CN111596658A - Multi-AGV collision-free operation path planning method and scheduling system - Google Patents

Abstract

The present invention discloses a path planning method and a scheduling system for the collision-free operation of multiple AGVs. The path planning method adopts a priority-based task scheduling strategy for scheduling, which can reasonably allocate tasks and effectively improve the efficiency of the scheduling system. At the same time, a priority-based AGV selection strategy is adopted, and the optimal path of each AGV is obtained through an A* algorithm, so that the operation path is relatively smooth, and the waiting time of the AGV is effectively reduced, thereby improving the operation efficiency of the entire system. The path planning method and scheduling system for the collision-free operation of multiple AGVs provided by the present invention can be widely used in manufacturing production plants, replacing traditional manual handling and conveyor belt transportation, accelerating the intelligent transformation of factories, reducing labor costs and greatly improving production efficiency.

Description

A path planning method and scheduling system for collision-free operation of multiple AGVs

technical field

The invention relates to the field of AGV (automatic guided vehicle) path planning, in particular to a path planning method and a scheduling system for multi-AGV collision-free operation based on the Internet of Things.

Background technique

With the gradual loss of my country's demographic dividend, the continuous rise in labor wages has put greater pressure on the cost side of 3C manufacturing companies. How to maintain a relatively low operating cost in the industry is the main problem faced by companies in the 3C manufacturing industry. From the perspective of the entire processing process of product production, most of the time materials are used for material storage, loading and unloading, transportation and the state to be processed. Therefore, improving the degree of factory automation and shortening non-processing time is the main way to reduce product cost. Within the framework of the Industry 4.0 functional factory, intelligent logistics is the core component of Industry 4.0 and an important link connecting supply and production. With the continuous development of intelligent manufacturing, intelligent manufacturing processes will be integrated into the manufacturing process to improve the efficiency of enterprises. competitiveness, and promote the transformation and upgrading of enterprises.

Automated Guided Vehicle (AGV) has become the main material distribution method in discrete production workshops. With the characteristics of high freedom of material handling paths, it has gradually replaced traditional manual handling as the main force of the logistics system.

At this stage, the amount of AGV deployment is increasing day by day, and the technical level of single AGV products is relatively mature. The core and focus of research in this field is to exert the collaborative operation ability of multiple AGVs. In the face of large-scale AGV cluster scheduling, the traditional centralized scheduling method calculates The load is heavy and the real-time performance is poor. It is urgent to find a new path planning method. Among them, the path planning of collision-free operation is the basic requirement of system scheduling. At present, scholars at home and abroad are studying the problem of multi-AGV scheduling, and have also proposed some algorithms, such as Traffic control method, route planning method based on time window, online scheduling strategy based on genetic algorithm. At present, the method of multi-AGV scheduling is still not mature enough, resulting in low utilization rate of AGV, long waiting time of AGV, unreasonable task allocation, resulting in low system efficiency and other shortcomings.

The Internet of Things technology enables each unit involved in production activities to acquire, process, integrate and exchange production process data, which opens up a new direction for the research on multi-AGV system scheduling technology.

SUMMARY OF THE INVENTION

The present invention provides a path planning method and a scheduling system for collision-free operation of multiple AGVs, so as to solve at least one problem existing in the prior art.

The present invention is achieved through the following technical solutions:

A path planning method for collision-free operation of multiple AGVs, the path planning method comprising:

S10, receiving tasks and assigning tasks, and assigning tasks according to the task scheduling strategy and the AGV selection strategy;

S20. According to the task allocation information, perform task scheduling according to the task scheduling strategy, and select an idle AGV to execute the task according to the AGV selection strategy, use the path planning algorithm to solve the optimal path of the AGV that needs to execute the task, and assign the optimal path to the AGV. as a planning path;

S30, if there is a running path, compare the planned path with the running path to determine whether there is a path conflict, if there is no path conflict, then go to step S50;

If there is a path conflict, predict the conflict time, if there is no time conflict, ignore the path conflict, and go to step S50; if there is a time conflict, determine the type of path conflict, if there is no opposite conflict in the path conflict type , then mark the conflicting nodes in the path conflict, and execute the corresponding waiting strategy according to the path conflict type, and go to step S50; if there is an opposite conflict in the path conflict type, mark the opposite conflicting node of the opposite conflict, and Use the path planning algorithm to solve the suboptimal path;

S40, using the suboptimal path as the planning path, and repeating step S30 until there is no more conflict in the planning path;

S50, the path planning is completed, a path list is created, the AGV runs according to the planned path, the running path information of the AGV is added to the path list, and the path information is updated in real time;

S60. Repeat the above steps S10 to S50 until the path planning of all tasks is completed.

