CN117234214A

CN117234214A - Automatic shuttle for stacking industrial goods

Info

Publication number: CN117234214A
Application number: CN202311374850.3A
Authority: CN
Inventors: 李成贵
Original assignee: Henan Bozhao Electronic Technology Co ltd
Current assignee: Henan Bozhao Electronic Technology Co ltd
Priority date: 2023-10-23
Filing date: 2023-10-23
Publication date: 2023-12-15

Abstract

The invention discloses an automatic shuttle for stacking industrial goods, relates to the technical field of automatic control, and mainly solves the problems of scheduling, distributing and queuing of the automatic shuttle. The automatic shuttle for stacking industrial goods comprises a control module, an acquisition module, a processing module, a communication module, a scheduling module, a driving module, an access module and a display module, wherein the path planning and the preferential selection are carried out through a genetic map algorithm, the goods distribution and the movement of the shuttle are realized through an autonomous allocation method, the real-time position update of the goods and the shuttle is realized through a time inertia algorithm, the queuing and the warehouse entry of the shuttle are realized through a double-chain network algorithm, the scheduling flexibility of the automatic shuttle is greatly improved, the blocking event in the queuing process is reduced, and the adaptability of the complex environment is improved.

Description

Automatic shuttle for stacking industrial goods

Technical Field

The invention relates to the technical field of automatic control, in particular to an automatic shuttle for stacking industrial goods.

Background

An automated shuttle (AS/RS) is an automated system for storing and retrieving goods. It has the advantages of high efficiency, safety, accuracy, space saving, etc. In the field of industrial manufacturing, AS/RS is commonly used for palletizing of goods, that is, various articles are stacked according to a rule, so AS to facilitate storage, transportation and sale. The automatic shuttle is very efficient in goods storage and transportation, and reduces waiting time and delay between operations, so that production efficiency is improved. The automatic shuttle car can accurately capture the position information of each article in a very wide variety of cargoes, and ensure the accuracy of cargo storage, transportation and stacking. The automatic shuttle car can reduce personnel cost, shorten production period and improve storage and retrieval efficiency of components. Automated shuttles reduce the space occupied by storage and transportation by stacking items in a stereoscopic warehouse.

However, there are also certain drawbacks to using AS/RS:

1. scheduling is not flexible enough: the scheduling of automated shuttles is usually implemented by preset scheduling algorithms, and the scheduling strategies of these algorithms are not flexible enough to cope with real-time changing production demands and environmental changes.

2. Blocking during queuing: during the rush hour, a large number of tasks may exist in the system to be processed, so that a blocking phenomenon may occur in the queuing process between the automatic shuttling vehicles, and the operation efficiency of the system is affected.

3. Cannot adapt to complex environments: although the automatic shuttle realizes effective automatic storage and retrieval in a simple environment, in warehouses with high capability and factories with various types of commodities, huge quantity of commodity types and operation change complexity have great limitations, and the system adaptability and application universality are poor.

Disclosure of Invention

Aiming at the defects of the technology, the invention discloses an automatic shuttle for stacking industrial goods, which performs path planning and preferential selection through a genetic map algorithm, realizes goods distribution and shuttle mobilization through an autonomous allocation method, realizes real-time position updating of goods and the shuttle through a time inertia algorithm, realizes queuing and warehouse entry of the shuttle through a double-chain network algorithm, greatly improves the scheduling flexibility of the automatic shuttle, reduces blocking events in the queuing process, and improves the adaptability of complex environments.

In order to achieve the technical effects, the invention adopts the following technical scheme,

an automatic shuttle for stacking industrial goods comprises a control module, an acquisition module, a processing module, a communication module, a scheduling module, a driving module, an access module and a display module;

the control module adjusts and controls the working state of the shuttle car through the embedded system;

the acquisition module acquires the length, width and height of a warehouse, the number of lanes, the length, width and height of a lane, the length, width and height of a goods shelf, the number of layers and the number of columns of the goods shelf, the number of shuttling vehicles, the picking time, the waiting time, the movement speed of the shuttling vehicles and the lifting machine and the weight of goods through a sensor network and a laser radar;

the processing module processes the collected data information through a data processing method;

the communication module carries out remote transmission on the collected data information through a TCP/IP protocol;

the scheduling module is used for planning the path of the shuttle and scheduling and distributing the shuttle; the dispatching module comprises a positioning unit, a distribution unit and a planning unit, wherein the planning unit adopts a genetic map algorithm to carry out path planning and preferential selection, the distribution unit realizes cargo distribution and shuttle dispatching through an autonomous dispatching method, the positioning unit realizes real-time position updating of cargoes and shuttles through a time inertia algorithm, and the output end of the positioning unit is connected with the input end of the planning unit; the output end of the planning unit is connected with the input end of the distribution unit;

The driving module realizes normal operation of the shuttle and emergency treatment through a motor driving model;

the storage and pickup module realizes the warehouse in and out of the shuttle and the carrying and stacking of goods through the automatic carrying device; the automatic carrying device realizes the warehouse-in and warehouse-out and the goods carrying and stacking operation of the shuttle car through the mechanical arm and the conveyor belt;

the display module realizes man-machine interaction, parameter setting and freight workload display through the touch screen and the display screen;

the output end of the control module is respectively connected with the acquisition module, the processing module, the communication module, the scheduling module, the driving module, the access module and the input end of the display module, the output end of the acquisition module is connected with the input end of the processing module, the output end of the processing module is connected with the input end of the communication module, the output end of the communication module is connected with the input end of the scheduling module, the output end of the scheduling module is connected with the input end of the driving module, the output end of the driving module is connected with the input end of the access module, and the output end of the access module is connected with the input end of the display module.

