Amr collaborative handling method and device based on multi-target cooperation, and storage medium

Abstract

The application relates to an AMR collaborative carrying method and device based on multi-target cooperation and a storage medium. The method comprises the following steps: based on acquired weight data of goods to be carried, shape data of the goods and bearing data of an AMR, the position data of multiple AMRs participating in collaborative carrying is calculated; the collision avoidance radius of each AMR participating in collaborative carrying is calculated as a whole, a collaborative operation path comprising multiple discrete path points is obtained through path searching and node reduction; for the collaborative operation path, a time-space joint parameterized trajectory is obtained based on a high-order Bezier curve; a master-slave relationship between the AMRs is established based on a chain master-slave rule, the time-space joint parameterized trajectory is taken as the trajectory of a virtual leader, and collaborative chain master-slave tracking control is used to realize AMR collaborative carrying. Compared with the prior art, the application has the advantages of effectively balancing spatial smoothness and time efficiency, high trajectory planning efficiency and the like.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of multi-agent, in particular to an AMR (Automated Mobile Robot) collaborative carrying method, device and storage medium based on multi-target cooperation. BACKGROUND

[0002] Multi-robot systems have attracted extensive attention in recent years, and have better flexibility and reliability compared with single robot systems. Multi-robot collaborative carrying is one of the important applications, aiming to solve the carrying task of larger or heavier objects that cannot be completed by a single robot, and thus plays an important role in production, manufacturing, warehousing and logistics.

[0003] At present, in actual industrial scenarios, the number of robots performing collaborative carrying and the cooperative formation are often fixed, so that the collaborative carrying task can only be performed in a specific scenario for specific goods, and the use scenario is very limited. In addition, most multi-robot collaborative carrying adopts discontinuous segmented trajectories for movement, stops at the reference point and adjusts, and then performs the next segment of action, which not only sacrifices the time efficiency, but also increases the risk of goods deviation or even falling due to uneven movement. Therefore, for the actual execution trajectory of collaborative carrying, smoothness and time efficiency are two important indicators, but smoothness and time consumption are two conflicting indicators, that is, pursuing higher smoothness means slower movement of the robot, increasing time consumption, and vice versa.

[0004] The existing trajectory planning method usually adopts a method of decoupling space characteristics and time characteristics for planning, determines the space path and then plans in the time dimension. This method of decoupling space and time fundamentally separates the space smoothness indicator and the time efficiency indicator, making it difficult to balance between the two.

[0005] In summary, there is currently a lack of a collaborative carrying method to solve or partially solve the problem of balancing space smoothness and time efficiency. SUMMARY

[0006] The present application is to overcome the defects of the prior art and provide an AMR collaborative carrying method, device and storage medium based on multi-target cooperation to solve or partially solve the problem of balancing space smoothness and time efficiency.

[0007] The object of the present application can be achieved by the following technical solutions:

[0008] In one aspect of the present application, an AMR collaborative carrying method based on multi-target cooperation is provided, comprising the following steps:

[0009] Based on the weight data, shape, and load-bearing data of the goods to be transported, the location data of multiple AMRs participating in the collaborative transport are calculated.

[0010] The collision avoidance radius of each AMR participating in the collaborative transport is calculated as a whole, and a collaborative operation path including multiple discrete path points is obtained through path search and node reduction.

[0011] For the aforementioned collaborative operation path, a spatiotemporally joint parameterized trajectory is obtained based on a higher-order Bézier curve;

[0012] The master-slave relationship between AMRs is established based on the chain master-slave rule. The spatiotemporally parameterized trajectory is used as the trajectory of the virtual leader. Collaborative chain master-slave tracking control enables collaborative transport of AMRs.

[0013] As a preferred technical solution, the multi-objective optimization problem is:

[0014] The path points in the collaborative operation path are connected sequentially to obtain the initial segmented uniform velocity trajectory of the entire transportation system. Near the reference points other than the starting point and the ending point, a spatiotemporally parameterized trajectory is obtained by solving a multi-objective optimization problem based on a high-order Bézier curve. The collaborative operation path is then transformed into a continuous smooth trajectory, achieving simultaneous planning of the spatial smoothness and time efficiency of the trajectory.

