093] 504. When the friction coefficient crosses the critical value within the friction coefficient range, start the hybrid mode. Keep the adsorption in the central area of the adsorption holes, deploy the clamping knuckles in the edge area, and the initial closing angle of the knuckles is positively correlated with the diagonal length of the circumscribed cube size.

[0094] In step 504, the hybrid mode is a mode started when the friction coefficient crosses the critical value. It combines the advantages of both adsorption and clamping methods to provide a more flexible grasping solution. Keeping the adsorption in the central area of the adsorption holes allows the adsorption holes to continue to work in the central area under the hybrid mode, providing basic adsorption force support. Deploying the clamping knuckles in the edge area strengthens the edge area of the package by using the clamping knuckles to ensure the stability and safety of grasping. The initial closing angle of the knuckles being positively correlated with the diagonal length of the circumscribed cube size means that the initial angle of the knuckles is adjusted according to the size of the package to adapt to packages of different sizes.

[0095] In the embodiment of the present application, when the friction coefficient crosses the critical value, the system automatically enters the hybrid mode. At this time, the central area of the adsorption holes continues to maintain the adsorption function, while the clamping knuckles are deployed in the edge area. The initial closing angle of the knuckles is adjusted according to the diagonal length of the circumscribed cube size of the package to ensure the stability and safety of grasping. This process requires precise sensor data and complex control algorithms for support, such as the controller for real-time adjustment of the clamping force. Finally, a grasping solution that takes into account the advantages of both adsorption and clamping is generated to ensure an efficient and safe operation process.

[0096] The following is a specific example: In a busy logistics center, when an intelligent logistics robot starts working, it first scans the cardboard boxes entering the sorting line in detail to identify their material hardness grade and friction coefficient range. If the material hardness is lower than the flexible threshold, activate the adsorption mode and dynamically adjust the distribution density of the adsorption hole array according to the friction coefficient range. If the material hardness exceeds the flexible threshold, switch to the clamping mode, calculate the upper limit of the contact pressure of the clamping knuckles and set the initial closing angle of the knuckles. For packages with complex friction coefficient distributions, the robot may enter the hybrid mode, using adsorption in the central area and clamping knuckles in the edge area. All these adjustments are achieved based on real-time monitored data and advanced algorithms, ensuring an efficient and safe operation process.

[0097] In summary, steps 501 to 504 significantly enhance the ability of intelligent logistics robots in the logistics center to handle different types of packages. It not only improves the sorting efficiency but also effectively reduces the risk of item damage, achieving the safety and accuracy of automated operations. By dynamically adjusting the grasping mode, the robot can flexibly respond to various situations, greatly enhancing the flexibility and reliability of the entire logistics system. Especially through the fine analysis of material hardness and friction coefficient, it ensures that each operation can reach the optimal state, thus guaranteeing the work efficiency while maximizing the protection of cargo safety. This method ensures the optimization of robot operation through meticulous data analysis and intelligent adjustment mechanisms.

[0098] To further solve the problem of selecting the appropriate clamping mode for intelligent logistics robots when handling packages with different friction coefficients and geometric shapes, this solution updates the action parameters of the logistics robot and the priority weights of dynamic sorting decisions through artificial intelligence algorithms. The aim is to continuously optimize the behavior pattern of the robot to make it more adaptable to the actual work requirements. Based on the learning and analysis of historical data, the system can self-adjust to improve the accuracy and rationality of decisions. This method greatly enhances the adaptive ability and work efficiency of logistics robots. In some embodiments, step 502, calculating the upper limit of the contact pressure of the clamping knuckle based on the friction coefficient interval and restricting the initial closing angle of the knuckle, includes: 601. Set the maximum contact pressure threshold of the clamping knuckle according to the upper limit value of the friction coefficient interval. When the friction coefficient crosses the high and medium friction domains, adopt a segmented increasing strategy to increase the pressure threshold growth rate; In step 601, the upper limit value of the friction coefficient interval represents the maximum value of the surface friction coefficient of the package, which is used to set the maximum contact pressure threshold of the clamping knuckle. A high friction coefficient allows a higher contact pressure without easy sliding. The maximum contact pressure threshold is the maximum pressure that the clamping knuckle can exert determined according to the upper limit value of the friction coefficient interval, ensuring that the package will not be damaged. This threshold is derived from experimental data or material properties. The segmented increasing strategy refers to a method of increasing the pressure threshold growth rate when the friction coefficient crosses the high and medium friction domains to adapt to the grasping requirements under different friction conditions. This strategy aims to optimize the grasping effect and avoid overpressure or underpressure.