As a further description of the above technical solution, the collection module includes a warehouse information collection unit, a roadway information collection unit, a shelf information collection unit, a shuttle information collection unit, a time information collection unit and a hoist information collection unit; the warehouse information acquisition unit acquires the size parameters of the warehouse through a laser scanner; the roadway information acquisition unit acquires the number and the size of the roadway through a visual identification method and a ranging sensor; the goods shelf information acquisition unit acquires the size and layout information of the goods shelf through the ranging sensor; the shuttle information acquisition unit acquires the shuttle state and parameters through the state monitoring circuit; the time information acquisition unit acquires picking time and waiting time through a system timer; the information acquisition unit of the elevator acquires the movement speed and the weight of goods of the elevator through a weight sensor, a displacement sensor, an acceleration sensor, a pressure sensor and a temperature sensor.

As a further description of the above technical solution, the processing module includes an encryption unit, a dividing unit, a classifying unit, a cleaning unit, a sorting unit and a storage unit, where the cleaning unit fills in data defects by adopting an interpolation algorithm and corrects portions of abnormal data by adopting an abnormality detection algorithm, the classifying unit classifies the cleaned data according to data types by using a hybrid clustering algorithm, the sorting unit sorts the classified data according to time by using a time-alignment method, the dividing unit divides the sorted data by using a data dividing algorithm, the encryption unit encrypts data blocks by using a hybrid encryption algorithm, the storage unit intelligently stores encrypted data information by using a metadata server and a data storage server, an output end of the cleaning unit is connected with an input end of the classifying unit, an output end of the classifying unit is connected with an input end of the sorting unit, an output end of the sorting unit is connected with an input end of the dividing unit, an output end of the dividing unit is connected with an input end of the encryption unit, and an output end of the encryption unit is connected with an input end of the storage unit.

As a further description of the above technical solution, the driving module includes a power supply unit, a motion unit, a frequency conversion unit and a braking unit, wherein the power supply unit provides energy for the shuttle vehicle through a solar panel and a wireless charging device, and the motion unit provides power for a motion mechanism of the shuttle vehicle through a brushless dc motor; the motion unit comprises a wheel group subunit, a lifting conveying subunit and an operation arm subunit; the wheel group subunit realizes the forward and backward movement, positioning and rotary movement of the shuttle car by controlling the motor and the gear; the wheel set subunit realizes position recording and positioning of the shuttle through the encoder; the lifting conveying subunit controls the lifting of the shuttle car and the conveying of goods through a control motor and a screw; the operation arm subunit performs grabbing and placing operations on cargoes through the air cylinder, the motor and the clamp; the speed of the shuttle is adjusted through intelligent converter by the frequency conversion unit, the emergency fault problem of the shuttle is handled through the mode of band-type brake by the braking unit, the input of motion unit is connected to the output of power supply unit, the input of frequency conversion unit is connected to the output of motion unit, the input of braking unit is connected to the output of frequency conversion unit.

As a further description of the above technical solution, the access module includes a scanning unit, a queuing unit and a handling unit, the scanning unit automatically identifies the goods number and the goods shelf position through a bar code scanner, the queuing unit realizes queuing and warehouse entry of the shuttle through a double-chain network algorithm, the handling unit grabs, handles and stacks the goods through a mechanical arm and a clamp, an output end of the scanning unit is connected with an input end of the queuing unit, and an output end of the queuing unit is connected with an input end of the handling unit.

As a further description of the above technical solution, the display module includes a configuration unit, an interaction unit, an announcement unit and an early warning unit, where the configuration unit implements remote parameter setting of the shuttle through a script program, the interaction unit performs multi-terminal remote monitoring of the shuttle through a touch screen, the announcement unit pushes a daily work report to a terminal device of a manager through an information popup window protocol, the early warning unit performs fault reminding of the shuttle through a buzzer, an output end of the configuration unit is connected with an input end of the interaction unit, an output end of the interaction unit is connected with an input end of the announcement unit, and an output end of the announcement unit is connected with an input end of the early warning unit.

As a further description of the above technical solution, the genetic map algorithm defines nodes and edges included in the image through a knowledge map, and presents the search space in a map form, and the genetic map algorithm obtains a set of path solution sets in the search space through a genetic algorithm path searching method, and updates the content of the knowledge map through a path solution set updating method; the genetic map algorithm judges the similarity between the path solution and the knowledge map by using a similar matching function and determines the position of the knowledge map, the similar matching function expands the knowledge map through the path solution to determine an optimizable path node, and a path strategy is added; the genetic map algorithm finally updates the path strategy through reinforcement learning so as to iteratively update the knowledge map and the path strategy until the algorithm loops 100 times; the similarity matching function is: （1)

in the formula (1), the components are as follows,representing a similarity function, ++>And->Expressed as meaning origin->Expressed as sense origin->Is (are) layered->Expressed as sense origin->Is (are) layered->Representing a time cost function, +.>Representing the time spent,/->Represents the distance between the origins of meaning;

and carrying out time statistics and evaluation through a time cost function in the matching process, wherein the time cost function is as follows:

（2）

In the formula (2), the amino acid sequence of the compound,representing the proportionality constant, +.>Representing an initial similarity value, +.>Representing a constant.

