WO2023087863A1 - Kinematic modeling strategy and path planning method for material conveying platform - Google Patents

Abstract

Provided is a kinematic modeling strategy and path planning method for a material conveying platform. The material conveying platform comprises hexagonal module bodies (1) provided with omnidirectional wheels (15), a workbench plate (2), connecting pieces (3), telescopic rod supports (4), a Wi-Fi module (5), and self-locking rollers (6). The kinematic modeling strategy and path planning method comprises the following steps: establishing, by means of a multi-kinematic modeling strategy, a kinematic model library for control; configuring a working environment in an operation interface of an upper computer; acquiring, by means of RFIDs (14), geographical location information of a material to be conveyed, and feeding back same to the upper computer; planning a global conveying path by means of an improved ant colony algorithm; planning a local conveying path by means of an improved artificial potential field method; and implementing dynamic switching between multiple kinematic models on the planned conveying path, and calculating corresponding rotating speeds and rotating directions of related omnidirectional wheels (15). The material conveying platform is convenient to assemble and disassemble, dynamic switching between different kinematic models enables the conveying to be more stable, the convergence speed of path planning can be increased, the probability of local optimum is reduced, and the conveying efficiency is high.

Description

A Kinematics Modeling Strategy and Path Planning Method for Material Transfer Platform technical field

The invention relates to logistics transmission, in particular to a kinematics modeling strategy and path planning method for a material transmission platform.

Background technique

With the rapid development of Internet technology and the rapid popularization of intelligent and information technology in production and logistics, the traditional logistics industry has injected "smart" genes, but in actual production, most transmission systems still remain in the manual or semi-automatic transmission stage , the overall level of automation is not high; at the same time, the omnidirectional wheel has broad development prospects due to its irreplaceable unique characteristics compared with other wheel structures. A hot spot; different kinematic modeling strategies are suitable for different systems; traditional ant colony algorithm and artificial potential field method are prone to problems such as slow convergence speed and falling into local minimum in the path planning process.

Luo Jijun proposed a sorter with double-chain anti-skid barriers. The goods are stably transported through the engagement of the sprocket and the chain. However, when a certain meshing link fails, the goods will cause accumulation, which seriously affects the sorting efficiency. Ni Qiao designed an automatic flip-type sorter. The sorting direction of this system is fixed, the flexibility is low, and it adopts the flip-type form. There are restrictions on the size and weight of the goods, otherwise it is easy to cause collision damage to the goods. Cao Xiangang uses the strategy of multi-robot cooperative control to sort the goods on the belt conveyor. The use of multiple robotic arms increases the system cost. If one robotic arm fails, the burden on other robotic arms will increase, which will easily lead to system collapse. . Wang Hongbin proposed a hybrid algorithm combining the improved A* algorithm and the dynamic window method. Although this method realizes the path planning of the robot, due to the multi-objective transportation method, the complexity of the algorithm is correspondingly increased, and the timeliness is low. Li Mengxi proposed a path planning method based on the heuristic information expansion node A* algorithm combined with the mixed ant colony algorithm, but did not study the situation of obstacles in depth, and the adaptability of the algorithm was not high. He Shaojia combined the particle swarm algorithm with the ant colony algorithm to reduce the time for ants to search for the path in the early stage, but the path optimization ability needs to be improved, and the algorithm is easy to fall into local minimum in the later stage of operation.

Through the research of existing technologies, it can be analyzed that most of the transmission systems are difficult to meet the diverse transmission requirements, the transmission efficiency is not high, and the one-time investment cost is huge; in terms of path planning, most of the improved path planning algorithms still have search The results are unstable, easy to fall into local minimum and other problems, and the optimal path cannot be guaranteed.

Contents of the invention

Purpose of the invention: The purpose of the invention is to provide a kinematics modeling strategy and path planning method for the material transfer platform, so as to realize all-round automatic transfer and the area and height of the workbench can meet different transfer requirements, and improve the path planning algorithm The convergence speed can reduce the probability of falling into a local minimum, shorten the transmission path, and reduce the transmission time.