[0099] In the embodiments of this application, first, measure the friction coefficient interval of the cardboard surface through a sensor and set the maximum contact pressure threshold of the clamping knuckle according to its upper limit value. Specifically, use a machine learning model to analyze the historical data set to predict the appropriate pressure threshold under different friction coefficient conditions. If the friction coefficient crosses the high and medium friction domains, adopt a segmented increasing strategy to gradually increase the growth rate of the pressure threshold. The final result is an optimized maximum contact pressure threshold setting based on the current friction coefficient interval, ensuring the optimal grasping effect under various friction conditions.

[0100] 602. Calculate the safety margin of the initial closing angle of the knuckle based on the distribution density of the regions with the maximum curvature in the geometric model. The safety margin is positively correlated with the product of the number of curvature extreme points and the lower limit value of the friction coefficient. In step 602, the distribution density of the regions with the maximum curvature refers to the number and distribution of the regions with the largest curvature change on the wrapped surface, and is used to calculate the safety margin of the initial closing angle of the knuckle. Regions with high curvature may require more careful clamping. The safety margin is the allowable range of the initial closing angle of the knuckle, and is positively correlated with the product of the number of curvature extreme points and the lower limit value of the friction coefficient, ensuring the safety of the clamping operation. The larger the safety margin, the more stable the clamping. The number of curvature extreme points refers to the number of points with the most significant curvature change on the wrapped surface, and is used to evaluate the complexity of the package. More extreme points mean that more delicate clamping control is required.

[0101] In the embodiments of the present application, a scanning technique is used to obtain the geometric model of the package and analyze the distribution density of the regions with the maximum curvature therein. Based on these data, the safety margin of the initial closing angle of the knuckle is calculated. The safety margin is directly proportional to the product of the number of curvature extreme points and the lower limit value of the friction coefficient. This step involves complex geometric analysis and mechanical simulation. Finite element analysis is used to simulate the stress distribution at different clamping angles to ensure that the knuckle can close at the optimal position and reduce potential damage to the package. The final output is an optimized configuration scheme for the initial closing angle of the knuckle to ensure efficient and safe clamping operation.

[0102] 603. Dynamically attenuate the maximum contact pressure threshold based on the shortest distance from the boundary of the fragile region to the clamping region in the surface stress distribution data. The attenuation coefficient is inversely proportional to the square of the shortest distance. In step 603, the surface stress distribution data refers to the stress distribution of each point on the surface of the package and is used to identify the fragile regions and their boundaries. These data help to adjust the clamping force and avoid damaging the package. The shortest distance from the boundary of the fragile region to the clamping region refers to the minimum distance between the boundary of the fragile region and the actual clamping region, and is used to dynamically attenuate the maximum contact pressure threshold. The closer the distance, the lower the pressure should be. The attenuation coefficient is inversely proportional to the square of the shortest distance and is used to adjust the maximum contact pressure threshold to ensure that the fragile region is not under excessive pressure. This proportional relationship ensures a progressive pressure adjustment.