As a further description of the above technical solution, the working method of the autonomous deployment method is as follows:

s1, sensing a task environment and extracting task information, wherein a shuttle senses the task environment through a laser sensor, and recognizes and segments a task area through a computer vision and image processing method so as to extract task position, start and finish time and article description information;

s2, dividing tasks according to the attribute, space and time requirements, and classifying and dividing data through a data mining, classifying algorithm and a rule engine so as to optimize and schedule subsequent tasks;

s3, task priority weight assignment, namely acquiring task priority weight through a weighted evaluation and multi-attribute decision method based on task attributes and requirements so as to indicate task importance and emergency indexes;

s4, constructing a multi-target planning model, and building an objective function and constraint conditions through linear planning, integer planning and a multi-target planning method to measure the advantages and disadvantages and feasibility of the task allocation result;

s5, defining an objective function and constraint conditions, and defining the objective function and the constraint conditions through a rule engine according to the task scheduling requirement and the performance limit of the shuttle;

S6, giving weight according to task priority, and giving corresponding weight to different objective functions and constraint conditions through a data analysis method according to the setting of task priority so as to weigh the relation between the different objective functions and constraint conditions;

s7, realizing optimal task allocation, and carrying out iterative computation and optimization through an ant colony algorithm to obtain an optimal task allocation result so as to meet task requirements;

and S8, distributing the task to the shuttle, and transmitting the task distribution result to the shuttle through a network communication protocol and a scheduling control method according to the task distribution result so as to perform actual execution operation.

As a further description of the above technical solution, the working method of the double-chain network algorithm is as follows:

r1, building a shuttle queuing task into an HOAN model through a data structure algorithm and recording task information, wherein each node represents a task and contains time and resource information of the task;

r2, calculating an initial value of the length of the automatic shuttle queue through a mathematical calculation method;

r3, randomly generating two wolves by a random number generator and representing a queuing scheduling scheme;

r4, carrying out local search on the position of the wolf cluster through a self-adaptive updating strategy so as to search the optimal solution position;

And R5, adjusting the position of each wolf through an automatic updating function according to the information and the updating strategy of the wolf group so as to realize the position movement of the group whole to the optimal solution, and recording the optimal result in each iteration through a log recorder.

In summary, by adopting the technical scheme, the invention has the beneficial effects that,

the invention discloses an automatic shuttle for stacking industrial goods, which is characterized in that path planning and preferential selection are carried out through a genetic map algorithm, goods distribution and shuttle mobilization are realized through an autonomous allocation method, real-time position updating of goods and the shuttle is realized through a time inertia algorithm, queuing and warehouse entry of the shuttle are realized through a double-chain network algorithm, scheduling flexibility of the automatic shuttle is greatly improved, blocking events in the queuing process are reduced, and adaptability to complex environments is improved.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings which are required in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings may be obtained from these drawings without inventive faculty for a person skilled in the art,

FIG. 1 is a schematic diagram of the overall architecture of the present invention;

FIG. 2 is a schematic diagram of a scheduling module structure;

FIG. 3 is a schematic diagram of a process module configuration;

FIG. 4 is a schematic diagram of a driving module structure;

FIG. 5 is a schematic diagram of an access module.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. It should be understood that the description is only illustrative and is not intended to limit the scope of the invention. In addition, in the following description, descriptions of well-known structures and techniques are omitted so as not to unnecessarily obscure the present invention.

1-5, an automatic shuttle for stacking industrial goods comprises a control module, an acquisition module, a processing module, a communication module, a scheduling module, a driving module, an access module and a display module;

In a further embodiment, the collection module comprises a warehouse information collection unit, a roadway information collection unit, a goods shelf information collection unit, a shuttle information collection unit, a time information collection unit and a lifter information collection unit; the warehouse information acquisition unit acquires the size parameters of the warehouse through a laser scanner; the roadway information acquisition unit acquires the number and the size of the roadway through a visual identification method and a ranging sensor; the goods shelf information acquisition unit acquires the size and layout information of the goods shelf through the ranging sensor; the shuttle information acquisition unit acquires the shuttle state and parameters through the state monitoring circuit; the time information acquisition unit acquires picking time and waiting time through a system timer; the information acquisition unit of the elevator acquires the movement speed and the weight of goods of the elevator through a weight sensor, a displacement sensor, an acceleration sensor, a pressure sensor and a temperature sensor.

In a further embodiment, the processing module includes an encryption unit, a dividing unit, a classifying unit, a cleaning unit, a sorting unit and a storage unit, the cleaning unit fills in data defects by adopting an interpolation algorithm and corrects portions of abnormal data by adopting an abnormality detection algorithm, the classifying unit classifies the cleaned data according to data types by adopting a hybrid clustering algorithm, the sorting unit sorts the classified data according to time, the dividing unit is used for dividing the sorted data into a plurality of identical data blocks according to batches, the encryption unit encrypts the data blocks by adopting a hybrid encryption algorithm, the storage unit intelligently stores encrypted data information by adopting a metadata server and a data storage server, an output end of the cleaning unit is connected with an input end of the classifying unit, an output end of the classifying unit is connected with an input end of the sorting unit, an output end of the sorting unit is connected with an input end of the dividing unit, an output end of the dividing unit is connected with an input end of the encryption unit, and an output end of the encryption unit is connected with an input end of the storage unit.