[0015] As a preferred technical solution, the multi-objective optimization problem is:

[0016]

[0017] st v b,k ≤v M

[0018] α b,k ≤α M

[0019] ω b,k ≤ω M

[0020]

[0021] Among them, T k To complete the spatiotemporal joint parameterization of trajectory B at reference point k k The time taken by K(t), k () is the curvature function, T k,z T k+1,z They are respectively with B k (t) the time taken by the adjacent uniform linear trajectory, and B k (t) represents the time taken for the adjacent uniform linear trajectory, v b,k αb,k , ω b,k are the velocity, angular velocity and acceleration of trajectory B k (t), respectively, v M , α M , ω M are the maximum velocity, maximum angular velocity and maximum acceleration, respectively, O new denotes the set of obstacle positions after safety inflation.

[0022] As a preferred technical solution, the position data of the plurality of AMRs participating in cooperative carrying is calculated by solving the following problem:

[0023]

[0024] ng≥G,

[0025]

[0026] A Conv ≥M

[0027] wherein n, P1,…P n are the number and positions of AMRs participating in cooperative carrying, A max , A min are the maximum intersection area and minimum intersection area, respectively, x R,i , y R,i , r are the horizontal coordinate, vertical coordinate and radius of the pallet of the i-th AMR, respectively, g represents the carrying weight of the AMR, G represents the weight of the goods to be carried, C i represents the effective carrying range of the i-th AMR, S represents the spatial set of the position of the goods, A Conv represents the area of the minimum convex hull of the coordinate set of all AMR pallet centers, and M represents the maximum area of the horizontal projection of the goods.

[0028] As a preferred technical solution, the process of path searching comprises the following steps:

[0029] Based on the position data of the plurality of AMRs, a safety collision avoidance radius is obtained by calculating the starting formation center position;

[0030] Based on the safety collision avoidance radius, the obstacles are subjected to safety inflation processing to obtain an updated coordinate set of the obstacles;

[0031] Based on the coordinate set, the starting formation center position and the end position, an initial discrete reference path is obtained based on the JPS algorithm to realize path searching.

[0032] As a preferred technical solution, the process of node reduction comprises the following steps:

[0033] For the initial discrete reference path obtained by path search, whether two nodes are feasible region is verified by using the convexity of function through iteration cycle, and then node reduction is realized

[0034] As a preferred technical scheme, the cooperative chain master-slave tracking control process comprises the following steps:

[0035] The tracking error of each AMR in the global coordinate system is calculated and converted into the AMR own coordinate system, and the linear velocity and angular velocity of the AMR are controlled based on the preset control law.

[0036] As a preferred technical scheme, the control law is:

[0037] υ i,c =υ i-1 +γ1Γ x,i

[0038] ω i,c =ω i-1 +γ2Γ θ,i +υ i-1 γ3Γ y,i sinΓ θ,i / Γ θ,i

[0039] Wherein, υ i,c , ω i,c are the input linear velocity and input angular velocity of the i-th AMR, γ1, γ2, γ3 are weight coefficients, Γ p,i =[Γ x,i Γ y,i ] T Indicates the position error of the i-th AMR, and Γ θ,i Indicates the orientation error of the i-th AMR.

[0040] Another aspect of the application provides an electronic device, comprising: one or more processors and a memory, the memory having one or more programs stored therein, the one or more programs comprising instructions for executing the aforementioned AMR cooperative carrying method based on multi-target cooperation.

[0041] Another aspect of the application provides a computer-readable storage medium comprising one or more programs for execution by one or more processors of an electronic device, the one or more programs comprising instructions for executing the aforementioned AMR cooperative carrying method based on multi-target cooperation.

[0042] Compared with the prior art, the application has at least one of the following beneficial effects:

[0043] (1) Achieve effective balance between spatial smoothness and time efficiency: The present application is aimed at cooperative operation path, based on high-order Bezier curve to get space-time joint parameterization trajectory, by adopting space-time joint parameterization curve design, using the characteristics of parameterization curve, designing the parameterization trajectory expression of the fusion of space characteristics and time characteristics, the spatial smoothness and time efficiency of the trajectory are planned at the same time, and the spatial smoothness and time efficiency are balanced.

[0044] (2) High efficiency of planning trajectory: The present application obtains the cooperative operation path including multiple discrete path points through path search and node reduction, through adopting node reduction mechanism, the complexity of reference path can be simplified on the basis of ensuring the safety of cooperative handling, thereby reducing the difficulty of planning trajectory, which has good practical application significance.

[0045] (3) The formation remains unchanged in cooperative handling: The present application realizes AMR cooperative handling through cooperative chain master-slave tracking control, utilizes chain total slave tracking principle to convert the cooperative group formed by multiple AMRs into chain structure, adopts two-by-two master-slave allocation method, respectively designs control law for each AMR, so that the overall cooperative formation remains unchanged. The cooperative chain master-slave tracking control decomposes the overall cooperative system into multiple subsystems, so that each subsystem can be independently designed and controlled. This modular design simplifies the analysis and design process of complex systems, reduces the consumption of computing resources, and improves the maintainability and scalability of the system.

[0046] (4) Strong applicability: The method can be applied to different actual working scenes according to different cargo characteristics and task demand changes. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 Flowchart of the AMR cooperative handling method based on multi-target cooperation in the embodiment. DETAILED DESCRIPTION

[0048] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the scope of protection of the present application.

[0049] Embodiment 1

[0050] To solve the problems in the prior art, the embodiment provides an AMR cooperative carrying method based on multi-target cooperation. Trajectory is executed based on spatiotemporal joint parameterization design, and the smoothness and time efficiency of the trajectory are planned, so that the trajectory is smooth enough to ensure the safety of carrying. The method can adaptively adjust the cooperative formation and the number of AMRs required to participate in cooperation according to task requirements and cargo attributes, and plan the smoothness and time efficiency of the trajectory, which are two conflicting targets, so that multiple AMRs can safely, smoothly and efficiently complete the cooperative carrying task in a high-density warehouse environment.

[0051] With reference to Figure 1 , the method comprises the following steps:

[0052] Step S1, determining the optimal cooperative formation.

[0053] Based on the attributes of the goods to be carried and the carrying capacity of the AMR itself, a cooperative allocation optimization problem is constructed for cooperative carrying requirements, and the number of AMRs required to participate in cooperation and the positions of the AMRs forming the cooperative carrying formation are obtained by solving the problem. The specific implementation process of this step is as follows:

[0054] Considering that n AMRs participate in the cooperative carrying task, each AMR has the same specifications, and adopts a lifting carrying method, that is, a tray with a radius of r is installed on the top of each AMR for carrying goods, and the center P of the tray i =[x R,i y R,i ] T is located at the center of the AMR, and also represents the position of the i-th AMR in the map. The rated carrying capacity of each AMR is g. For the goods to be carried, the mass is G, the maximum area on the horizontal plane is M, and the corresponding position set in the two-dimensional map is S.

[0055] For the cooperative carrying task, three conditions need to be met: first, the trays of all AMRs should not overlap or collide, that is:

[0056]

[0057] Second, the total carrying capacity of all AMRs needs to be greater than the mass of the goods itself, that is:

[0058] ng≥G

[0059] Third, the trays of all AMRs participating in the cooperative carrying task should have effective contact with the goods to be carried, and the cooperative area formed by all AMRs should be greater than the maximum area of the goods placement horizontal plane.

[0060] To meet the above requirements, first define the effective bearing range of each AMR as:

[0061] C i = {[x i y i ] | (x i -x R,i ) 2 +(y i -y R,i ) 2 ≤ r 2}

[0062] The condition for the AMR to be in effective contact with the goods is:

[0063]

[0064] And the area of intersection of each AMR with the goods position S can be represented as:

[0065] A i = ∫∫dxdy, [x, y] ∈ C i ∩S, i = 1, … n

[0066] Then, the coordinate set of the center of the pallet is defined as:

[0067] P = {[x1, y1], [x2, y2], …, [x n y n ]}

[0068] The minimum convex hull containing the point set P is obtained by Andrew's Monotone Chain algorithm as C P = Conv(P). Define the vertex set of C P as {[x c,1 y c,1 ], [x c,2 y c,2 ], …, [x c,k y c,k ]}, where k is the number of convex vertices, then the area of the convex hull is:

[0069]

[0070] The following conditions should be met at all times:

[0071] A Conv ≥ M

[0072] In addition, in order to make the collaborative handling task evenly distributed to each AMR. First define the maximum intersection area A max and the minimum intersection area A min as follows

[0073] A max = max(A1, A2,..., An) n )

[0074] A min = min(A1, A2,..., An) n )

[0075] When the difference between them is as small as possible, the goods area carried by each AMR tends to be average.

[0076] In summary, in order to obtain the best cooperative formation and save the number of AMRs, while meeting the above-mentioned conditions, the following optimization problem can be constructed:

[0077]

[0078] ng≥G,

[0079]

[0080] A Conv ≥M

[0081] For the above-constructed optimization problem, a heuristic algorithm (such as particle swarm algorithm) is used to solve it, and the minimum number of AMRs n and the position of each AMR P1,..., Pn that can complete the cooperative carrying task are obtained. The positions of these AMRs together form a cooperative carrying formation.

[0082] Step S2, safe cooperative operation path generation.

[0083] Based on the cooperative carrying formation, all AMRs participating in carrying are regarded as a whole, and its safe collision avoidance radius is calculated, and according to the requirements of the carrying task, a path search algorithm is used and a node reduction principle is designed to generate discrete path reference points (i.e. path points). The specific implementation process of this step is as follows:

[0084] Based on the cooperative formation generated in step S1, the formation center P0 is determined:

[0085]

[0086] Then based on P0, P1,..., Pn, the safe collision avoidance radius δ is determined: n

[0087] δ = max (‖P0P1‖, ‖P0P2‖,..., ‖P0Pn‖) n

[0088] Use δ to make a safe inflation for obstacles in the two-dimensional map:

[0089] ​​​

[0090] where [O x O y ] T ∈O represents the position of obstacles in the environment map, and then the original obstacle information in the environment map is updated to O new = Θ∪O. According to O new , P0, and the target position P N , the initial discrete reference path Φ ini = {Ψ0, Ψ1,..., Ψ N} from the starting point to the target point can be obtained by using the JPS algorithm or other path search algorithms, where Ψ0= P0, Ψ N =P N . Based on Φ ini and O new , the original nodes are further reduced by using Algorithm 1 to generate the simplified discrete reference path Φ out .

[0091] Algorithm 1: Reference path node simplification algorithm for cooperative handling system

[0092]

[0093]

[0094] where Length() represents the length of the variable in (),.Append represents adding the elements in () to the tail, and.Delete represents deleting the elements in () from the original variable. Finally, the simplified discrete path is obtained as where represents the simplified path reference point, and The algorithm verifies whether the two reference nodes are feasible by using the convexity of the function through the iteration loop, and then realizes the simplification.

[0095] Step S3, space-time joint parameterization curve trajectory planning.

[0096] The discrete path reference points are connected in order to obtain the initial segmented uniform velocity trajectory of the handling whole, and the space-time joint parameterization trajectory is designed near the reference points to convert the originally discontinuous polyline trajectory into a continuous smooth trajectory, and the spatial smoothness and time efficiency of the trajectory are planned at the same time to ensure the balance of the two. The specific implementation process of this step is as follows:

[0097] According to Φ out and O new , the space-time joint parameterization curve is used to Φ out except and All other path reference points The trajectory is planned in the vicinity of the group and a smooth trajectory is generated for the entire collaborative group.

[0098] for Its coordinates are [ρ x ρ y ] T First, to The trajectory is defined as a uniform linear trajectory with velocity v. k-1 , to The angle between the line connecting the two points and the horizontal direction is ∈ k Similarly, to The trajectory is defined as a uniform linear trajectory with velocity v. k , to The angle between the line connecting the two points and the horizontal direction is ∈ k+1 Then select the 5th-order Bézier curve B. k As The corresponding parameterized curve to be spatiotemporally coupled is defined as follows:

[0099]

[0100] Where λ represents the path parameters, therefore the above only represents spatial path information and does not involve temporal characteristics; δ u,k u = 0, 1, ..., 5 represents the control points of the current Bézier curve, δ 0,k This indicates the starting point of the trajectory segment, while δ 5,k This indicates the end point of that trajectory segment.

[0101] It should be noted that this embodiment uses a 5th-order Bézier curve for illustration. Other higher-order curves may be used instead, provided there is no conflict.

[0102] Define the completed trajectory B k The time taken is T k To enable simultaneous planning of the spatial and temporal characteristics of the trajectory, the following spatiotemporal joint parameterization curve is designed:

[0103]

[0104] The control points are represented as follows:

[0105]

[0106] in This is the scaling factor. The control points are represented as shown in T above. kThe function, thus yielding the spatiotemporally joint parameterized trajectory B. k (t). Regarding B k (t), whose spatial smoothness can be determined by the curvature K determined by the time consumption. k (T k )express:

[0107]

[0108] And B k The time characteristic corresponding to (t) is T. k B k The dynamics (including velocity v) corresponding to (t) b,k angular velocity ω b,k acceleration α b,k Then, according to the principle of differential flatness, we can obtain:

[0109] v b,k =B k (t)′

[0110] α b,k =B k (t)″

[0111]

[0112] With B k (t) The curvature of consecutive uniform linear trajectories is zero, while their time characteristics are determined by adjacent B... k (t) determines: with B k The time taken by the adjacent uniform linear trajectory (t) is:

[0113]

[0114] And with B k The time taken for the adjacent uniform linear trajectory after (t) is:

[0115]

[0116] In summary, both the spatial and temporal characteristics of the overall trajectory can be expressed as T. k A function of k = 1, 2, ..., m-1, therefore, only T needs to be considered. k Planning allows for the simultaneous consideration of both the spatial smoothness and temporal efficiency of the overall trajectory. The corresponding optimization problem is shown below:

[0117]

[0118] st v b,k ≤ν M

[0119] α b,k ≤α M

[0120] ω b,k ≤ω M

[0121]

[0122] Where v M α M ω M These represent the maximum speed, maximum angular velocity, and maximum acceleration, respectively.

[0123] The aforementioned nonlinear optimization problem can be solved using the IPOPT nonlinear solver to obtain the optimal T. k k = 1, 2, ..., m-1, to be entered into B k (t) can then be obtained from The segmented trajectory corresponding to the spatiotemporal joint parameterization is obtained, and the straight line trajectory connected to each Bézier curve segment is determined based on the beginning and end endpoints of each segment. Finally, the overall spatiotemporal joint parameterization trajectory is obtained and defined as Υ(t).

[0124] Step S4, collaborative chain master-slave tracking control.

[0125] The collaborative group formed by multiple AMRs is transformed into a chain structure. A master-slave allocation method is used, and a control law is designed for each AMR to ensure that the overall collaborative formation remains unchanged. The specific implementation process of this step is as follows:

[0126] From step S1, it is known that the AMRs participating in the cooperative transport are R1, R2, ... R n Then, based on the chain-like master-slave principle, the overall collaborative formation is transformed into a pairwise master-slave formation, that is, R2 is the slave of R1, R3 is the slave of R2, and so on, until all AMRs participating in the collaborative transport are assigned master-slave relationships according to the chain-like master-slave principle. The overall reference trajectory Y(t) generated in step S3 serves as the trajectory of the virtual leader R0, and R0 serves as the master of R1. Therefore, the entire collaborative group will move along Y(t) from the initial position to the target point.

[0127] The expected formation relationship between master and servant The initial positions P0, P1, ..., P determined based on steps S1 and S2 are then... n Decide:

[0128] τ x,i =p x,(i-1) -p x,i

[0129] τ y,i =p y,(i-1) -py,i

[0130] for each R i , i = 1,..., n, the corresponding tracking error in the Cartesian global coordinate system is:

[0131] E i (t) = [E x,i E y,i ] T = Y(t) - R1(t) - τ i , i = 1

[0132] E i (t) = [E x,i E y,i ] T = R i-1 (t) - R i (t) - τ i , i = 2,..., n

[0133] θ E,i (t) = θ i-1 (t) - θ i (t), i = 1,..., n

[0134] where E i (t) represents the AMR position error, θ E,i (t) represents the AMR orientation error, R i (t) represents the real-time position of the i-th AMR during the transport process, and θ i (t) represents the real-time orientation of the i-th AMR during the transport process.

[0135] Since the above tracking error is defined in the Cartesian global coordinate system, in order to design a tracking control law for each AMR, the above error in the global coordinate system is converted to the AMR's own coordinate system through the following transformation.

[0136] Γ p,i = [Γ x,i Γ y,i ] T = [E x,i cosθ i (t) + E y,i sinθ i (t) E y,i cosθ i (t) - E x,i sinθ i (t)] T

[0137] Γ θ,i = θ E,i (t)

[0138] where Γ p,i represents the position error, Γ θ,i represents the orientation error. The derivative of Γ p,i and Γ θ,i gives the dynamic error equation of each AMR itself:

[0139]

[0140] where υ i represents the R i actual linear velocity, ω i represents the R i actual angular velocity.

[0141] Therefore, the control law υ i,c (input linear velocity) ω i,c (input angular velocity) is designed as follows so that the tracking error of each AMR tends to zero, thereby ensuring that the cooperative carrying formation is invariant:

[0142] υ i,c = υ i-1 + γ1Γ x,i

[0143] ω i,c = ω i-1 + γ2Γ θ,i + υ i-1 γ3Γ y,i sinΓ θ,i / Γ θ,i

[0144] where γ1, γ2, γ3 are weight coefficients.

[0145] In summary, the embodiment establishes a complete collaborative handling technical framework including cooperative formation determination, safe cooperative operation path generation, space-time joint parameterized curve trajectory planning, and cooperative chain master-slave tracking control. Based on the collaborative handling method, after the collaborative handling task requirements are determined, the AMR autonomously determines the required number of AMRs and forms a cooperative formation according to the information of the goods to be handled. Then, a simplified discrete path reference point is formed in the environment map using a safety inflation radius. Based on the discrete reference path point, a space-time joint parameterized curve is designed for the cooperative handling group to generate a reference trajectory with high smoothness and low time consumption. Finally, the chain master-slave tracking control method is used to drive each AMR to maintain the cooperative formation unchanged until the target point is reached. The present application determines the cooperative formation in different scenarios by constructing a cooperative optimization problem. On the other hand, by designing a space-time joint parameterized trajectory, the smoothness and time efficiency of the trajectory are planned, enabling multiple AMRs to safely, smoothly, and efficiently complete the collaborative handling task in a high-density warehouse environment. The method has the following beneficial effects:

[0146] (1) The number of AMRs used in the collaborative handling technology and the cooperative formation can be changed according to different goods characteristics and task requirements, suitable for different actual working scenarios.

[0147] (2) The path reference point reduction mechanism can simplify the complexity of the reference path while ensuring the safety of collaborative handling, thereby reducing the difficulty of planning the trajectory and having good practical application significance.

[0148] (3) Based on the space-time joint parameterized curve design, the parameterized trajectory expression is designed by utilizing the characteristics of the parameterized curve, integrating the spatial characteristics and time characteristics, planning the spatial smoothness and time efficiency of the trajectory, and balancing the two.

[0149] (4) Based on the chain master-slave cooperative formation control, the cooperative group formed by multiple AMRs is converted into a chain structure using the chain master-slave tracking principle, and a two-by-two master-slave allocation method is used to design a control law for each AMR, so that the overall cooperative formation remains unchanged.

[0150] Embodiment 2

[0151] The embodiment provides an electronic device, including one or more processors and a memory, the memory storing one or more programs, the one or more programs including instructions for executing the AMR collaborative handling method based on multi-target cooperation as described in embodiment 1.

[0152] Embodiment 3

[0153] The embodiment provides a computer readable storage medium, comprising one or more programs for execution by one or more processors of an electronic device, the one or more programs comprising instructions for performing the AMR collaborative handling method based on multi-target cooperation as described in embodiment 1.

[0154] The above merely provides a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of various equivalent modifications or replacements within the technical range disclosed by the present application, and these modifications or replacements shall be encompassed within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A cooperative transport method for AMRs based on multi-objective collaboration, characterized in that, Includes the following steps: Based on the weight data, shape, and load-bearing data of the goods to be transported, the location data of multiple AMRs participating in the collaborative transport are calculated. The collision avoidance radius of each AMR participating in the collaborative transport is calculated as a whole, and a collaborative operation path including multiple discrete path points is obtained through path search and node reduction. For the aforementioned collaborative operation path, a spatiotemporally joint parameterized trajectory is obtained based on a higher-order Bézier curve; A master-slave relationship is established between AMRs based on a chain-like master-slave principle. The spatiotemporally parameterized trajectory is used as the trajectory of the virtual leader. Collaborative chain-like master-slave tracking control enables collaborative transport by AMRs. The multi-objective optimization problem is: By connecting the path points in the cooperative operation path sequentially, an initial segmented uniform velocity trajectory for the entire transport is obtained. Near reference points other than the starting and ending points, a spatiotemporally parameterized trajectory is obtained by solving a multi-objective optimization problem based on higher-order Bézier curves. This transforms the cooperative operation path into a continuous and smooth trajectory, achieving simultaneous planning of spatial smoothness and temporal efficiency of the trajectory. The multi-objective optimization problem is as follows: in, To complete the reference point The spatiotemporally joint parameterized trajectory at the location The time consumed For curvature function, , They are respectively with The time taken by the previously adjacent uniform linear trajectories, and The time taken for the next adjacent uniform linear trajectory, , , Trajectories Velocity, angular velocity, and acceleration , , These are the maximum speed, maximum angular velocity, and maximum acceleration, respectively. This represents the set of obstacle locations after safety expansion. The location data of multiple AMRs involved in collaborative handling are calculated by solving the following problem: in, , These represent the number and location of the AMRs involved in the collaborative handling. , These are the maximum intersection area and the minimum intersection area, respectively. , , They represent the first The x-coordinate, y-coordinate, and radius of the center of the tray of each AMR. This indicates the load-bearing capacity of the AMR. Indicates the weight of the goods to be transported. Indicates the first The effective load-bearing range of an AMR A spatial set representing the location of goods. Let represent the area of ​​the smallest convex hull of the set of coordinates of all AMR pallet centers. This represents the maximum area of ​​the cargo projected onto a horizontal plane. Each AMR and cargo location The area of ​​the intersection is: The coordinates of the center of the tray are set as follows: The set of contained points is obtained using Andrew's Monotone Chain algorithm. The minimum convex hull is , The set of vertices is ,in Let be the number of vertices of the convex hull, and let its area be: ， The path search process includes the following steps: Based on the position data of the multiple AMRs, the safe collision avoidance radius is obtained by calculating the center position of the initial formation. Based on the safe collision avoidance radius, the obstacle is subjected to a safety expansion process to obtain an updated set of obstacle coordinates; Based on the coordinate set, the initial formation center position, and the endpoint position, an initial discrete reference path is obtained using the JPS algorithm, thus enabling path search. The node deletion process includes the following steps: For the initial discrete reference path obtained from the path search, the convexity of the function is used to verify whether there is a feasible region between two nodes through iterative loops, thereby realizing node reduction. The collaborative chain master-slave tracking control process includes the following steps: The tracking error of each AMR in the global coordinate system is calculated and transformed to the AMR's own coordinate system. Based on a preset control law, the linear velocity and angular velocity of the AMR are controlled. The control law is as follows: in, , The first The input linear velocity and input angular velocity of each AMR , , These are the weighting coefficients. Indicates the first Position error of each AMR Indicates the first Orientation error of each AMR.

2. An electronic device, characterized in that, include: One or more processors and a memory, wherein the memory stores one or more programs, the one or more programs including instructions for executing the multi-target cooperative AMR transport method as described in claim 1.

3. A computer-readable storage medium, characterized in that, It includes one or more programs that are executed by one or more processors of an electronic device, the one or more programs including instructions for executing the multi-target cooperative AMR transport method as described in claim 1.

Images