[0103] In the embodiments of the present application, stress distribution data on the surface of the package is obtained through stress sensors or simulation software, and the fragile areas and their boundaries are identified. Based on this information, the shortest distance from the boundary of the fragile area to the clamping area is calculated, and the maximum contact pressure threshold is dynamically attenuated accordingly. The attenuation coefficient is inversely proportional to the square of the shortest distance, ensuring that the pressure is appropriately reduced when approaching the fragile area to avoid damage. This step involves the application of data analysis and real-time monitoring technologies, and a controller is used for precise pressure adjustment. Finally, a pressure threshold adjustment scheme adapted to the current package characteristics is formed to ensure the safety of the fragile area during the clamping process.

[0104] 604. Based on the safety margin, when the upper limit of the contact pressure of the clamping knuckle reaches the threshold warning range, high-frequency and small-amplitude vibrations of the contact surface are triggered, and the pressure gradient distribution of the clamping knuckle is monitored in real time. If it is detected that the pressure concentration area overlaps with the area with the largest curvature in the geometric model, the opening compensation angle of the adjacent knuckle is increased to disperse the stress.

[0105] In step 604, the threshold warning range refers to the warning range triggered when the upper limit of the contact pressure of the clamping knuckle approaches the set maximum value, and is used to initiate further safety measures. This range provides early warning to prevent overpressure. High-frequency and small-amplitude vibrations are a way of slight vibrations, which are used to disperse the local stress concentration phenomenon generated during the clamping process. The vibrations help to evenly distribute the pressure and prevent local overload. The opening compensation angle refers to the opening angle of the adjacent knuckle increased when it is detected that the pressure concentration area overlaps with the area with the largest curvature in the geometric model to disperse the stress. This method can effectively avoid damage caused by local stress concentration.

[0106] In the embodiments of the present application, when the upper limit of the contact pressure of the clamping knuckle reaches the threshold warning range, the system automatically triggers high-frequency and small-amplitude vibrations of the contact surface. This process is realized by a micro-vibration device installed on the knuckle. At the same time, the pressure gradient distribution of the clamping knuckle is monitored in real time. If it is detected that the pressure concentration area overlaps with the area with the largest curvature in the geometric model, the opening compensation angle of the adjacent knuckle is immediately increased to disperse the stress. This step involves the application of advanced sensing technologies and feedback control mechanisms, such as using capacitive sensors for real-time pressure monitoring. The finally output is an optimized clamping strategy configuration scheme to ensure an efficient and safe operation process.

[0107] The following is a specific example: In a busy logistics center, when an intelligent logistics robot starts working, it first conducts a detailed scan of the cartons entering the sorting line to identify their material hardness grade and friction coefficient range. Assuming that the material hardness of the carton exceeds the flexible threshold, the robot switches to the clamping mode. Next, the robot sets the maximum contact pressure threshold of the clamping knuckles according to the upper limit value of the friction coefficient range and adopts a segmented increasing strategy to increase the pressure threshold growth rate. Then, it uses scanning technology to analyze the geometric features of the package surface and calculates the safety margin of the initial closing angle of the knuckles. For the identified fragile areas, the maximum contact pressure threshold is dynamically attenuated to ensure that the fragile areas are not subjected to excessive pressure. Finally, during the clamping process, if the contact pressure approaches the threshold warning range, the robot triggers high-frequency micro-amplitude vibrations and monitors the pressure distribution in real time. When necessary, it increases the opening compensation angle of adjacent knuckles to ensure an efficient and safe operation process.

[0108] In summary, steps 601 to 604 significantly improve the clamping ability and safety of the intelligent logistics robot when handling different types of packages. Through fine friction coefficient analysis, geometric feature recognition, and real-time monitoring and adjustment, the stability and reliability of the clamping operation are ensured. Especially when facing complex surfaces and fragile items, this method can effectively reduce the damage risk and improve the safety and accuracy of automated operations. In addition, by dynamically adjusting the contact pressure and knuckle angles, the robot can flexibly respond to various situations, greatly enhancing the flexibility and efficiency of the entire logistics system. This refined adjustment strategy not only guarantees work efficiency but also maximally protects the safety of the goods.