Specifically, in the step S10, the task scheduling strategy includes: assigning priorities to tasks, establishing a task list, and arranging tasks in the order of their priorities to form a task waiting queue. The release order of the tasks is arranged, and when the tasks are scheduled, they are scheduled according to the arrangement order of the task list.

Specifically, in the step S10, the AGV selection strategy includes: traversing all idle AGVs, using a path planning algorithm to solve the optimal path of each idle AGV according to the current position point of the AGV and the task transfer point, and comparing the various idle AGVs The optimal path of the AGV, select the idle AGV with the shortest running distance as the optimal AGV to perform the task.

Preferably, in the step S30, the path conflict types include pursuit conflict, node conflict and opposite conflict;

The waiting policy corresponding to the path conflict type is:

If it is a pursuit conflict, the AGV running later will stop running, and the AGV running first will run to a safe distance before running;

If there is a node conflict, the AGV will be prioritized. When a node conflict is about to occur, the priority of each AGV will be judged. The AGV with low priority will stop running, and the AGV with high priority will first pass through the conflicting node before running the priority. Low AGV.

The present invention also provides a multi-AGV dispatching system, the dispatching system includes a dispatching control device and two or more AGVs, the AGV includes a main controller, a navigation module connected to the main controller, an RFID sensor, an infrared Obstacle avoidance module, power detection module, IoT communication module, motor drive module and power supply module;

The dispatching control device is provided with an Internet of Things module and an Internet of Things communication module, the Internet of Things communication module is connected to the dispatching control device through the Internet of Things module, and the dispatching control device communicates with each AGV through the Internet of Things communication module, and uses the above-mentioned The path planning method performs scheduling control of each AGV.

The present invention provides a path planning method and a scheduling system for multi-AGV collision-free operation, which adopts a priority-based task scheduling strategy for scheduling, which can reasonably allocate tasks and effectively improve the efficiency of the scheduling system. The optimal AGV selection strategy is realized through the Internet of Things technology, and the optimal path of each AGV is obtained through the A* algorithm, so that the running path is relatively smooth, effectively reducing the waiting time of the AGV, thereby improving the operating efficiency of the entire system. The path planning method and scheduling system for collision-free operation of multiple AGVs based on the Internet of Things provided by the present invention can be widely used in manufacturing factories, replace traditional manual handling and conveyor belt transportation, accelerate the intelligent transformation of factories, and reduce labor costs Greatly improve production efficiency.

Description of drawings

1 is a schematic flowchart of a route planning method according to Embodiment 1 of the present invention;

FIG. 2 is a schematic block diagram of the structure of a scheduling system according to Embodiment 2 of the present invention.

Detailed ways

In order to facilitate the understanding of those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

Example 1

As shown in FIG. 1, a path planning method for collision-free operation of multiple AGVs, the path planning method includes:

S10, receiving tasks and assigning tasks, and assigning tasks according to the task scheduling strategy and the AGV selection strategy;

S20. According to the task allocation information, perform task scheduling according to the task scheduling strategy, and select an idle AGV to execute the task according to the AGV selection strategy, use the path planning algorithm to solve the optimal path of the AGV that needs to execute the task, and assign the optimal path to the AGV. as a planning path;

S30, if there is a running path, compare the planned path with the running path to determine whether there is a path conflict, if there is no path conflict, then go to step S50;

If there is a path conflict, predict the conflict time, if there is no time conflict, ignore the path conflict, and go to step S50; if there is a time conflict, determine the type of path conflict, if there is no opposite conflict in the path conflict type , then mark the conflicting nodes in the path conflict, and execute the corresponding waiting strategy according to the path conflict type, and go to step S50; if there is an opposite conflict in the path conflict type, mark the opposite conflicting node of the opposite conflict, and Use the path planning algorithm to solve the suboptimal path;

S40, using the suboptimal path as the planning path, and repeating step S30 until there is no more conflict in the planning path;

S50, the path planning is completed, a path list is created, the AGV runs according to the planned path, the running path information of the AGV is added to the path list, and the path information is updated in real time;

S60. Repeat the above steps S10 to S50 until the path planning of all tasks is completed.