Technical solution: The present invention provides a material transmission platform. The working platform is embedded with a hexagonal module body with universal wheels, which can be infinitely spliced and expanded, and the workbench panel is enlarged or reduced accordingly. Connected by connectors; the gear train on the hexagonal unit module box on the workbench is in an equilateral triangle layout, and the module boxes are arranged in a honeycomb-like arrangement; the real-time switching of the kinematic model makes the transmission more stable; The material transmission system adopts the path planning method of the improved ant colony algorithm and the artificial potential field method, which can make the algorithm search process more targeted, improve the convergence speed, reduce the probability of falling into a local minimum value, and correspondingly improve the transmission efficiency of the transmission platform.

A number of connecting rods are fixed on the inner side of the lower surface of the module box to fix the motor embedded in the upper surface of the module body; there are a number of mounting holes connected by bolts on the upper surface of the module box for the installation and disassembly of the module; The size of the material is fixed, the surface area is not larger than the surface area of the module box, and it can cover three omnidirectional wheels.

There are six height-adjustable brackets installed under the workbench, which are connected to the workbench panel through the flange connection plate, and self-locking rollers are installed under the brackets at the four corners to realize the position adjustment of the entire transmission platform.

The control system uses STM32 single-chip microcomputer as the main control chip. The upper computer controller communicates with each module controller through the CAN bus protocol. The workbench is equipped with a Wi-Fi module for connecting to the Internet of Things, which can remotely monitor material transmission on the mobile terminal. Process; the materials to be transported are affixed with RFID electronic tags, and the geographical location information of the transported materials can be extracted by using the RFID receiver at the entrance of the platform.

The present invention provides a kinematics modeling strategy and path planning method for a material transfer platform, and its control method includes the following steps:

(1) Adopt the multi-kinematics modeling strategy to complete the establishment of the kinematics model library for control;

(2) The total number of unit modules configured in the host computer, including the side length of the hexagonal module and the number of X and Y axes, stipulates the material transmission speed;

(3) Obtain the geographic location information of the transported material through RFID, and feed it back to the host computer;

(4) Using the improved ant colony algorithm to plan the global transmission path;

(5) Using the improved artificial potential field method to plan the local transmission path;

(6) On the planned transmission path, the dynamic switching of multiple kinematic models is realized, and the corresponding rotational speed and steering of the relevant omni-directional wheels are calculated.

The material transmission platform is connected to the Internet of Things through the Wi-Fi module, and the entire transmission process of the material can be remotely monitored through the mobile terminal.

In the step (2), the steps for establishing a kinematics model under the condition that the adjacent wheels are asymmetrically distributed when the two modules participate in the transmission are:

(2.1) Select the center point o _(l) to establish the world coordinate system. The x _(l) axis increases from bottom to top, and the y _(l) axis increases from left to right. v _ox(l) means v _{o( l)} The component velocity on the x _(l) axis, v _{oy (l)} represents the component velocity of v _{o (l)} on the y _(l) axis, v _{o (l)} represents the linear velocity of point o _(l) , w _o(l) represents the angular velocity of point o _(l) , θ _(l) represents the angle between v _o(l) and x _(l) axis, where l=1,2,3, represents the number of modules involved in material transmission number;

(2.2) Establish the velocity components v x(2) of the three omnidirectional wheel velocities v ₁₍₂₎ , v ₂₍₂₎ and v ₃₍₂₎ on the x ₍₂₎ axis and y _{( 2)} axis of the body coordinate system , v _y(2) and the relationship between v _ox(2) and v _ey(2) in the world coordinate system:

In the formula, r ₍₂₎ represents the vertical distance between v _{y (2)} direction and point o ₍₂₎ , which is a constant;

(2.3) From v _qy(2) =w _q(2) R(q=1,2,3), calculate the relationship between each wheel speed w _q(2) and angular velocity w _o(2) :

In the formula, R represents the equivalent radius of the omnidirectional wheel;

(2.4) Formula (4) is written in matrix form, that is, the inverse kinematics equation:

In the step (2), the steps of establishing a kinematics model under the condition that the adjacent wheels are asymmetrically distributed when the three modules participate in the transmission are:

Inverse kinematics equation:

Calculate the speed of point o _(l) based on the rotational speeds of the three omnidirectional wheels to form a positive kinematic equation:

The specific process of the global path planning of the step (4) improving the ant colony algorithm is as follows:

(4.1) Environment modeling: The terrain map adopts the 0/1 method of rectangular rasterization, 0 indicates the passable position, 1 indicates the obstacle position, the upper left corner is the starting point position, the lower right corner is the target point position, and a right angle is established Coordinate system: the values of the x-axis and y-axis increase from left to right and from top to bottom, respectively, and the custom grid is 20×20, that is, x=y=i, where i∈[1,20] and i∈ N, and place a generation of ants. When a hexagonal module fails, it is regarded as an obstacle and represented by 1. If the obstacle is not full of a grid, it is still considered to occupy a grid;

(4.2) Improve the heuristic function of the ant colony algorithm: by introducing the sum of the distance between the current module position and the next module position and the distance between the next module position and the target module position, and adding weight factors λ ₁ and λ ₂ to the heuristic function Improvements are made to make the algorithm search process more targeted and reduce the probability of falling into a local minimum. The specific improvement methods are as follows:

In the formula, m is the current module position of the material, n is the next module position where the material may be transmitted, D is the target point position of the material transmission, η′ _mn is the improved distance heuristic function, d _nD is the Euclidean number between n and D Reed distance, and considering the uniqueness of the platform structure, set λ ₁ +λ ₂ = 2, and λ ₁ , λ ₂ >0;

(4.3) Improve the pheromone inspiration factor and distance expectation function factor of the ant colony algorithm, considering that the pheromone content in the early path is too low, which increases the blindness of the ant search path, and in the later stage, due to the excessive accumulation of pheromone, it shrinks The selectable range of the path causes the algorithm to fall into a local optimum. The importance of the pheromone heuristic factor and the distance expectation function factor is different at different times. The adaptive update strategy of the design factor is designed. The specific improvement method is as follows:

In the formula, K is the total number of iterations of the algorithm, k is the current number of iterations, α is the initial pheromone inspiration factor, β is the initial distance expectation function factor, and u is a constant. Considering the uniqueness of the platform structure, set u=2e /3.

(4.4) Determine the current feasible module road set, introduce the roulette algorithm to establish and update the optimized probability model between transmission modules, and optimize the probability model

The specific expression is:

In the formula, A _j represents the module set that the material can reach in the next step,

is the pheromone concentration on the optimized transmission path,

is the improved distance heuristic function;

(4.5) Update the pheromone content on the transmission path, the specific expression is:

τ _mn (t+1)＝(1-ρ)τ _mn (t)+Δτ _mn (t) (13)

In the formula, ρ is the pheromone volatilization coefficient, Δτ _mn is the sum of the pheromones released by two adjacent modules,

is the pheromone increment of two adjacent modules, L _j is the path length of ant j, and Q is the pheromone enhancement coefficient;

(4.6) Calculate the feasible solution obtained by the current iteration, compare it with the feasible solution obtained by the previous generation, record the optimal solution, and plan the global shortest path for material transmission after the number of iterations is over.

In the local path planning, fuzzy repulsion points are added to improve the traditional artificial potential field method. The additional repulsion can make the material escape from the local minimum point. The specific process of adding fuzzy repulsion points is as follows:

(5.1) First, draw a circle R1 with the object center r as the center of the circle, and the maximum distance _ρ0 of the obstacle action as the radius; construct a straight line l ₁ passing through the obstacle point F closest to r and the starting point O at the same time; record the distance between l ₁ and the x-axis Angle θ; determine the number c of obstacles falling in the circle R1; if c=1, execute step (5.2), if c=i (i=2,3,4...) then execute step (5.3);

(5.2) The number of obstacles is 1; find the distance L ₁ between the object and F; then draw a circle R2 with r as the center and L ₁ as the radius; rotate l ₁ forward or counterclockwise based on the center r The straight line l ₂ is constructed by the θ angle; the node farther away from F among the intersection points of l ₂ and circle R2 is the fuzzy repulsion point Q;

(5.3) The number of obstacles is multiple; record the distance L ₂ between the two obstacles closest to the center r of the object, the center point f between them, and the maximum size h of the object outline (known parameters); simultaneously calculate The distance L _min between F and r; draw a circle R3 with r as the center and L _min as the radius;

If h<L ₂ , construct the line l ₃ passing through f and r, and the node farther away from F in the intersection of l ₃ and circle R3 is the fuzzy repulsion point Q;

If h≥L ₂ , then construct a straight line l ₄ that passes through F and r, and use l ₄ as the center of the circle r as a reference, and construct a straight line l ₅ by rotating the angle θ positively or counterclockwise; the intersection point of l ₅ and the circle R3 is far away from F The node of is the fuzzy repulsion point Q.