In a specific embodiment, data to be processed is input into the input end of the cleaning unit, the cleaning unit fills in data defects through an interpolation algorithm and abnormal data is corrected through an abnormal detection algorithm, and the integrity and accuracy of the data are ensured. The cleaning unit outputs cleaned data, is connected with the input end of the classifying unit, the classifying unit classifies the cleaned data according to data types by adopting a hybrid clustering algorithm, the classified data is output, the classifying unit is connected with the input end of the classifying unit, the classifying unit ranks the classified data according to time to ensure the time sequence of the data, the sequenced data is output, the classifying unit is connected with the input end of the classifying unit, the classifying unit divides the sequenced data into a plurality of identical data blocks according to batches, the balance and the readability of the data are ensured, the classified data are output, the input end of the encrypting unit is connected, the encrypting unit encrypts the data blocks through a hybrid encryption algorithm, the safety of the data is ensured, the encrypting unit outputs the encrypted data as a final output result of the processing module, and the storing unit intelligently stores the final output result.

In a further embodiment, the driving module comprises a power supply unit, a motion unit, a frequency conversion unit and a braking unit, wherein the power supply unit provides energy for the shuttle vehicle through a solar panel and a wireless charging device, and the motion unit provides power for a motion mechanism of the shuttle vehicle through a brushless direct current motor; the motion unit comprises a wheel group subunit, a lifting conveying subunit and an operation arm subunit; the wheel group subunit realizes the forward and backward movement, positioning and rotary movement of the shuttle car by controlling the motor and the gear; the wheel set subunit realizes position recording and positioning of the shuttle through the encoder; the lifting conveying subunit controls the lifting of the shuttle car and the conveying of goods through a control motor and a screw; the operation arm subunit performs grabbing and placing operations on cargoes through the air cylinder, the motor and the clamp; the speed of the shuttle is adjusted through intelligent converter by the frequency conversion unit, the emergency fault problem of the shuttle is handled through the mode of band-type brake by the braking unit, the input of motion unit is connected to the output of power supply unit, the input of frequency conversion unit is connected to the output of motion unit, the input of braking unit is connected to the output of frequency conversion unit.

In a specific embodiment, the power supply unit provides energy to the movement unit: the energy is provided for the shuttle vehicle through the solar panel and the wireless charging device. The solar panel converts solar energy into electric energy, and the wireless charging device can wirelessly transmit the electric energy to the shuttle by utilizing an electromagnetic induction principle. The motion unit provides power for the frequency conversion unit: a brushless direct current motor is adopted as a movement mechanism of the shuttle car to provide power, and the movement unit outputs the power to the frequency conversion unit. The speed of the shuttle is adjusted by the frequency conversion unit: the frequency conversion unit adjusts the speed of the shuttle according to the signals and controls the running speed and the running stability of the shuttle. The brake unit deals with the emergency fault problem of the shuttle: the emergency fault problem of the shuttle is solved through the band-type brake, and the shuttle is prevented from sliding or deviating from a track due to the fault.

Further, the access module comprises a scanning unit, a queuing unit and a carrying unit, the scanning unit automatically identifies the goods number and the goods shelf position through the bar code scanner, the queuing unit realizes queuing and warehouse entry of the shuttle through a double-chain network algorithm, the carrying unit grabs, carries and stacks goods through a mechanical arm and a clamp, the output end of the scanning unit is connected with the input end of the queuing unit, and the output end of the queuing unit is connected with the input end of the carrying unit.

In a specific embodiment, the scanning unit automatically identifies the goods number and shelf location: and automatically scanning the bar code on the goods by a bar code scanner to determine the number of the goods and the position of the goods shelf. The queuing unit realizes queuing and warehouse entry of the shuttle car: queuing and warehouse entry of the shuttle car are realized by adopting a double-chain network algorithm. And distributing the shuttle to corresponding task points according to the task priority and the goods shelf position, and ensuring the timely delivery and storage of goods. The carrying unit grabs, carries and stacks goods: and grabbing, carrying and stacking cargoes through the mechanical arm and the clamp. And according to the output of the scanning unit and the instruction of the queuing unit, taking the goods off the goods shelf, carrying the goods to the entrance or unloading the goods from the shuttle to the appointed goods shelf. Connection between scanning unit and queuing unit: the output end of the scanning unit is connected with the input end of the queuing unit, and the serial numbers and the goods shelf positions of goods are input into the queuing unit for task allocation and scheduling of the shuttle. Connection between queuing unit and handling unit: the output end of the queuing unit is connected with the input end of the carrying unit, the instruction of the shuttle is transmitted to the carrying unit, the actions of the mechanical arm and the clamp are controlled, and the carrying and stacking of goods are completed.

Further, the display module includes a configuration unit, an interaction unit, an announcement unit and an early warning unit, the configuration unit realizes remote parameter setting of the shuttle through a script program, the interaction unit carries out multi-terminal remote monitoring of the shuttle through a touch screen, the announcement unit pushes a daily work report to terminal equipment of a manager through an information popup window protocol, the early warning unit carries out fault reminding of the shuttle through a buzzer, an output end of the configuration unit is connected with an input end of the interaction unit, an output end of the interaction unit is connected with an input end of the announcement unit, and an output end of the announcement unit is connected with an input end of the early warning unit.