[0109] To solve the problem of accurately obtaining the surface material type, three-dimensional geometric profile, and surface stress distribution of different types of packages by an intelligent logistics robot, this solution combines the force closure stability score, timing parameters, and joint torque balance factor to generate a multi-objective optimization weight set and finally outputs a dynamic sorting strategy. The core idea is to formulate the optimal grasping plan by comprehensively considering various factors such as mechanical stability, time synchronization, and mechanical safety. This method fully considers various variables during the package handling process to ensure that each operation can achieve the best effect, effectively improving the accuracy and safety of package handling. In some embodiments, the obtaining of the surface material type data, three-dimensional geometric profile data, and surface stress distribution data of the package to be sorted by the logistics robot in step 101 includes: 701. Collect the surface reflectance map of the package through the multi-spectral sensor of the logistics robot, extract the multi-band spectral response difference to match the surface material type data, and enable polarization filtering to suppress interference in the high-reflection area; In step 701, the multispectral sensor is a device capable of collecting the reflectance of light at different wavelengths and is used to identify the material of the package surface. The materials are distinguished by analyzing the differences in reflectance in different bands. The surface material type data is information extracted based on the differences in multi-band spectral responses and describes the types and characteristics of the package surface materials. These data help determine the appropriate grasping strategy. Polarization filtering refers to a technique used to suppress the interference in highly reflective areas and improve the accuracy of material identification. By filtering out unnecessary reflected light, the recognition effect of key information is enhanced.

[0110] In the embodiment of the present application, first, the multispectral sensor is used to scan the package surface to collect the reflectance maps at different wavelengths. A machine learning algorithm is used to analyze these reflectance differences to match the surface material type data. For highly reflective areas, the polarization filtering technique is enabled to reduce the interference of reflected light and ensure the accuracy of material identification. The final result is an accurate surface material type data set, providing basic information for subsequent operations. This process involves complex optical analysis and the application of machine learning models, ensuring high-precision material identification.

[0111] 702. Use the fusion of structured light and stereo vision to reconstruct the three-dimensional point cloud, extract the curvature features to generate three-dimensional geometric contour data, and start multi-angle scanning to complement the point cloud for blind areas; In step 702, the fusion reconstruction of structured light and stereo vision refers to combining structured light projection and stereo vision technologies to generate a three-dimensional point cloud model of the package. This method can provide high-precision geometric information. The three-dimensional geometric contour data is the curvature features extracted based on the three-dimensional point cloud model and describes the shape and size of the package. These data help plan the optimal grasping path. Blind area complementation refers to using the multi-angle scanning method to supplement the complete three-dimensional point cloud data for areas that cannot be directly observed. This step ensures the integrity of the model.

[0112] In the embodiment of the present application, the structured light projection technology is used to project a specific pattern on the package surface, and combined with a stereo camera to capture images to generate an initial three-dimensional point cloud model. Then, computer vision algorithms are used to extract the curvature features in the model to generate three-dimensional geometric contour data. For parts with blind areas, start the multi-angle scanning strategy to supplement the point cloud data from different perspectives to ensure the integrity of the model. The final output is a three-dimensional geometric contour data set containing all details, providing accurate geometric information for gripping and handling. This process involves complex image processing and three-dimensional reconstruction technologies, ensuring high-precision geometric modeling.

[0113] 703. Load a pressure gradient through a flexible tactile array, record the strain rate and stress relaxation curve to generate surface stress distribution data, and trigger high-frequency sampling when a stress mutation is detected to obtain high-frequency sampling stress data; In step 703, the flexible tactile array is a pressure sensor array mounted on the robot gripper, used to detect the pressure changes on the surface of the package. These sensors can record the pressure distribution in real time. The surface stress distribution data is used to record the strain rate and stress relaxation curve, describing the stress state of each point on the package surface. This data helps to optimize the grasping force and avoid damaging the package. High-frequency sampling means that when a stress mutation is detected, a high-frequency sampling mechanism is triggered to capture more detailed pressure change information. This mechanism improves the resolution and reliability of the data.