A computer storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned kinematics modeling strategy and path planning method oriented to a material transfer platform are realized.

A computer device, including a memory, a processor, and a computer program stored on the memory and operable on the processor, when the processor executes the computer program, the above-mentioned kinematics construction oriented to a material transfer platform is realized. Modular strategy and path planning method.

Beneficial effect: compared with the prior art, the present invention has the following advantages:

1. The workbench can be expanded and height-adjusted according to the needs of material transmission volume. The module positioning is accurate. The system is suitable for different environments and realizes full automation. It is equipped with a Wi-Fi module for connecting to the Internet of Things platform, which can be remotely accessed on the mobile terminal. Monitoring, equipped with RFID electronic tags, automatically extracts the geographic location information of the transported materials, and realizes non-contact data communication;

2. The host computer realizes the real-time switching of different kinematic models, and solves the speed and steering of each wheel to make the transmission more stable;

3. In terms of path planning, the improved ant colony algorithm and artificial potential field method are proposed. Compared with the traditional algorithm, the convergence speed is improved, the probability of falling into a local minimum is reduced, and the transmission path is correspondingly shortened, reducing the transmission time.

Description of drawings

Figure 1 is a schematic diagram of the overall structure of the transmission platform;

Fig. 2 is a schematic structural diagram of the omnidirectional wheel module;

Fig. 3 is the concrete flowchart of control method;

Figure 4 is a schematic diagram of kinematic model solution in the case of asymmetric distribution of adjacent wheels when two modules participate in the transmission;

Figure 5 is a schematic diagram of kinematics model solution in the case of asymmetric distribution of adjacent wheels when three modules participate in the transmission;

Fig. 6 is the specific flowchart of the global path planning under the improved ant colony algorithm of the present invention;

Fig. 7 is a schematic diagram of the fuzzy repulsion point when there is a single obstacle;

Figure 8 is a schematic diagram of fuzzy repulsion points when there are multiple obstacles, wherein Figure 8a is a schematic diagram when h<L ₂ , and Figure 8b is a schematic diagram when h≥L ₂ ;

Fig. 9 is a comparison diagram of the trajectory of the traditional ant colony algorithm and the improved ant colony algorithm of the present invention, wherein Fig. 9a is the trajectory of the traditional ant colony algorithm, and Fig. 9b is the trajectory of the improved ant colony algorithm;

Fig. 10 is the contrast graph of the convergence curve of traditional ant colony algorithm of the present invention and improved ant colony algorithm;

Figure 11 is a comparison diagram of path planning when the artificial potential field method is locally optimal before and after improvement, where Figure 11a is a schematic diagram when h<L ₂ , and Figure 11b is a schematic diagram when h≥L ₂ .

Detailed ways

The technical solution of the present invention will be further described below in conjunction with the accompanying drawings.

As shown in Figures 1-2, the kinematics model for the material transfer platform of the present invention includes a hexagonal module body 1 equipped with omnidirectional wheels, a workbench panel 2, a connector 3, a telescopic rod bracket 4, and a Wi-Fi Module 5 and self-locking roller 6, the workbench is composed of multiple hexagonal module bodies 1 spliced together, the platform can be infinitely expanded and spliced, and the hexagonal module body 1 has a slot for positioning when connecting with the connecting piece 3, the connecting piece 3 is provided with a notch for positioning when connecting with the module box 11. When the hexagonal module body 1 is installed on the workbench panel 2, insert the slot into the notch on the connector 3, and the edge of the module box 11 Two installation holes 16 are respectively provided on both sides of the opening in the circumferential direction. The hexagonal module body 1 and the connector 3 are connected by bolts. The positioning process is simple and accurate, and can prevent the hexagonal module body 1 from moving up and down. The pole bracket 4 is connected to the workbench panel 2 through the flange plate. The telescopic pole bracket 4 is telescopic and used to adjust the height of the workbench. The Wi-Fi module 5 can be connected to the Internet of Things platform for remote monitoring on the mobile terminal. Self-locking rollers 6 Make the whole platform have good mobility and stability.