In the embodiment of the present invention, the path planning method for collision-free operation of multiple AGVs is applied to a multiple AGV scheduling system, and the multiple AGV scheduling system includes a scheduling control device and two or more AGVs. Preferably, the dispatching control device and each AGV are provided with an Internet of Things communication module, and the Internet of Things communication module is used for communication between AGVs and for communication between each AGV and the dispatching control device, that is, through the Internet of Things technology. Multi-AGV interconnection, that is to say, the AGV dispatching system according to the embodiment of the present invention communicates and interconnects based on the Internet of Things.

In the path planning calculation of the AGV, it can be the A* algorithm, the Floyd algorithm, the Dijkstra algorithm, etc. In this embodiment, the path planning algorithm is preferably the A* algorithm; in addition, in this embodiment, the shortest path is taken as the most Excellent path.

The following is a brief description of the principle of path planning for a single AGV based on the A* algorithm: The A* algorithm is an algorithm based on the Dijkstra algorithm and the BFS (Best First Search) algorithm. It uses a heuristic function to estimate the starting point on the plane. The distance cost value of the end point is a heuristic search algorithm. The heuristic search is to search for each reachable position in the state space for evaluation, get the best position, and then search from this position to the target. In this way, a large number of unnecessary search paths can be avoided, and the operation efficiency of the algorithm can be improved. The key of the A* algorithm is the heuristic function. Using different estimation functions, the path results may be different. The heuristic function is defined as:

f(n)=g(n)+h(n)

In the formula, f(n) is the evaluation function of node n; g(n) is the cost function of node n, which represents the cost value of moving from the starting point to the currently specified node; h(n) is the estimation function of node n, which represents The estimated moving cost from the current node to the end position, the predicted path here ignores the existence of obstacles, and the h value is obtained by some distance calculation methods. In this embodiment, the Manhattan distance is taken as the estimated value of h, that is, the sum of the horizontal distance and the vertical distance from the current node to the termination point. The formula is as follows:

h(n)=|X _n -X _end |+|Y _n -Y _end |

The biggest advantage of the A* algorithm is that it uses a heuristic function, which makes the path search directional, the search direction intelligently tends to the end point, the number of search nodes is small, and the speed is fast. The specific algorithm implementation steps are:

Step 1: Initialize, input starting point, target point and map information. Create OPEN list and CLOSE list, put the starting point in the OPEN list.

Step 2: Check if the OPEN list is empty, if not, enter the loop. If the OPEN list is empty, it means that the pathfinding fails and the program ends.

Step 3: Traverse the OPEN list, find the node with the smallest f value as the current processing node, delete it from the OPEN list, and add it to the CLOSE list. If the current processing node is the target point, it means that the pathfinding is completed, and the loop is exited (skip to step 6).

Step 4: Find the nodes that can be reached around the current processing node, search from the four directions of up, down, left and right, check whether it is in the CLOSE list, and ignore it if it is. If not, check if it is in the OPEN list. If it is not in the OPEN list, set the current processing node to its parent node. Calculate its f, g, and h values. G is obtained by adding the g value of its parent node to the cost value of one grid. The h value is obtained from the Manhattan distance, and f is equal to the sum of g and h. If it is already in the OPEN list, check whether the g value will be smaller if the new path is used to reach it. If the new g value is larger than the original, it means that running with the new path is not a wise choice, and no operation is required. Go to the next step. If the new g value is smaller, it means that it is better to run on the new path, then modify its parent node to the current processing node, and recalculate the g and f values.

Step 5: If the pathfinding has not been completed, go back to Step 2.

Step 6: Return to the starting point one by one through the parent node of the target point to get the complete path.