[0114] In the embodiment of the present application, a pressure gradient is loaded through the flexible tactile array to record the strain rate and stress relaxation curve of the package surface in real time, generating surface stress distribution data. If a stress mutation is detected, the system automatically switches to the high-frequency sampling mode to increase the sampling frequency to capture more details. This step involves the application of advanced sensing technologies and data analysis methods, such as the dynamic time warping algorithm for processing stress curves. Finally, a detailed surface stress distribution data set is formed to help avoid damage caused by excessive pressure. This process involves complex data acquisition and signal processing technologies, ensuring high-precision stress monitoring.

[0115] 704. Synchronize the multi-band spectral response difference, multi-angle scanning completed point cloud, and high-frequency sampling stress data to a three-dimensional coordinate system to generate a cross-modal correlation mapping table.

[0116] In step 704, the cross-modal correlation mapping table is a correlation mapping table generated by synchronizing the multi-band spectral response difference, multi-angle scanning completed point cloud, and high-frequency sampling stress data to a unified three-dimensional coordinate system. This table provides comprehensive data integration for comprehensive analysis. The three-dimensional coordinate system refers to the spatial reference framework used to integrate various sensor data, ensuring the consistency and comparability of the data. This coordinate system is the basis for all data.

[0117] In the embodiment of the present application, various data obtained in steps 701 to 703 are synchronized to a unified three-dimensional coordinate system. Specifically, coordinate transformation algorithms (such as homogeneous coordinate transformation) are used to align data from different sources and generate a cross-modal correlation mapping table. This process involves complex mathematical calculations and data fusion technologies to ensure that all information can be comprehensively analyzed under the same spatial reference framework. The final result is a comprehensive and consistent data set, providing a solid foundation for subsequent decision-making and operations. This process involves advanced data fusion and coordinate transformation technologies, ensuring a high degree of data consistency.

[0118] The following is a specific example: In a busy logistics center, when an intelligent logistics robot starts working, it first conducts a detailed scan of the cartons entering the sorting line. The robot uses a multi-spectral sensor to collect the reflectance spectra of the package surface. By analyzing the differences in multi-band spectral responses, it enables polarization filtering to suppress interference in high-reflectance areas. Then, it adopts a technology that fuses structured light and stereo vision reconstruction to generate a three-dimensional point cloud model, and starts multi-angle scanning to complete the point cloud in blind areas. Subsequently, it loads a pressure gradient through a flexible tactile array, records the strain rate and stress relaxation curves to generate surface stress distribution data, and triggers high-frequency sampling when stress mutations are detected. Finally, all the collected data is synchronized to a three-dimensional coordinate system to generate a cross-modal correlation mapping table, providing accurate data support for subsequent efficient sorting.

[0119] In summary, steps 701 to 704 significantly improve the ability of the intelligent logistics robot to obtain key information when processing different types of packages. It not only improves the sorting efficiency but also effectively reduces the risk of item damage, achieving the safety and accuracy of automated operations. Through the application of advanced technologies such as multi-spectral sensors, the fusion of structured light and stereo vision reconstruction, and flexible tactile arrays, the high precision and reliability of material identification, geometric modeling, and stress monitoring are ensured. Especially through cross-modal data fusion, the robot can flexibly handle various complex situations, greatly enhancing the flexibility and efficiency of the entire logistics system. This method not only guarantees work efficiency but also maximally protects the safety of goods. This refined data acquisition and processing strategy ensure that each package can be safely and efficiently processed.