In a specific embodiment, the configuration unit remotely sets parameters: parameters of the shuttle are set through the script program, including task allocation, speed adjustment, working modes and the like, and system parameters can be conveniently and remotely controlled and adjusted. Multi-terminal remote monitoring of an interactive unit shuttle: the multi-terminal remote monitoring of the shuttle is carried out through the touch screen, the running state, the position and the task information of the shuttle can be displayed in real time, fault points are rapidly positioned, and the efficiency and the accuracy of storage and logistics are improved. Daily work report pushing by the bulletin unit: daily work reports are pushed to terminal equipment of the manager in a popup window mode, wherein the daily work reports comprise task completion conditions and employee performances, and comprehensive data support and decision basis are provided for the manager. Fault reminding of the early warning unit shuttle: the fault reminding of the shuttle is carried out through the buzzer and the voice notification, the abnormal condition of the shuttle is quickly reflected, and the efficiency and timeliness of fault processing are improved.

In a further embodiment, the genetic map algorithm comprises the following working methods: firstly, establishing a knowledge graph, defining included nodes and edges, presenting a search space in a graph form, then carrying out genetic algorithm path search in the search space, obtaining a set of path solution sets, updating the path solution sets obtained by the genetic algorithm search on the knowledge graph, judging the similarity between the path solution and the knowledge graph by using a similar matching function, determining the position of the knowledge graph, expanding the knowledge graph by using the path solution, evaluating and analyzing the path solution, determining an optimizable path node, adding a path strategy, updating the path strategy by reinforcement learning, and finally continuously searching paths from the new search space, updating the knowledge graph and the path strategy until the algorithm circulates 100 times; the similarity matching function is:

（1)

and in the matching process, carrying out time statistics and evaluation by adopting a time cost function, wherein the time cost function is as follows:

（2）

In a specific embodiment, the principle of the genetic map algorithm is: modeling a path planning area, abstracting information such as roads, barriers, traffic rules, road conditions and the like into nodes, and forming a knowledge graph by the relation between the nodes. The knowledge graph can be constructed and updated by an automatic means and expert knowledge, and richer and more accurate knowledge support is provided. The genetic algorithm generates a group of path planning schemes through combination and mutation operation, the nodes involved in the schemes are corresponding to actual elements and relations in the map and the knowledge graph, the adaptability of each scheme is evaluated, the scheme with higher adaptability is selected as the father of the next generation, and the like until convergence or reaching a preset stop condition. And introducing a reinforcement learning algorithm to optimize the path on the basis of a path planning scheme generated by the genetic algorithm. The reinforcement learning treats path planning as a decision process by training the learning agent, and gives a target reward aiming at different states and actions, so that the learning agent can acquire feedback information of the path from the rewards, continuously adjust the path planning scheme and improve the feasibility and efficiency of the path. Mapping the optimized path planning scheme into an actual environment, executing path planning, acquiring the actual performance and effect of the path planning, inputting feedback information into an algorithm, adjusting parameters and models of the algorithm, providing more accurate knowledge and guidance for the next path planning, and comparing the knowledge and the sleep consensus algorithm with a knowledge graph and the sleep consensus algorithm under the condition that other conditions are the same, as shown in table 1:

According to the table 1, the node arrangement sequence of the knowledge graph algorithm is A, B, C, D, and the time is 367ms; the node arrangement sequence of the sleep consensus algorithm is B, D, A, C, and the time is 574ms; the node arrangement sequence of the genetic map algorithm is A, C, B, D, and the time is 68ms. Through comparison, the genetic map algorithm is found to be shortest in time under the condition of the same node number, which indicates that the genetic map algorithm has a good optimizing effect.

The pseudo code of the knowledge graph algorithm is realized as follows:

defining node types

enum NodeType {

Start, End, Intersection, Turn

}

Defining node attributes

table Node {

type: NodeType;

position: Point;

neighbor: array < Node >;// neighbor Node

}

Defining edge attributes

table Edge {

startNode: Node;

endNode: Node;

weight: float;// weight, representing path distance or cost

}

Defining genetic algorithm parameters

int potential_size=50;// POPULATION SIZE

float multiple_rate=0.1;// variability

float cross-speed=0.8;// CROSSOVER RATE

int elitism_count=5;// elite individual number

Definition of knowledge graph

table KnowledgeGraph {

nodes: Array<Node>;

edges: Array<Edge>;

}

Defining path policies

table PathPolicy {

startNode: Node;

endNode: Node;

Policy: array < float >;// probability value for each action

}

Establishing knowledge graph

buildKnowledgeGraph() {

graph = new KnowledgeGraph();

Generating nodes and edges and adding knowledge graph

return graph;

}

Path search using genetic algorithm

geneticAlgorithm(graph) {

The POPULATION = initial POPULATION:// initial POPULATION

for (i=0; i < 100; i++) {// cycle 100 times

FitnessValues=evaluateFitness (map);// calculate fitness value

newPopulation = [ ]// New population

Keep some elite individuals

elites = selectElites(population, fitnessValues, ELITISM_COUNT);