In order to allocate tasks reasonably and effectively improve the efficiency of the system, it is necessary to set a reasonable task scheduling strategy. In this embodiment, in the step S10, the task scheduling strategy preferably includes: assigning priorities to tasks, establishing a task list, and arranging tasks according to the order of their priorities to form a task waiting queue. If the tasks have the same priority , the tasks are arranged according to the order in which the tasks are released, and when the tasks are scheduled, they are scheduled according to the order in which the task list is arranged.

Optionally, the task scheduling strategy may further include: setting a timer, starting the timer when the optimal path is solved by using the path planning algorithm, and starting the next timer according to the order of the task list after the timer reaches a preset time. task scheduling.

In practical applications, the priority of tasks is manually set. For general tasks, they are assigned a low priority, while for special urgent tasks, they are assigned a high priority for priority processing. If the priorities of the tasks are relatively the same, they will be sorted according to the time of release of the tasks. As an option, tasks with low priority can be subdivided as needed, for example, low priority can be divided into lowest priority, common low priority, etc. Correspondingly, tasks with high priority can also be divided into The tasks are subdivided again as needed, such as the highest priority, normal high priority, etc.

How to select idle AGVs to perform tasks, make multiple AGVs operate in coordination, and avoid collisions, etc., so it is necessary to reasonably select idle AGVs to perform tasks, so this embodiment adopts an AGV selection strategy to achieve. In this embodiment, the AGV selection strategy preferably includes: traversing all idle AGVs, and using a path planning algorithm (the A* algorithm in this embodiment) to find the optimal value of each idle AGV according to the current position point of the AGV and the task transfer point. Path, by comparing the optimal paths of each idle AGV, select the idle AGV with the shortest running distance as the optimal AGV to perform the task.

Preferably, the AGV selection strategy may further include: before the AGV accepts the task scheduling, the power level of the AGV is detected, and if the power level of the AGV is lower than the preset value, the AGV is charged until the power level of the AGV is greater than or equal to The default value is acceptable for task scheduling. Through this strategy, the situation that the AGV cannot continue to run due to insufficient power during the execution of the task can be effectively avoided, and the occurrence of road blockage or more complicated situations can be avoided.

In this embodiment, in the step S30, the path conflict types include pursuit conflict, node conflict and opposite conflict. Among them, the pursuit conflict refers to two AGVs traveling in the same direction. If the speed of the AGV running behind is greater than the speed of the AGV in front, a pursuit collision will occur; node conflict refers to two AGVs at the same time Arriving at the same node, after passing through the node, the driving direction is different. If no corresponding measures are taken at this time, the two AGVs will collide at the node; the opposite conflict means that when the two AGVs are driving in the opposite direction, due to the same segment The path is only allowed to pass through one AGV at the same time, which will cause opposite collisions, resulting in a deadlock phenomenon.

Based on the production workshop environment and the actual operation of the AGV, the above three types of path conflicts include almost all path conflicts. In this embodiment, the solution strategy for the above-mentioned three path conflicts is as follows: the corresponding waiting strategy is adopted for the pursuit conflict and the node conflict, and the waiting strategy cannot be adopted for the opposite conflict, because the deadlock phenomenon will lead to road blockage, Manual intervention is required to solve the problem. Therefore, if there is an opposite conflict in the optimal path solved, it is necessary to re-plan the path and solve the sub-optimal path.

In order to make the running path relatively smooth and effectively reduce the waiting time of the AGV, the corresponding waiting strategy adopted in this embodiment for the path conflict type is:

If it is a pursuit conflict, the AGV running later will stop running, and the AGV running first will run to a safe distance before running;

If there is a node conflict, the AGV will be prioritized. When a node conflict is about to occur, the priority of each AGV will be judged. The AGV with low priority will stop running, and the AGV with high priority will first pass through the conflicting node before running the priority. Low AGV.

Specifically to this embodiment, when the priority is allocated to the AGV in this embodiment, the priority allocation method is as follows:

Step a: Calculate all conflicting nodes of each AGV to other AGVs, and classify the conflicting nodes;

Step b: Compare the number of conflicts of each AGV, assign the AGV with a large number of conflicts as a low priority, and assign an AGV with a small number of conflicts as a high priority;

Step c: If the number of conflicts is the same, compare the number of pursuit conflicts, assign a low priority to those with a large number of pursuit conflicts, and assign a high priority to those with a small number of pursuit conflicts.