The hexagonal module body 1 includes a module box 11, a connecting rod 12, a tray 13, an RFID module 14, an omnidirectional wheel 15 and a mounting hole 16. The inner side of the module box 11 is fixedly connected to the top of the connecting rod 12, and the bottom of the three connecting rods 12 A tray 13 is fixed at the end, and an RFID module 14 is installed in the center of the hexagonal module body 1, which can automatically extract the geographical location information of the transported materials and realize non-contact data communication.

As shown in Figure 3, the specific process of the control method of the present invention is:

The multi-kinematics modeling strategy is adopted to complete the establishment of the kinematics model library for control; the number of total unit modules is configured in the host computer, including the side length of the hexagonal module and the number of X and Y axes, and the material transmission speed is specified; through RFID Obtain the geographical location information of the transported materials and feed it back to the host computer; use the improved ant colony algorithm to plan the global transmission path; use the improved artificial potential field method to plan the local transmission path; on the planned transmission path, realize the multi-kinematics model Switch dynamically, and calculate the speed and steering corresponding to the relevant omni-directional wheels. The material transmission platform is connected to the Internet of Things through the Wi-Fi module, and the entire transmission process of the material can be remotely monitored through the mobile terminal.

As shown in Figure 4, a kinematics model is established for the asymmetric distribution of adjacent wheels when two modules participate in the transmission:

Velocity components _v x _{(2 ) , v y ( 2)} The relationship between v _ox(2) and v _ey(2) in the world coordinate system is as follows:

In the formula, r ₍₂₎ represents the vertical distance between v _{y (2)} direction and point o ₍₂₎ , which is a constant;

From v _qy(2) ＝w _q(2) R(q=1,2,3), calculate the relationship between each wheel speed w _q(2) and angular velocity w _o(2) :

In the formula, R represents the equivalent radius of the omnidirectional wheel;

The above formula is written in matrix form, that is, the inverse kinematics equation:

As shown in Figure 5, the kinematics model is established when the adjacent wheels are not symmetrically distributed when the three modules participate in the transmission:

Inverse kinematics equation:

Calculate the speed of point o _(l) based on the rotational speeds of the three omnidirectional wheels to form a positive kinematic equation:

As shown in Figure 6, the improved ant colony algorithm flowchart of the present invention:

The initialization parameters include environment modeling: the terrain map adopts the 0/1 method of rectangular rasterization, 0 indicates the passable position, 1 indicates the obstacle position, the upper left corner is the starting point position, the lower right corner is the target point position, and a right angle is established Coordinate system: the values of the x-axis and y-axis increase from left to right and from top to bottom, respectively, and the custom grid is 20×20, that is, x=y=i, where i∈[1,20] and i∈ N, and place a generation of ants. When a hexagonal module fails, it is regarded as an obstacle and represented by 1. If the obstacle is not full of a grid, it is still considered to occupy a grid;

By introducing the distance between the current module position and the next module position and the sum of the distance between the next module position and the target module position, and adding weight factors λ ₁ and λ ₂ to improve the heuristic function, making the algorithm search process more targeted , and reduce the probability of falling into a local minimum, the specific improvement method is as follows:

In the formula, m is the current module position of the material, n is the next module position where the material may be transmitted, D is the target point position of the material transmission, η′ _mn is the improved distance heuristic function, d _nD is the Euclidean number between n and D Reed distance, and considering the uniqueness of the platform structure, set λ ₁ +λ ₂ = 2, and λ ₁ , λ ₂ >0;

Considering that the pheromone content in the early path is too low, which increases the blindness of the ant's search path, and in the later stage, due to the accumulation of too much pheromone, the optional range of the path is reduced, causing the algorithm to fall into a local optimum, and the design factor is adaptively updated Strategies, the specific improvement methods are as follows:

In the formula, K is the total number of iterations of the algorithm, k is the current number of iterations, α is the initial pheromone inspiration factor, β is the initial distance expectation function factor, and u is a constant. Considering the uniqueness of the platform structure, set u=2e /3;

Determine the current feasible module road set, introduce the roulette algorithm to establish and update the optimized probability model between transmission modules, and the optimized probability model

The specific expression is:

In the formula, A _j represents the module set that the material can reach in the next step,

is the pheromone concentration on the optimized transmission path,

is the improved distance heuristic function;