[0120] Figure 2 The following is a schematic structural diagram of a logistics robot control system based on artificial intelligence provided by an embodiment of the present application. As Figure 2 shown, the system includes: An acquisition module 21, which obtains the surface material type data, three-dimensional geometric contour data, and surface stress distribution data of the package to be sorted through the logistics robot; A determination module 22, which determines the material hardness grade and friction coefficient range of the package to be sorted based on the surface material type data; An identification module 23, which reconstructs the geometric model of the package to be sorted based on the three-dimensional geometric contour data and calculates the size of the minimum circumscribed cube of the package, and identifies the boundary of the fragile area on the surface of the package to be sorted based on the surface stress distribution data; The generation module 24 adjusts the adsorption mode or clamping mode of the logistics robot according to the material hardness grade and the friction coefficient range, adjusts the deployment angle of the logistics robot in combination with the minimum circumscribed cube size of the package, and at the same time restricts the contact point distribution and movement trajectory of the grasping action of the logistics robot according to the fragile area boundary, and generates a dynamic sorting strategy in combination with the current sorting line flow rate, the target sorting area capacity, and the robot joint load status; The update module 25 updates the action parameters of the logistics robot and the priority weights of the dynamic sorting decision through an artificial intelligence algorithm based on the execution result of the dynamic sorting strategy, so as to update the dynamic sorting strategy of the logistics robot.

[0121] Figure 2 The described logistics robot control system based on artificial intelligence can execute Figure 1 The described logistics robot control method based on artificial intelligence in the illustrated embodiment, the implementation principle and technical effects will not be elaborated. For the logistics robot control system based on artificial intelligence in the above embodiment, the specific ways in which each module and unit perform operations have been described in detail in the embodiment related to the method, and will not be elaborated here.

[0122] In a possible design, Figure 2 The logistics robot control system based on artificial intelligence in the illustrated embodiment can be implemented as a computing device, such as Figure 3 shown, the computing device may include a storage component 31 and a processing component 32; The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32.

[0123] The processing component 32 is used for the above Figure 1 The logistics robot control method in the illustrated embodiment.

[0124] Among them, the processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components for executing the above method.

[0125] The storage component 31 is configured to store various types of data to support the operation of the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disc.

[0126] Of course, the computing device may also necessarily include other components, such as input / output interfaces, display components, communication components, etc.

[0127] The input / output interface provides an interface between the processing component and the peripheral interface module, and the above-mentioned peripheral interface module may be an output device, an input device, etc.

[0128] The communication component is configured to facilitate communication between the computing device and other devices in a wired or wireless manner, etc.

[0129] Among them, the computing device may be a physical device or an elastic computing host provided by a cloud computing platform, etc. At this time, the computing device may refer to a cloud server, and the above-mentioned processing component, storage component, etc. may be basic server resources leased or purchased from a cloud computing platform.

[0130] The embodiment of the present application also provides a computer storage medium storing a computer program, and when the computer program is executed by a computer, it can implement the above Figure 1 shown embodiment of an artificial intelligence-based logistics robot control method.

[0131] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0132] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0133] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0134] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A control method for a logistics robot based on artificial intelligence, characterized in that, Including: Obtaining surface material type data, three-dimensional geometric contour data, and surface stress distribution data of the package to be sorted by a logistics robot; Determining the material hardness grade and friction coefficient range of the package to be sorted based on the surface material type data; Reconstructing the geometric model of the package to be sorted based on the three-dimensional geometric contour data and calculating the size of the minimum circumscribed cube of the package, and identifying the boundary of the fragile area on the surface of the package to be sorted based on the surface stress distribution data; Adjusting the adsorption mode or clamping mode of the logistics robot according to the material hardness grade and friction coefficient range, adjusting the unfolding angle of the logistics robot in combination with the size of the minimum circumscribed cube of the package, and at the same time constraining the distribution of contact points and the motion trajectory of the grasping action of the logistics robot according to the boundary of the fragile area, and generating a dynamic sorting strategy in combination with the current sorting line flow rate, the capacity of the target sorting area, and the load state of the robot joints; Based on the execution result of the dynamic sorting strategy, updating the action parameters of the logistics robot and the priority weight of the dynamic sorting decision through an artificial intelligence algorithm to update the dynamic sorting strategy of the logistics robot.