By adopting such a waiting strategy and prioritizing AGVs, the path can be made smoother, and the waiting time of AGVs can be effectively reduced, thereby improving the operating efficiency of the entire system.

Example 2

This embodiment provides a multi-AGV dispatching system, as shown in FIG. 2 , which includes a dispatching control device and more than two AGVs; specifically, the AGV includes a main controller, and the main controller is connected to the main controller Navigation module, RFID sensor, infrared obstacle avoidance module, power detection module, Internet of Things communication module, motor drive module and power supply module; the dispatching control device is provided with an Internet of Things communication module, and the dispatching control device communicates with the Internet of Things communication module through the Internet of Things communication module. Each AGV communicates. Preferably, the navigation module is preferably a magnetic navigation sensor. When the scheduling system performs scheduling control on each AGV, it uses the path planning method described in Embodiment 1 for scheduling. For the specific path planning method, please refer to Embodiment 1, which will not be repeated here.

Among them, in Figure 2, only a schematic block diagram of the structure of one of the AGVs is drawn. In fact, the number of AGVs can be combined into 2, 3, 5, 10, etc. The specific number can be configured according to actual needs.

In this embodiment, the main controller of the AGV is responsible for receiving the instructions of the scheduling control device, analyzing the data of each sensor, and completing the control of all the actions of the AGV. The magnetic navigation sensor (the model is preferably D-MNSV6-X8) converts the deviation of the AGV relative to the tape path into 8-bit data and transmits it to the main controller by sensing the magnetic field strength around the tape on the preset path. The controller completes the guiding function through the algorithm. The RFID sensor is used to read the radio frequency tag on the map when the AGV is running, so as to complete the real-time positioning of the AGV. The infrared obstacle avoidance sensor is used to scan and detect obstacles in a short-distance fan-shaped area in front of the AGV. Once an obstacle is detected, the main controller is immediately triggered to interrupt the program to prevent the AGV from colliding and protect the production workshop equipment. The power detection module is used to detect and display the power of the AGV. When the power value is lower than the preset value, an alarm signal is generated to prompt the AGV to charge, and to control the AGV to charge at the corresponding place. The Internet of Things communication module is used for communication between AGVs and for communication between each AGV and a dispatch control device. That is, in the present invention, the interconnection of multiple AGVs is realized through the Internet of Things technology. The use of Internet of Things communication technology can make the data transmission and acquisition and data processing capabilities of the AGV dispatching system stronger, and also make the interaction between the AGV dispatching system and other equipment more convenient and more powerful.

In this embodiment, the scheduling control device may be a PC, and the PC is loaded with software for scheduling and controlling the AGV, and the relevant scheduling operations on the AGV are implemented through the software. In this embodiment, it adopts magnetic guidance as the AGV navigation method. Compared with other navigation methods, magnetic guidance navigation has the advantages of high navigation accuracy, low cost and high stability; magnetic guidance navigation belongs to the existing There are technologies, which will not be described in detail here.

The multi-AGV path planning method provided in Embodiment 1 of the present invention and the scheduling system provided in Embodiment 2 use a priority-based task scheduling strategy for scheduling, which can allocate tasks reasonably and effectively improve the efficiency of the scheduling system. The priority AGV selection strategy, and the optimal path of each AGV is obtained through the A* algorithm, so that the running path is relatively smooth, effectively reducing the waiting time of the AGV, thereby improving the operating efficiency of the entire system. The multi-AGV path planning method and scheduling system based on the Internet of Things provided by the embodiments of the present invention can be widely used in manufacturing factories, replacing traditional manual handling and conveyor belt transportation, accelerating the intelligent transformation of factories, reducing labor costs and greatly improving Productivity.

The contents mentioned in the above embodiments are preferred embodiments of the present invention, and are not intended to limit the present invention, and any obvious replacements are within the protection scope of the present invention without departing from the concept of the present invention.