Judging whether a generation of ants has completed the path, if not completed, return to the stage of determining the candidate road set, if completed, select the best path in the generation of ants;

After the optimal path of a generation of ants is determined, it is judged whether the maximum number of iterations has been reached, and if not, the pheromone content on the transmission path is updated. The specific expression is:

τ _mn (t+1)＝(1-ρ)τ _mn (t)+Δτ _mn (t)

In the formula, ρ is the pheromone volatilization coefficient, Δτ _mn is the sum of the pheromones released by two adjacent modules,

is the pheromone increment of two adjacent modules, L _j is the path length of ant j, and Q is the pheromone enhancement coefficient;

If the maximum number of iterations is reached, the best path is selected to complete the shortest global path planning.

As shown in Figure 7, a schematic diagram of the fuzzy repulsion point for a single obstacle in the present invention:

Draw a circle R1 with the center of the object r as the center and the maximum distance _ρ0 of the obstacle as the radius; construct a straight line l ₁ passing through the obstacle point F closest to r and the starting point O at the same time; record the angle θ between l ₁ and the x-axis; Determine the number c of obstacles falling within the circle R1;

The number of obstacles is 1, find the distance L ₁ between the object and F; draw a circle R2 with r as the center and L ₁ as the radius; use l ₁ as the center of the circle r as the reference, and rotate the angle θ positively or counterclockwise to construct a straight line l ₂ . The node farther away from F in the intersection of l ₂ and circle R2 is the fuzzy repulsion point Q, which rotates counterclockwise in the figure.

As shown in Figure 8, a schematic diagram of fuzzy repulsion points when multiple obstacles in the present invention:

There are multiple obstacles, record the distance L ₂ between the two obstacles closest to the object center r, the center point f between them, and the maximum size h of the object outline (known parameters); simultaneously find F and r The distance L _min between them; draw a circle R3 with r as the center and L _min as the radius.

As shown in Fig. 8(a), if h<L ₂ , construct the straight line l ₃ passing through f and r, and the node farther away from F among the intersection points of l ₃ and circle R3 is the fuzzy repulsion point Q.

As shown in Fig. 8(b), if h≥L ₂ , then construct a straight line l ₄ passing through F and r, and construct a straight line l ₅ by rotating l ₄ with the center r as a reference, by an angle θ positively or counterclockwise. The node farther away from F in the intersection of l ₅ and circle R3 is the fuzzy repulsion point Q, as shown in Figure 8(b), which rotates counterclockwise in the figure.

As shown in Figure 9, the traditional ant colony algorithm of the present invention and the motion trajectory comparison figure of the improved ant colony algorithm:

The traditional ant colony algorithm has many inflection points, 13, and the running time is about 13.2 seconds. At the coordinates (8,10) in the figure, it passes through the center of two obstacles, which is an unreasonable path on this transmission platform. It should be Get rid of; Compared with its shortcoming, the ant colony algorithm motion locus inflection point improved by the present invention is less, 7, and running time is about 8.1 seconds, and inflection point number reduces 46.1%, and search efficiency improves 38.7%.

As shown in Figure 10, the traditional ant colony algorithm and the improved ant colony algorithm of the present invention are compared with the convergence trajectory:

The convergence curve of the traditional ant colony algorithm has large fluctuations, slow convergence speed and tends to be stable at about 60 iterations. At the same time, the algorithm falls into a local minimum, and the shortest path is about 35. Compared with its shortcomings, the invention of this paper The improved ant colony algorithm has small fluctuations, fast convergence speed and completes convergence at about 31 iterations, and the shortest path is about 30, shortening the length of the global path by 14.3%, increasing the convergence speed by 51.7%, and the transmission of the entire transmission platform Efficiency increases accordingly.

As shown in Figure 11(a), when the item falls into the local optimum and h<L ₂ , the fuzzy repulsion point is introduced to make the item pass between the two obstacles and escape the local minimum point, and at the same time, it can also make the item transportation avoid obstacles The path is optimal, that is, the path passed is the shortest.

It can be seen from Figure 11(b) that under the same conditions, when h≥L ₂ , the item will bypass the obstacle side, effectively avoiding the obstacle.