2. The method according to claim 1, wherein The generating a dynamic sorting strategy by constraining the distribution of contact points and the motion trajectory of the grasping action of the logistics robot according to the boundary of the fragile area, and combining the current sorting line flow rate, the capacity of the target sorting area, and the load state of the robot joints includes: Based on the boundary of the fragile area, delineating a candidate contact point area on the surface of the geometric model, excluding the coordinate points outside the safety distance of the boundary expansion of the fragile area, and generating an initial contact point set; Screening candidate contact point pairs that meet the condition of the coincidence degree between the grasping force direction and the projection of the package center of gravity from the initial contact point set, and calculating the force closure stability score of the candidate contact point pairs; Decomposing the grasping motion trajectory in the geometric model coordinate system into a horizontal translation component and a longitudinal following component, calculating the speed compensation amount of the horizontal translation component based on the sorting line flow rate, and generating the timing parameters of the longitudinal following component; Calculating the joint torque balance factor of the candidate contact point pairs according to the reciprocal of the ratio of the real-time current data of the logistics robot joints to the historical load, fusing the force closure stability score, the timing parameters, and the joint torque balance factor to generate a multi-objective optimization weight set, and outputting a dynamic sorting strategy based on the multi-objective optimization weight set.

3. The method according to claim 2, wherein The fusing the force closure stability score, the timing parameters, and the joint torque balance factor to generate a multi-objective optimization weight set includes: Determining the initial weight allocation of the force closure stability score, the timing parameters, and the joint torque balance factor based on the material hardness grade and the sorting line flow rate, and increasing the score weight when the material hardness grade is higher than the threshold; Converting the force closure stability score into a standardized interval, decomposing the timing parameters into a synchronization error tolerance and a time window matching degree, and combining the joint torque balance factor to generate a multi-dimensional parameter vector; Correcting the initial weight allocation according to the change rate of the capacity of the target sorting area, and superimposing the corrected initial weight allocation and the multi-dimensional parameter vector to generate a comprehensive score of the candidate solution. Truncate the joint torque balance factor according to the difference between the real-time current peak value of the joint of the logistics robot and the historical load threshold, and generate a multi-objective optimization weight set based on the comprehensive score of the candidate solutions.

4. The method according to claim 3, characterized in that, Converting the force closure stability score to a standardized interval, decomposing the timing parameters into a synchronization error tolerance and a time window matching degree, and generating a multi-dimensional parameter vector in combination with the joint torque balance factor, including: Dynamically map the force closure stability score to a standardized interval based on the score distribution range of the historical grasping case library; When decomposing the timing parameters, calculate the synchronization error tolerance according to the ratio of the sorting line flow rate to the size of the circumscribed cube of the package, and generate the time window matching degree based on the theoretical movement time from the center of gravity of the geometric model to the opening of the sorting container; Dynamically adjust the gain of the joint torque balance factor according to the fluctuation variance between the real-time current data of the joint and the historical average current; Align the dimensions of the standardized interval, synchronization error tolerance, time window matching degree, and the adjusted joint torque balance factor according to the priority rules of the logistics scenario to generate a multi-dimensional parameter vector including mechanical stability, timing synchronization, and mechanical safety.

5. The method according to claim 1, wherein Adjusting the adsorption mode or clamping mode of the logistics robot according to the material hardness grade and the friction coefficient interval, including: Set the adsorption force safety threshold based on the material hardness grade. When the material hardness grade is lower than the flexible threshold, activate the adsorption mode and dynamically adjust the distribution density of the adsorption hole array according to the friction coefficient interval; When the material hardness grade is higher than the flexible threshold, switch to the clamping mode. Calculate the upper limit of the contact pressure of the clamping knuckle based on the friction coefficient interval and constrain the initial closing angle of the knuckle. In the clamping mode, adjust the force closure direction of the knuckle based on the angle between the center of gravity projection of the geometric model and the action line of the clamping force; In the adsorption mode, match the texture of the adsorption contact surface according to the lower limit value of the friction coefficient interval, and increase the activation number of auxiliary adsorption holes in the local low-friction area; When the friction coefficient interval crosses the critical value, start the hybrid mode. Keep the adsorption in the central area of the adsorption hole, deploy the clamping knuckles in the edge area, and the initial closing angle of the knuckle is positively correlated with the diagonal length of the circumscribed cube size.