New individuals are generated

for (j = 0; j < POPULATION_SIZE - ELITISM_COUNT; j++) {

parent1 = select part (position, fitnessValues);// select parent1

parent 2=selection part (position, fitnessValues);// selection of parent2

child=cross-server (parent 1, parent2, cross-server);// cross

child = MUTATION (child, mutation_rate)

New population (child);// addition to New populations

}

newPoplation. Add (elites);// elite addition individuals

Position = newPosition;// update population

}

return population,// return Path solution set

}

Updating path solution sets onto knowledge-graph

updateKnowledgeGraph(paths, graph) {

for (path in paths) {

for (i = 0; i < path.size()-1; i++) {

startNode = path.get(i);

endNode = path.get(i+1);

Edge = new Edge (startNode, endNode, 1.0);// weight defaults to 1.0

if (!graph.nodes.contains(startNode)) {

graph.nodes.add(startNode);

}

if (!graph.nodes.contains(endNode)) {

graph.nodes.add(endNode);

}

if (!graph.edges.contains(edge)) {

graph.edges.add(edge);

}

return graph;

}

Determining similarity between path solution and knowledge graph, and determining knowledge graph position

findBestMatchingPath(paths, graph) {

bestPath = null;

bestMatchingRate = 0.0;

for (path in paths) {

matchingRate=calclulatenonodeMatchingRate (path, graph);// compute node similarity

if (matchingRate > bestMatchingRate) {

bestMatchingRate = matchingRate;

bestPath = path;

}

Finding the best position and returning

}

Knowledge graph expansion by utilizing path solution

expandKnowledgeGraph(path, graph) {

Addition of nodes and edges to knowledge-graph

}

Evaluating and analyzing the path solution, determining optimized path nodes, and adding path strategies

evaluatePaths(paths, graph, policy)

Further embodiment, the working method of the autonomous deployment method is as follows: the working method of the autonomous allocation method comprises the following steps:

In a specific embodiment, the principle of the self-tuning method is: the environment information of the task execution area is acquired through means such as a sensor network, a task space is abstracted into a group of nodes, basic information such as positions, types and priorities of tasks are described, the basic information is connected with other task nodes to form a task network, and data support is provided for task distribution. The task allocation problem is regarded as a multi-objective optimization problem, a plurality of objectives such as task response time, task execution priority, task processing cost and the like are considered, an optimal scheme of task allocation is obtained through a multi-objective planning algorithm, and the task allocation scheme is adjusted by continuously learning the task execution condition and the task change based on an autonomous learning and decision mechanism of an agent. For example, for newly appeared tasks or changes of existing tasks, the task allocation scheme can be recalculated through autonomous perception and multi-objective planning, so that efficient execution and reasonable allocation of the tasks are ensured. By monitoring and feedback, the execution condition of task allocation is checked and evaluated, the task allocation scheme and parameters are adjusted according to the feedback information, and the algorithm performance and task execution efficiency are optimized, as shown in table 2:

according to Table 2, the time required for node A to complete the task is the shortest and the time required for node D is the longest in the actually required task allocation scheme. In the task allocation scheme of the autonomous sensing algorithm, the task completion time of the nodes B and D is increased compared with the actually required task allocation scheme, and the task completion time of the nodes A and C is reduced. Compared with the actually required task allocation scheme, the task allocation scheme of the autonomous allocation method has the advantages that the time required by each node to complete the task is basically consistent, the total time consumption is shortest, and the total time consumption is only 38ms. Thus, autonomous deployment methods perform best in reducing the overall time consumption of task allocation.

The pseudo code of the autonomous deployment method is realized as follows:

defining task information

table Task {

position: Point;

startTime: Time;

endTime: Time;

itemDesc: String;

category, int;// task classification

priority: float;// task priority weight

}

Multi-objective planning model

maximaze f1 (x 1, x2,., xn)// first objective function

maximaze f2 (x 1, x2,., xn)// second objective function

...

maximaze fm (x 1, x2,., xn)// mth objective function

subject to

constraint1 (x 1, x2,., xn) <=c1// first constraint

constraint2 (x 1, x2,., xn) <=c2// second constraint

...

constraint (x 1, x2,) xn) <=cn// nth constraint

Definition task classification

int CATEGORY1 = 1;

int CATEGORY2 = 2;

int CATEGORY3 = 3;

Task priority weights

float PRIORITY1 = 0.5;

float PRIORITY2 = 0.3;

float PRIORITY3 = 0.2;

Performance index of/(shuttle)

int capability;// space Capacity

int speed;// vehicle speed

int maxDistance;// maximum distance travelled

Classifying all tasks according to attribute, location and time requirements and giving corresponding priority

classifyTasks() {

task 1 = [ ];// task of category one

task 2= [ ];// task of category two

task 3= [ ];// task of category three

for (task in allTasks) {

if (task.category == CATEGORY1) {

task.priority = PRIORITY1;

tasks1.add(task);

} else if (task.category == CATEGORY2) {

task.priority = PRIORITY2;

tasks2.add(task);

} else if (task.category == CATEGORY3) {

task.priority = PRIORITY3;

tasks3.add(task);

}

return (tasks1, tasks2, tasks3);

}

Multi-objective planning model construction

buildMOPModel() {

(tasks 1, tasks2, tasks 3) =classifytasks ();// categorize all tasks by attribute, location and time requirements

Defining objective functions and constraints

max f1=sum (t 1in tasks 1) t1.Priority t1 endtime;// minimize task completion time

min f2=sum (t 2 in tasks 2) t2.Priority t2. Starttime;// maximize task start time

max f3=sum (t 3 in tasks 3) t3.Priority t3.Itemdesc. Length;// minimize article description length

subject to

sum (t 1 in tasks 1) t1.Priority t1.itemdesc. Length < = capability;// space capacity constraint

sum (t 2 in tasks 2) t2.Priority distance (t 2. Position)/speed < = t2. Starttime;// maximum distance travelled constraint)

Sum (t 1 in tasks1, t2 in tasks2, t3 in tasks 3) t1. Priority+t2. Priority+t3. Priority=1;// priority constraint

...