Claims (9)

1. A material transmission platform, characterized in that it includes a hexagonal module body equipped with omnidirectional wheels, a workbench panel, connectors, telescopic rod brackets, a Wi-Fi module and self-locking rollers, and the workbench panel consists of several six The hexagonal module body is spliced, and the hexagonal module body has a slot for positioning when connecting with the connector. The connector is provided with a gap for positioning when connecting with the hexagonal module body. The workbench panel is connected with a Wi-Fi module on the workbench panel, and self-locking rollers are connected to the bottom of the workbench panel.
2. A material transfer platform according to claim 1, wherein the hexagonal module body includes a module box, a connecting rod, a tray, an RFID module, omnidirectional wheels and mounting holes, and the center of the module box is equipped with an RFID module , There are several regularly arranged installation holes on the surface of the module box, the inner side of the module box is fixedly connected with the top of the connecting rod, the bottom end of the three connecting rods is fixed with a tray, and the bottom of the module box is provided with three omnidirectional wheels.
3. A kinematics modeling strategy and path planning method for a material transfer platform, characterized by its control method, comprising the following steps:

   (1) Adopt the multi-kinematics modeling strategy to complete the establishment of the kinematics model library for control;

   (2) The total number of unit modules configured in the host computer, including the side length of the hexagonal module and the number of X and Y axes, stipulates the material transmission speed;

   (3) Obtain the geographical location information of the transported materials through RFID, and feed it back to the host computer;

   (4) Using the improved ant colony algorithm to plan the global transmission path;

   (5) Using the improved artificial potential field method to plan the local transmission path;

   (6) Realize the dynamic switching of multi-kinematics models on the planned transmission path, and solve the rotational speed and steering corresponding to the relevant omni-directional wheels;

   The material transmission platform is connected to the Internet of Things through the Wi-Fi module, and the entire transmission process of the material can be remotely monitored through the mobile terminal.
4. The control method according to claim 3, characterized in that, in the step (2), the step of establishing a kinematics model under the condition that the adjacent wheels are asymmetrically distributed when the two modules participate in the transmission is:

   (2.1) Select the center point o _(l) to establish the world coordinate system. The x _(l) axis increases from bottom to top, and the y _(l) axis increases from left to right. v _ox(l) means v _{o( l)} The component velocity on the x _(l) axis, v _{oy (l)} represents the component velocity of v _{o (l)} on the y _(l) axis, v _{o (l)} represents the linear velocity of point o _(l) , w _o(l) represents the angular velocity of point o _(l) , θ _(l) represents the angle between v _o(l) and x _(l) axis, where l=1,2,3, represents the number of modules involved in material transmission number;

   (2.2) Establish the velocity components v x(2) of the three omnidirectional wheel velocities v ₁₍₂₎ , v ₂₍₂₎ and v ₃₍₂₎ on the x ₍₂₎ axis and y _{( 2)} axis of the body coordinate system , v _y(2) and the relationship between v _ox(2) and v _ey(2) in the world coordinate system:

   

   

   

   In the formula, r ₍₂₎ represents the vertical distance between v _{y (2)} direction and point o ₍₂₎ , which is a constant;

   (2.3) From v _qy(2) =w _q(2) R(q=1,2,3), calculate the relationship between each wheel speed w _q(2) and angular velocity w _o(2) :

   

   In the formula, R represents the equivalent radius of the omnidirectional wheel;

   (2.4) Formula (4) is written in matrix form, that is, the inverse kinematics equation:

   
5. The control method according to claim 3, characterized in that, in the step (2), a kinematics model is established under the condition that the adjacent wheels are asymmetrically distributed when the three modules participate in the transmission:

   Inverse kinematics equation:

   

   Calculate the speed of point o _(l) based on the rotational speeds of the three omnidirectional wheels to form a positive kinematic equation:

   
6. The control method according to claim 3, wherein the step (4) is specifically:

   (4.1) Environment modeling: The terrain map adopts the 0/1 method of rectangular rasterization, 0 indicates the passable position, 1 indicates the obstacle position, the upper left corner is the starting point position, the lower right corner is the target point position, and a right angle is established Coordinate system: the values of the x-axis and y-axis increase from left to right and from top to bottom, respectively, and the custom grid is 20×20, that is, x=y=i, where i∈[1,20] and i∈ N, and place a generation of ants;