6. The method according to claim 5, characterized in that, Calculating the upper limit of the contact pressure of the clamping knuckle based on the friction coefficient interval and constraining the initial closing angle of the knuckle, including: Set the maximum contact pressure threshold of the clamping knuckle according to the upper limit value of the friction coefficient interval, and adopt a segmented increasing strategy to increase the pressure threshold growth rate when the friction coefficient crosses the high and medium friction domains; Calculate the safety margin of the initial closing angle of the knuckle based on the distribution density of the area with the maximum curvature in the geometric model, and the safety margin is positively correlated with the product of the number of curvature extreme points and the lower limit value of the friction coefficient; Dynamically attenuate the maximum contact pressure threshold based on the shortest distance from the boundary of the fragile area to the clamping area in the surface stress distribution data, and the attenuation coefficient is inversely proportional to the square of the shortest distance. Based on the safety margin, when the upper limit of the contact pressure of the clamping knuckle reaches the threshold warning range, high-frequency and small-amplitude vibration of the contact surface is triggered to monitor the pressure gradient distribution of the clamping knuckle in real time. If it is detected that the pressure concentration area overlaps with the area with the largest curvature in the geometric model, the opening compensation angle of the adjacent knuckle is increased to disperse the stress.

7. The method according to claim 1, wherein The acquisition of the surface material type data, three-dimensional geometric contour data, and surface stress distribution data of the package to be sorted by the logistics robot includes: Collect the surface reflectance map of the package through the multi-spectral sensor of the logistics robot, extract the multi-band spectral response differences to match the surface material type data, and enable polarization filtering to suppress interference in the highly reflective area; Use the fusion of structured light and stereo vision to reconstruct the three-dimensional point cloud, extract the curvature features to generate three-dimensional geometric contour data, and start multi-angle scanning in the blind area to complete the point cloud; Load the pressure gradient through the flexible tactile array, record the strain rate and stress relaxation curve to generate the surface stress distribution data, and trigger high-frequency sampling when stress mutation is detected to obtain the high-frequency sampling stress data; Synchronize the multi-band spectral response differences, multi-angle scanning completed point cloud, and high-frequency sampling stress data to the three-dimensional coordinate system to generate a cross-modal correlation mapping table.

8. A logistics robot control system based on artificial intelligence, characterized in that, Including: An acquisition module that acquires the surface material type data, three-dimensional geometric contour data, and surface stress distribution data of the package to be sorted through the logistics robot; A determination module that determines the material hardness grade and friction coefficient range of the package to be sorted based on the surface material type data; An identification module that reconstructs the geometric model of the package to be sorted based on the three-dimensional geometric contour data and calculates the size of the minimum circumscribed cube of the package, and identifies the boundary of the fragile area on the surface of the package to be sorted based on the surface stress distribution data; A generation module that adjusts the adsorption mode or clamping mode of the logistics robot according to the material hardness grade and friction coefficient range, adjusts the unfolding angle of the logistics robot in combination with the size of the minimum circumscribed cube of the package, and at the same time restricts the contact point distribution and movement trajectory of the grasping action of the logistics robot according to the boundary of the fragile area, and generates a dynamic sorting strategy in combination with the current sorting line flow rate, the capacity of the target sorting area, and the robot joint load status; An update module that updates the action parameters of the logistics robot and the priority weight of the dynamic sorting decision through an artificial intelligence algorithm based on the execution result of the dynamic sorting strategy to update the dynamic sorting strategy of the logistics robot.

9. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for controlling a logistics robot based on artificial intelligence as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, There is a computer program stored, and when the computer program is executed by the computer, it implements a method for controlling a logistics robot based on artificial intelligence as described in any one of claims 1 to 7.

Images