Weights defining objective functions and constraints

w1 = 0.6;

w2 = 0.3;

w3 = 0.1;

Multi-objective planning model for/(and) minimization

minimize w1 * f1 + w2 * f2 + w3 * f3;

}

Multi-objective execution/planning algorithm

execute() {

solution = solveMOPModel ();// solution multi-objective planning model

assignTasks (solution);// assigning results to shuttles with corresponding requirements and collaboration capabilities

}

In a further embodiment, the working method of the double-chain network algorithm is as follows:

In a specific embodiment, the principle of the double-stranded network algorithm is: HOAN is a Directed Acyclic Graph (DAG) based model that treats storage shelves as nodes and automated shuttle request tasks as edges. The model divides the task allocation modes into two types: the forward task and the reverse task respectively correspond to the storage rack memory and the goods taking task. The HOAN model prioritizes tasks of the common storage area for short tasks. Meanwhile, the model also considers the influence of the moving time and the operating time of the automatic shuttle. DLA is an algorithm based on group intelligence, and comprises two processes of group behavior simulation and group evolution simulation. The algorithm uses the length of an automatic shuttle queue as an fitness function, uses a double-wolf group model as a basis, and searches for an optimal solution by continuously searching a state space, as shown in table 3:

According to the task allocation scheme of the semi-open loop algorithm shown in table 3, the node D is the longer time required to complete the task, and the node a is the shortest time required to complete the task. In the task allocation scheme of the double wolf's swarm algorithm, the task completion time of the node C is shortest, and the task completion time of the node D is longest. In the task allocation scheme of the double-chain network algorithm, the shortest time required for completing the task is node C, and the longest time required for completing the task is node B. It can be seen that different algorithms assign tasks in different ways and priorities, resulting in differences in the time required for each node to complete the task. The double-stranded network algorithm performs optimally in terms of the time required to complete the task.

The pseudocode implementation of the double-stranded network algorithm is:

defining a shuttle queuing task table including time information and resource information

table Task {

time: int;

resource: int;

}

Defining queue length for individual shuttles

int shuttleQueueLength;

The initial wolf group is/is defined, and each solution is the length of the shuttle queue

int[] wolf1, wolf2;

Woods/initialization and shuttle queue length

initialize() {

shattlequeue length = calcluteschuttellequeue length ()// calculate initial queue length from HOAN model

wolf1= generateRandomSolution (shuttleQueueLength);// randomly generating a solution for the first wolf

wolf2= generateRandomSolution (shuttleQueueLength);// randomly generating a second wolf solution

}

Update of the current solution

updateSolution(currentSolution, wolf) {

/(where currentSolution is the current solution and wolf is the current wolf)

newsolution=currentsolution;// new solution equals the current solution

for (i = 0; i < length(newSolution); i++) {

Trimming of each element in the solution

if (random() < wolf[1]) {

newSolution [ i ] + = wolf [2] (rand () -0.5);// random trimming

}

return newSolution;

}

Execution queuing scheduling algorithm

execute() {

Bestsolution=null;// initialize the optimal solution to null

for (i=0; i < maxItation; i++) {// iteration maxItation times

Updating wolf group information according to adaptive update strategy

wolf1 = updateSolution(wolf1, currentWolves);

wolf2 = updateSolution(wolf2, currentWolves);

Calculating the adaptation value of the updated wolf group

fitness1 = calculateFitness(wolf1);

fitness2 = calculateFitness(wolf2);

Ultrafresh optimal solution

if (fitness1 < bestFitness) {

bestFitness = fitness1;

bestSolution = wolf1;

}

if (fitness2 < bestFitness) {

bestFitness = fitness2;

bestSolution = wolf2;

}

return bestSolution,// return to optimal solution

}

While specific embodiments of the present invention have been described above, it will be understood by those skilled in the art that these specific embodiments are by way of example only, and that various omissions, substitutions, and changes in the form and details of the methods and systems described above may be made by those skilled in the art without departing from the spirit and scope of the invention. For example, it is within the scope of the present invention to combine the above-described method steps to perform substantially the same function in substantially the same way to achieve substantially the same result. Accordingly, the scope of the invention is limited only by the following claims.