   (4.2) Improve the heuristic function of the ant colony algorithm by introducing the sum of the distance between the current module position and the next module position and the distance between the next module position and the target module position, and adding weight factors λ ₁ and λ ₂ to the heuristic function Improvements are made to make the algorithm search process more targeted and reduce the probability of falling into a local minimum. The specific improvement methods are as follows:

   

   

   In the formula, m is the current module position of the material, n is the next module position where the material may be transmitted, D is the target point position of the material transmission, η′ _mn is the improved distance heuristic function, d _nD is the Euclidean number between n and D Reed distance, and considering the uniqueness of the platform structure, set λ ₁ +λ ₂ = 2, and λ ₁ , λ ₂ >0;

   (4.3) Improve the pheromone inspiration factor and distance expectation function factor of the ant colony algorithm, considering that the pheromone content in the early path is too low, which increases the blindness of the ant search path, and in the later stage, due to the excessive accumulation of pheromone, it shrinks The selectable range of the path causes the algorithm to fall into a local optimum, and the design factor adaptive update strategy, the specific improvement method is as follows:

   

   

   In the formula, K is the total number of iterations of the algorithm, k is the current number of iterations, α is the initial pheromone inspiration factor, β is the initial distance expectation function factor, and u is a constant. Considering the uniqueness of the platform structure, set u=2e /3;

   (4.4) Determine the current feasible module road set, introduce the roulette algorithm to establish and update the optimized probability model between transmission modules, and optimize the probability model

   

   The specific expression is:

   

   In the formula, A _j represents the module set that the material can reach in the next step,

   

   is the pheromone concentration on the optimized transmission path,

   

   is the improved distance heuristic function;

   (4.5) Update the pheromone content on the transmission path, the specific expression is:

   τ _mn (t+1)＝(1-ρ)τ _mn (t)+Δτ _mn (t) (13)

   

   

   In the formula, ρ is the pheromone volatilization coefficient, Δτ _mn is the sum of the pheromones released by two adjacent modules,

   

   is the pheromone increment of two adjacent modules, L _j is the path length of ant j, and Q is the pheromone enhancement coefficient;

   (4.6) Calculate the feasible solution obtained by the current iteration, compare it with the feasible solution obtained by the previous generation, record the optimal solution, and plan the global shortest path for material transmission after the number of iterations is over.
7. The control method according to claim 3, wherein the step (5) is specifically:

   (5.1) First, draw a circle R1 with the object center r as the center of the circle, and the maximum distance _ρ0 of the obstacle action as the radius; construct a straight line l ₁ passing through the obstacle point F closest to r and the starting point O at the same time; record the distance between l ₁ and the x-axis Angle θ; determine the number c of obstacles falling in the circle R1; if c=1, execute step (5.2), if c=i (i=2,3,4...) then execute step (5.3);

   (5.2) The number of obstacles is 1; find the distance L ₁ between the object and F; then draw a circle R2 with r as the center and L ₁ as the radius; rotate l ₁ forward or counterclockwise based on the center r The straight line l ₂ is constructed by the θ angle; the node farther away from F among the intersection points of l ₂ and circle R2 is the fuzzy repulsion point Q;

   (5.3) The number of obstacles is multiple; record the distance L ₂ between the two obstacles closest to the center r of the object, the center point f between them, and the maximum size h of the object outline; at the same time, calculate the distance between F and r The distance L _min ; draw a circle R3 with r as the center and L _min as the radius;

   If h<L ₂ , construct the line l ₃ passing through f and r, and the node farther away from F in the intersection of l ₃ and circle R3 is the fuzzy repulsion point Q;

   If h≥L ₂ , then construct a straight line l ₄ passing through F and r, and use l ₄ as the center of circle r as a reference, and rotate l _{5 in a forward or counterclockwise direction by an angle θ to construct a straight line l 5} ; the intersection point of l ₅ and circle R3 is far away from F The node of is the fuzzy repulsion point Q.
8. A computer storage medium, on which a computer program is stored, characterized in that, when the computer program is executed by a processor, the kinematics modeling oriented material transfer platform as described in any one of claims 3-7 is realized Strategy and path planning methods.
9. A computer device, comprising a memory, a processor, and a computer program stored on the memory and operable on the processor, characterized in that, when the processor executes the computer program, any of claims 3-7 can be realized. A kinematics modeling strategy and path planning method for a material transfer platform described in one item.

Images