CN-121020064-B - AGV warehouse-in trolley and warehouse-in path planning method thereof

CN121020064BCN 121020064 BCN121020064 BCN 121020064BCN-121020064-B

Abstract

The invention discloses an AGV warehouse-in trolley and a warehouse-in path planning method thereof, belonging to the technical field of warehouse logistics, and comprising a scheduling planning module, a perception reconstruction module, a control navigation module, a communication coordination module, a resource allocation module, a positioning adjustment module, a leveling execution module, a recovery energy storage module and an energy management module; the invention improves the efficiency and the robustness of the path, realizes the conflict detection and the priority scheduling of each AGV warehouse-in workshop, reduces the congestion and the collision, and ensures the stable and reliable warehouse-in action; meanwhile, the feasibility and the execution stability of the fine tuning path are ensured, the dynamic, reasonable and efficient allocation and recovery of resources are realized, the overall scheduling efficiency is improved, and further, the task continuity is ensured.

Inventors

ZHOU JIAN
XU GUANGHUI
Li Luanluan

Assignees

宁波发现国际物流有限公司

Dates

Publication Date: 20260512
Application Date: 20250815

Claims (6)

1. The AGV warehouse-in trolley is characterized by comprising a scheduling planning module, a perception reconstruction module, a control navigation module, a communication coordination module, a resource allocation module, a positioning adjustment module, a leveling execution module, a recovery energy storage module and an energy management module; the scheduling planning module is used for constructing a dynamic weight matrix and planning a basic path of each AGV warehouse-in trolley; The perception reconstruction module is used for acquiring pose information and surrounding environment information of each AGV warehouse-in trolley in real time so as to dynamically generate a driving path; the control navigation module is used for calculating the wheel speed and steering instruction of each AGV warehousing trolley and controlling each AGV warehousing trolley to avoid the obstacle in real time; the communication coordination module is used for constructing a communication foundation of task-energy-traffic flow coupling scheduling and generating a resolution strategy; The resource allocation module is used for dynamically allocating the use authority and time quota of the resources according to reservation requests sent by the AGVs; The positioning adjustment module is used for acquiring visual characteristics of the goods shelves and calculating pose deviation between each AGV warehousing trolley and a target goods position; The leveling execution module is used for controlling each AGV to carry out dynamic leveling and executing the storage operation of the tray; The recovery energy storage module is used for converting waste heat into electric energy and storing the generated electric energy; The energy management module is used for preheating or assisting in heat dissipation of each AGV warehouse-in trolley battery pack; the specific steps of the perception reconstruction module for dynamically generating the driving path are as follows: S1.1, acquiring a warehouse static map, taking a channel intersection, a shelf clearance and a charging station position as nodes, taking a communication channel between the nodes as edges, acquiring a starting point and a target cargo space area in a task instruction, and simultaneously calculating the actual cost from the starting point to the corresponding node and the Manhattan distance from each node to the target cargo space to generate a basic path, wherein the specific calculation formula of the actual cost is as follows: In the formula (I), in the formula (II), Representing the actual cost of the device, Representing the first in the path The weight coefficient of the segment edge is calculated, Representing the first of the base paths The actual physical length of the segment edge; the specific calculation formula of the Manhattan distance is as follows: ; In the formula, Representing the manhattan distance of the person, Representing nodes Coordinates in the warehouse plane coordinate system, Representing target cargo space area Is defined by the reference point coordinates of (a); S1.2, smoothing the generated basic path, and acquiring the current position of each AGV warehousing trolley, the congestion condition of a peripheral channel, the dynamic obstacle coordinates, the electric quantity of each AGV warehousing trolley and the task emergency update and simultaneously acquiring the parameters of a dynamic weight matrix in real time; S1.3, splitting a basic path into a plurality of micro path sections according to preset intervals, calculating the comprehensive score of each micro path section, marking each micro path section with the score lower than a preset threshold value as a section to be optimized, and searching each communication path in a preset range around each section to be optimized to generate a plurality of alternative micro path sections; S1.4, calculating toughness values of alternative micro-path segments, selecting alternative micro-path segments with first rank as alternative paths according to high-to-low arrangement, embedding the alternative paths into a basic path, and correcting joint points of the alternative paths and the basic path through a smoothing algorithm to generate a complete path, wherein the specific calculation formula of the toughness values is as follows: ; In the formula, The toughness value is indicated by the terms "toughness value", The width-adaptation coefficient is represented by a number of bits, Representing the coefficient of stability of the history, Representing the temporal compatibility coefficient.
2. The AGV warehouse entry cart of claim 1, wherein the communication coordination module generates the resolution strategy by: S2.1, extracting all nodes in a complete path in real time, carrying out time prediction on each path node, generating a corresponding predicted arrival time window, simultaneously acquiring reservation information of all AGV warehousing trolleys, and synchronizing to an external central dispatching system and adjacent AGV warehousing trolleys in a path overlapping area through a wireless network; S2.2, the receiver writes information into a local path reservation cache table, compares the overlapping condition of path nodes, predicts conflict if a plurality of AGV warehouse-in trolleys occur simultaneously in a time window of the same path node, and judges the conflict level of the current AGV warehouse-in trolleys according to a preset dividing rule and the high level, the medium level and the low level; and S2.3, calculating corresponding resource priority scores according to task urgency, energy criticality, path replaceability, path toughness and historical waiting factors of each collision AGV warehousing trolley, arranging from high to low, reserving original paths of the AGVs with scores higher than a preset threshold according to ranking sequence, and adjusting paths of other AGVs.
3. The AGV warehouse entry trolley according to claim 2, wherein the positioning adjustment module calculates the pose deviation between each AGV warehouse entry trolley and the target cargo space by the following steps: S3.1, acquiring front end structure image data of a shelf, acquiring coordinate information according to the front end structure image data of each shelf, and converting the image coordinates into space pose under the coordinate system of each AGV warehouse-in trolley through calibration parameters so as to output corresponding three-dimensional coarse positioning information; S3.2, acquiring historical butt joint postures of each AGV relative to a goods location, comparing the historical butt joint postures with the self postures of each AGV warehousing trolley and the goods location postures in the three-dimensional coarse positioning information, and calculating the posture deviation amount of each AGV warehousing trolley and a target goods location, wherein the specific calculation formula of the posture deviation amount is as follows: the specific calculation formula of the pose deviation amount comprises the following steps: ; In the formula, The deviation of the pose is represented, Indicating the actual position of an AGV warehouse-in trolley The coordinates of the two points of the coordinate system, Indicating the actual position of each AGV warehouse-in trolley The coordinates of the two points of the coordinate system, Indicating the expected position of each AGV warehouse-in trolley The coordinates of the two points of the coordinate system, Indicating the expected position of each AGV warehouse-in trolley Coordinates; the specific calculation formula of the attitude angle deviation amount comprises the following steps: ; In the formula, The amount of deviation of the attitude angle is expressed, The actual attitude angle of each AGV warehouse-in trolley is represented, The expected attitude angle of each AGV warehouse-in trolley is represented; s3.3, comparing the calculated pose deviation amount with a target pose of a cargo space, calculating compensation adjustment amounts required by all AGVs, and carrying out micro-correction on the pose of each AGV warehouse-in trolley according to the calculated compensation adjustment amounts, wherein the specific calculation formula of the compensation adjustment amounts is as follows: ; In the formula, Representing the position compensation adjustment quantity, comprising three components, which correspond to the X-axis position compensation, the Y-axis position compensation and the attitude angle compensation in sequence, The X, Y axis position compensation scaling factor is represented, Representing the attitude angle compensation scaling factor, X, Y coordinates representing the pose of the cargo space object, Represents the current actual X, Y coordinates of the AGV warehouse-in trolley, Representing the attitude angle of the cargo space object, And the current actual attitude angle of each AGV warehouse-in trolley is represented.
4. An AGV warehouse entry trolley warehouse entry path planning method for realizing the AGV warehouse entry trolley function according to any one of claims 1-3, which is characterized by comprising the following specific steps: collecting and analyzing warehouse-in task information, and generating a basic path of each AGV warehouse-in trolley according to a warehouse static map; II, collecting the position and surrounding environment information of each AGV warehouse-in trolley in real time, and evaluating and adjusting the generated basic path; collecting path reservation information sent by each AGV warehouse-in trolley, carrying out conflict prediction according to each path reservation information, and simultaneously coordinating each conflict; IV, collecting real-time occupation conditions of resources, reserving resource use application of each AGV warehouse-in trolley, and carrying out dynamic quota allocation; And V, calculating pose deviation between each AGV warehouse-in trolley and the target cargo space so as to generate a corresponding fine adjustment path.
5. The method for planning a warehouse entry path of an AGV warehouse entry trolley according to claim 4, wherein the specific steps of performing the dynamic quota allocation in step IV are as follows: S4.1, collecting and distributing corresponding identifiers for various management resources, detecting whether AGVs enter a warehouse to occupy resources in a management resource area in real time, simultaneously calling occupation records of various resources in1 hour from a resource local database, and counting average occupation time length and peak use time period; S4.2, collecting resource reservation requests of all AGVs in-warehouse trolleys, arranging each resource reservation request according to an old-to-new sequence, marking each resource reservation request as a to-be-processed state, evaluating the priority of each resource reservation request, and arranging the resource reservation requests according to a high-to-low sequence; And S4.3, sequencing the priority of each resource reservation request based on various resource occupation conditions, distributing use quotas for various resources to generate corresponding quota instructions, then issuing the quota instructions to the corresponding AGV warehousing trolley, and if the AGV warehousing trolley with the allocated quota is in failure stagnation or task cancellation, recovering the original allocated quota and reassigning the original allocated quota to the AGV warehousing trolley with the first ranking of the priority of the resource reservation request.
6. The method for planning a warehouse entry path of an AGV warehouse entry trolley according to claim 5, wherein the specific step of generating the corresponding fine adjustment path in step V is as follows: S5.1, collecting reference pose data of a target cargo position and image feature points of a goods shelf, extracting pixel coordinates of the image feature points of the goods shelf, correcting the pixel coordinates of the image feature points of the goods shelf through distortion coefficients, and converting the corrected pixel coordinates into coordinates under a three-dimensional coordinate system through a perspective projection formula; S5.2, converting the coordinates in the three-dimensional coordinate system into the coordinates in the AGV warehousing trolley coordinate system to obtain the current plane position of each AGV warehousing trolley, calculating the pose deviation amount of each AGV warehousing trolley and the target cargo space, and setting constraint parameters of a fine adjustment path according to the physical properties of each AGV warehousing trolley and the target cargo space environment; S5.3, generating corresponding fine adjustment paths based on the calculated pose deviation and constraint parameters of the set fine adjustment paths, checking each fine adjustment path in real time, if the fine adjustment paths pass the check, entering an execution stage, otherwise, returning to regeneration, and then converting the fine adjustment paths passing the check into driving wheel control instructions and sending the driving wheel control instructions to corresponding AGV warehouse-in trolleys, wherein the specific formula of the fine adjustment path check is as follows: ; In the formula, Indicating that each AGV warehouse-in trolley is in fine adjustment process and the th The real-time distance between the individual surrounding obstacles, Real-time coordinates in the fine adjustment process of the AGV warehouse-in trolley are measured, Represent the first Coordinates of the individual obstacles.

Description

AGV warehouse-in trolley and warehouse-in path planning method thereof Technical Field The invention relates to the technical field of warehouse logistics, in particular to an AGV warehouse-in trolley and a warehouse-in path planning method thereof. Background Along with the development of intelligent logistics and intelligent storage technologies, in modern storage logistics systems, automatic guided vehicles become one of important equipment for realizing automation and intellectualization of material handling, particularly in intelligent warehouses, AGVs bear key tasks such as cargo handling, warehousing and ex-warehouse, and the like, traditional AGV path planning methods are mostly based on static maps, paths do not have dynamic sensing capability, problems such as path congestion, resource conflict, poor butting precision and the like are easily generated in complex environments, meanwhile, allocation management of resources is rough, comprehensive consideration of task urgency, vehicle states and resource history use conditions is lacked, and in addition, in the shelf butting process, the AGVs are often difficult to accurately complete warehousing actions due to environmental changes and image errors, and the operation efficiency and the accuracy are affected. The existing AGV warehouse-in trolley and warehouse-in path planning method thereof have low path planning efficiency and lack of coordination on the AGV warehouse-in trolley, and the existing AGV warehouse-in trolley and warehouse-in path planning method thereof have insufficient path fine tuning precision and single resource management mode. Disclosure of Invention The invention aims to solve the defects in the prior art, and provides an AGV warehouse-in trolley and a warehouse-in path planning method thereof. In order to achieve the above purpose, the present invention adopts the following technical scheme: an AGV warehouse-in trolley comprises a scheduling planning module, a perception reconstruction module, a control navigation module, a communication coordination module, a resource allocation module, a positioning adjustment module, a leveling execution module, a recovery energy storage module and an energy management module; the scheduling planning module is used for constructing a dynamic weight matrix and planning a basic path of each AGV warehouse-in trolley; The perception reconstruction module is used for acquiring pose information and surrounding environment information of each AGV warehouse-in trolley in real time so as to dynamically generate a driving path; the control navigation module is used for calculating the wheel speed and steering instruction of each AGV warehousing trolley and controlling each AGV warehousing trolley to avoid the obstacle in real time; the communication coordination module is used for constructing a communication foundation of task-energy-traffic flow coupling scheduling and generating a resolution strategy; The resource allocation module is used for dynamically allocating the use authority and time quota of the resources according to reservation requests sent by the AGVs; The positioning adjustment module is used for acquiring visual characteristics of the goods shelves and calculating pose deviation between each AGV warehousing trolley and a target goods position; The leveling execution module is used for controlling each AGV to carry out dynamic leveling and executing the storage operation of the tray; The recovery energy storage module is used for converting waste heat into electric energy and storing the generated electric energy; the energy management module is used for preheating or assisting in heat dissipation of each AGV warehouse-in trolley battery pack. As a further scheme of the invention, the specific steps of the perception reconstruction module for dynamically generating the driving path are as follows: S1.1, acquiring a warehouse static map, taking a channel intersection, a shelf clearance and a charging station position as nodes, taking a communication channel between the nodes as edges, acquiring a starting point and a target cargo space area in a task instruction, and simultaneously calculating the actual cost from the starting point to the corresponding node and the Manhattan distance from each node to the target cargo space to generate a basic path, wherein the specific calculation formula of the actual cost is as follows: In the formula, Representing the actual cost of the device,Representing the first in the pathThe weight coefficient of the segment edge is calculated,Representing the first of the base pathsThe actual physical length of the segment edge; the specific calculation formula of the Manhattan distance is as follows: In the formula, Representing the manhattan distance of the person,Representing nodesCoordinates in the warehouse plane coordinate system,Representing target cargo space areaIs defined by the reference point coordinates of (a); S1.2, smoothing the generated