Claims (10)

1. A path planning method for collision-free operation of multiple AGVs is characterized by comprising the following steps:

s10, receiving and distributing tasks, and distributing the tasks according to a task scheduling strategy and an AGV selection strategy when distributing the tasks;

s20, according to the task allocation information, performing task scheduling according to the task scheduling strategy, selecting an idle AGV to execute a task according to an AGV selection strategy, solving the optimal path of the AGV needing to execute the task by using a path planning algorithm, and taking the optimal path as a planning path;

s30, if a running path exists, comparing the planned path with the running path, judging whether a path conflict exists, and if no path conflict exists, turning to the step S50;

if a path conflict exists, predicting conflict time, and if no time conflict exists, ignoring the path conflict, and proceeding to step S50; if the time conflict exists, judging the type of the path conflict, if the opposite conflict does not exist in the path conflict type, marking the conflict node in the path conflict, executing a corresponding waiting strategy according to the path conflict type, and turning to the step S50; if the path conflict type has a directional conflict, marking a directional conflict node of the directional conflict, and solving a suboptimal path by using a path planning algorithm;

s40, taking the suboptimal path as a planned path, and repeating the step S30 until no opposite conflict exists in the planned path;

s50, completing path planning, creating a path list, enabling the AGV to run according to the planned path, adding running path information of the AGV into the path list, and updating the path information in real time;

and S60, repeating the steps S10-S50 until the path planning of all tasks is completed.

2. The path planning method according to claim 1, wherein in the step S10, the task scheduling policy includes: the method comprises the steps of distributing priorities of tasks, establishing a task list, arranging according to the sequence of the priorities of the tasks to form a task waiting queue, arranging according to the issuing sequence of the tasks if the priorities of the tasks are the same, and sequentially scheduling according to the arranging sequence of the task list when the tasks are scheduled.

3. The path planning method according to claim 2, wherein the task scheduling policy further comprises: and setting a timer, starting the timer when the optimal path is solved by using a path planning algorithm, and starting next task scheduling according to the sequence of the task list after the timing reaches the preset time.

4. The path planning method according to claim 1, wherein in step S10, the AGV selection strategy includes: and after all idle AGVs are traversed, solving the optimal path of each idle AGV according to the current position point and the task transfer point of the AGV by using a path planning algorithm, and selecting the idle AGV with the shortest running distance as the optimal AGV to execute the task by comparing the optimal paths of the idle AGVs.

5. The path planning method according to claim 4, wherein the AGV selection strategy further comprises: and the AGV performs electric quantity detection before receiving the task scheduling, if the electric quantity of the AGV is lower than a preset value, the AGV is charged, and the task scheduling can be received until the electric quantity of the AGV is larger than or equal to the preset value.

6. The path planning method according to claim 1, characterized in that: in the step S30, the path conflict types include a chase conflict, a node conflict, and a direction conflict;

the waiting policy corresponding to the path conflict type is as follows:

if the collision is chased, the AGV which runs later stops running, and runs again after the AGV which runs earlier runs to a safe distance;

if the node conflict exists, priority distribution is carried out on the AGVs, when the node conflict occurs, the priority of each AGV is judged, the AGV with low priority stops running, and the AGV with high priority runs again after the AGV with high priority passes through the conflict node preferentially.

7. The path planning method according to claim 4, characterized in that: when the AGV is prioritized, the priority assignment method is as follows:

step a: calculating all conflict nodes of other AGVs by each AGV, and classifying the conflict nodes;

step b: comparing the number of conflicts of each AGV, distributing the AGV with a large number of conflicts as a low priority, and distributing the AGV with a small number of conflicts as a high priority;

step c: if the number of collisions is the same, the number of collision pursuits is compared, the number of collision pursuits is assigned a low priority, and the number of collision pursuits is assigned a high priority.

8. The path planning method according to any one of claims 1 to 7, characterized in that: the path planning algorithm is an A-x algorithm.

9. A multiple AGV dispatching system, characterized in that: the dispatching system comprises a dispatching control device and more than two AGVs,

the AGV comprises a main controller, a navigation module, an RFID sensor, an infrared obstacle avoidance module, an electric quantity detection module, an Internet of things communication module, a motor driving module and a power supply module, wherein the navigation module is connected with the main controller;

the dispatching control device is provided with an internet of things communication module, communicates with each AGV through the internet of things communication module, and dispatches and controls each AGV by using the path planning method according to any one of claims 1 to 8.

10. The route planning method according to claim 9, characterized in that: the navigation module is a magnetic navigation sensor.

Images