Claims

1. An automatic shuttle for industry goods pile up neatly, its characterized in that: the system comprises a control module, an acquisition module, a processing module, a communication module, a scheduling module, a driving module, an access module and a display module;

2. An automated shuttle for industrial palletizing of goods according to claim 1, wherein: the acquisition module comprises a warehouse information acquisition unit, a roadway information acquisition unit, a goods shelf information acquisition unit, a shuttle information acquisition unit, a time information acquisition unit and a lifter information acquisition unit; the warehouse information acquisition unit acquires the size parameters of the warehouse through a laser scanner; the roadway information acquisition unit acquires the number and the size of the roadway through a visual identification method and a ranging sensor; the goods shelf information acquisition unit acquires the size and layout information of the goods shelf through the ranging sensor; the shuttle information acquisition unit acquires the shuttle state and parameters through the state monitoring circuit; the time information acquisition unit acquires picking time and waiting time through a system timer; the information acquisition unit of the elevator acquires the movement speed and the weight of goods of the elevator through a weight sensor, a displacement sensor, an acceleration sensor, a pressure sensor and a temperature sensor.

3. An automated shuttle for industrial palletizing of goods according to claim 1, wherein: the processing module comprises an encryption unit, a dividing unit, a classifying unit, a cleaning unit, a sorting unit and a storage unit, wherein the cleaning unit fills in data defects by adopting an interpolation algorithm and corrects abnormal data by adopting an abnormal detection algorithm, the classifying unit classifies the cleaned data according to data types by adopting a hybrid clustering algorithm, the sorting unit sorts the classified data according to time by adopting a time arrangement method, the dividing unit divides the sorted data by adopting a data segmentation algorithm, the encryption unit encrypts data blocks by adopting a hybrid encryption algorithm, the storage unit intelligently stores encrypted data information by adopting a metadata server and a data storage server, the output end of the cleaning unit is connected with the input end of the classifying unit, the output end of the classifying unit is connected with the input end of the sorting unit, the output end of the sorting unit is connected with the input end of the dividing unit, the output end of the dividing unit is connected with the input end of the encryption unit, and the output end of the encryption unit is connected with the input end of the storage unit.

4. An automated shuttle for industrial palletizing of goods according to claim 1, wherein: the driving module comprises a power supply unit, a motion unit, a frequency conversion unit and a braking unit, wherein the power supply unit provides energy for the shuttle car through a solar panel and a wireless charging device, and the motion unit provides power for a motion mechanism of the shuttle car through a brushless direct current motor; the motion unit comprises a wheel group subunit, a lifting conveying subunit and an operation arm subunit; the wheel group subunit realizes the forward and backward movement, positioning and rotary movement of the shuttle car by controlling the motor and the gear; the wheel set subunit realizes position recording and positioning of the shuttle through the encoder; the lifting conveying subunit controls the lifting of the shuttle car and the conveying of goods through a control motor and a screw; the operation arm subunit performs grabbing and placing operations on cargoes through the air cylinder, the motor and the clamp; the speed of the shuttle is adjusted through intelligent converter by the frequency conversion unit, the emergency fault problem of the shuttle is handled through the mode of band-type brake by the braking unit, the input of motion unit is connected to the output of power supply unit, the input of frequency conversion unit is connected to the output of motion unit, the input of braking unit is connected to the output of frequency conversion unit.

5. An automated shuttle for industrial palletizing of goods according to claim 1, wherein: the access module comprises a scanning unit, a queuing unit and a carrying unit, wherein the scanning unit automatically identifies the goods number and the goods shelf position through a bar code scanner, the queuing unit realizes queuing and warehouse entry of the shuttle through a double-chain network algorithm, the carrying unit grabs, carries and stacks goods through a mechanical arm and a clamp, the output end of the scanning unit is connected with the input end of the queuing unit, and the output end of the queuing unit is connected with the input end of the carrying unit.

6. An automated shuttle for industrial palletizing of goods according to claim 1, wherein: the display module comprises a configuration unit, an interaction unit, an announcement unit and an early warning unit, wherein the configuration unit realizes remote parameter setting of the shuttle through a script program, the interaction unit carries out multi-terminal remote monitoring of the shuttle through a touch screen, the announcement unit pushes a daily work report to terminal equipment of a manager through an information popup window protocol, the early warning unit carries out fault reminding of the shuttle through a buzzer, the output end of the configuration unit is connected with the input end of the interaction unit, the output end of the interaction unit is connected with the input end of the announcement unit, and the output end of the announcement unit is connected with the input end of the early warning unit.

7. An automated shuttle for industrial palletizing of goods according to claim 1, wherein: the genetic map algorithm defines nodes and edges contained in an image through a knowledge map, and presents a search space in a map form, and the genetic map algorithm obtains a group of path solution sets in the search space through a genetic algorithm path search method and updates the content of the knowledge map through a path solution set updating method; the genetic map algorithm judges a path solution and a knowledge graph by using a similar matching functionThe similarity of the spectrums and the position of the knowledge graph are determined, the similarity matching function expands the knowledge graph through path solution to determine optimized path nodes, and a path strategy is added; the genetic map algorithm finally updates the path strategy through reinforcement learning so as to iteratively update the knowledge map and the path strategy until the algorithm loops 100 times; the similarity matching function is: （1)

in the formula (1), the components are as follows,representing a similarity function, ++>And->Expressed as meaning origin->Expressed as sense origin->Is (are) layered->Expressed as a sense originIs (are) layered->Representing a time cost function, +.>Representing the time spent,/->Represents the distance between the origins of meaning;

（2）

8. An automated shuttle for industrial palletizing of goods according to claim 1, wherein: the working method of the autonomous allocation method comprises the following steps:

9. An automated shuttle for industrial palletizing of goods according to claim 5, wherein: the working method of the double-chain network algorithm comprises the following steps: