A collaborative control method for multiple AGVs that collaboratively carry heavy objects

Abstract

The present invention discloses a cooperative control method for multiple AGVs that cooperate to carry heavy objects, belonging to the field of automated logistics technology. This method uses a high-dimensional environmental description vector as a model condition and inputs it into an improved conditional diffusion model that introduces a linkage constraint feature channel in the reverse denoising stage of the traditional conditional diffusion model to obtain a preliminary solution. Then, using popular theory and high-dimensional differential geometry constraints, the preliminary solution is corrected. Next, according to the corrected solution, dynamic environment perception and linkage state prediction, online feedback and local path adjustment mechanisms, and a multi-overload synchronous control strategy are adopted during control operation to ensure the cooperative control of multiple AGVs. The method of the present invention obtains an initial path based on the improved conditional diffusion model, improving the feasibility and cooperation of the generated results. Moreover, a high-dimensional differential geometry dynamic control method is used to control the vehicle to optimize the vehicle speed, acceleration, and steering angle, ensuring the dynamic consistency of multi-vehicle synchronous movement.

Description

Technical Field

[0001] The present invention relates to the field of automated logistics technology, and in particular to a collaborative control method for multiple AGVs that collaboratively carry heavy objects. Background Art

[0002] The traditional single-vehicle handling method relies on an automated guided vehicle (AGV) to carry materials along a fixed path. The path and control of the material handling by a single vehicle are usually simple, and it is only suitable for carrying small-sized and light-weight materials. In modern manufacturing and logistics scenarios, the handling of large-sized and heavy-weight materials places higher demands on the automation and flexibility of the production line than in existing general scenarios. Traditional handling methods usually rely on single-vehicle handling or manual operation along a fixed path, which is inefficient and difficult to adapt to complex and changing production environments. How to efficiently and safely complete the transportation of large-sized and heavy-weight materials in a limited space has become a problem that needs to be solved in the process of intelligent upgrading of production lines.

[0003] Therefore, in some heavy-load scenarios in the prior art, a multi-vehicle collaborative handling method is adopted to improve the handling efficiency and stability by sharing the weight and optimizing the path, so that multiple automatic guided vehicles work together to jointly undertake the handling task. For multi-vehicle collaborative systems, the existing path planning technology usually calculates the moving path of each vehicle based on a static algorithm to ensure that there is no conflict between vehicles and to carry out collaborative handling. In order to cope with dynamic environmental changes (such as obstacles), dynamic path adjustment and obstacle avoidance technologies are introduced in the prior art. These technologies can change the path in real time to avoid dynamic obstacles and ensure the smooth progress of the handling task. In multi-vehicle collaborative handling, the existing technology adopts a synchronous control strategy to ensure the consistency of the speed, position and direction of movement of each vehicle. The synchronous control is usually achieved through centralized control or distributed control strategies to maintain the stability of the handling. Error detection and correction mechanisms are also introduced in some existing systems, using sensors and feedback systems to monitor vehicle deviations and perform error correction to maintain the accuracy and stability of the handling.

[0004] Heavy-duty automated guided vehicles are gradually being introduced into the handling of large-sized and heavy-weight materials due to their high flexibility and automation characteristics. Compared with traditional single-vehicle handling, multi-vehicle collaborative operation has certain advantages in heavy-duty scenarios. It can improve handling efficiency and system stability by sharing weight, optimizing paths and coordinating movements. However, due to the particularity of large-sized materials, multi-vehicle coordinated handling still faces the following challenges in actual applications:

[0005] Complexity of path planning: In the handling of large-sized materials, due to the large volume and heavy weight of the materials, multiple vehicles need to jointly carry and transport the materials in a coordinated manner. This way makes it impossible for a single vehicle to complete path planning independently, and traditional static path planning algorithms are difficult to adapt to the path planning among multiple vehicles in a highly dynamic and complex handling scenario in an actual dynamic situation that needs to comprehensively consider material size constraints, limitations of the handling space, and potential obstacles in the dynamic environment.

[0006] Synchronization and accuracy of multi-vehicle linkage: When handling large-sized materials, the synchronous coordination among single vehicles is relatively important. If there are deviations in the speed, position, or direction of the vehicles, it may lead to unstable material handling and even cause safety accidents. In addition, the handling of large-sized materials requires the vehicles to achieve high-precision synchronous movement in space, which poses certain requirements for the control and coordination between vehicles.

[0007] Real-time adjustment ability in a dynamic environment: Large-sized and heavy materials usually need to be flexibly handled between production lines, and the dynamic changes in the production environment (such as the appearance of obstacles, road congestion, temporary scheduling changes, etc.) demand real-time response during the handling process. If the multi-vehicle system cannot quickly adjust the path and movement strategy according to the real-time environment, it will lead to the failure of the handling task or a significant decrease in transportation efficiency.

[0008] Error accumulation and reliability of feedback: During the multi-vehicle cooperation process, due to reasons such as sensor errors and communication delays between vehicles, there will be deviations in position or path planning. Especially for large-sized materials, subtle errors may be amplified, thus having a significant impact on the stability of handling. Therefore, how to construct a reliable error correction and feedback mechanism during the handling process is the key to ensuring the stability and accuracy of handling.

[0009] In summary, the problem of designing a processing method that can generate reliable and collaborative path planning results during the joint handling process of multiple heavy-load automatic guided vehicles, and ensure the dynamic balance of synchronization and load distribution of the vehicles during the movement process becomes an urgent problem to be solved in the enterprise production process. Summary of the Invention

[0010] In view of this, the purpose of the present invention is to provide a cooperative control method for multiple AGVs that cooperate to handle heavy objects, so as to solve the problem of the linkage constraints of path planning among multiple vehicles in a highly dynamic and complex handling scenario in an actual dynamic situation that is difficult to handle by existing static methods based on optimization or planning.

[0011] To achieve the above purpose, the method of the present invention proposes an improved conditional diffusion model, introduces a linkage constraint feature channel, and embeds geometric constraints, load limitations, and channel width limitations to enhance the feasibility and collaboration of the generated results. Specifically, the method includes the following steps:

[0012] 1) Take the kinematic characteristic information of multiple AGVs integrating the handling to be performed, the real-time feasible region information of the multiple AGVs and the materials they jointly handle, and the high-dimensional environment description vector of the real-time environment information and the real-time load information of the multiple AGVs collected by sensors as model conditions, and input them into an improved conditional diffusion model to obtain a preliminary plan including the initial paths and action sequences of the multiple AGVs to jointly perform handling.

[0013] The improved conditional diffusion model adds penalty terms for geometric collision of materials, overloading penalty, and plant aisle restriction penalty to the reverse prediction equation in the reverse denoising stage of the traditional conditional diffusion model.

[0014] 2) Use the popular theory to perform an overall continuous description of the initial paths and action sequences, correct the obtained overall continuous description results, obtain the corrected values of the speed, acceleration, and steering angle of each AGV when satisfying the constraint conditions in the manifold space, and determine the motion instructions of each AGV according to the corrected values to perform coordinated motion control of the multiple AGVs.

[0015] The constraint conditions are that the constraint function of the coupled motion with balanced forces between the multiple AGVs and the materials is established, and the high-dimensional differential geometry constraint mapping constraint including multiple constraints such as AGV speed balance, load distribution balance, and material attitude stability is established.

[0016] 3) In the coordinated motion control of the multiple AGVs, perform short-term prediction of the future state according to the real-time detection results of the environmental sensors, evaluate the safety risks during the operation of the multiple AGVs according to the short-term prediction results, and perform different adjustment strategies according to different safety risk assessment results to ensure the safe operation of the multiple AGVs.

[0017] Design a multi-load synchronous control law based on the coupled dynamics equation, and achieve dynamic balance of load distribution by real-time distributing vehicle outputs.

[0018] The method of the present invention has the following advantages: the high-dimensional environment description vector of the method of the present invention not only includes real-time environmental information but also integrates vehicle dynamics characteristics, ensuring the comprehensive integration of dynamic sensor information, and the vector can characterize the collaborative handling state of multiple-carrier automatic guided vehicles and large-sized materials, external environmental interference and geometric feasible domain constraints, and can dynamically reflect the collaborative handling state of multiple vehicles and large-sized materials. In addition, in the present invention, by introducing a linkage constraint feature channel in the reverse denoising stage of the traditional conditional diffusion model, embedding geometric constraints, load restrictions and channel width restrictions, the feasibility and coordination of the generated results are significantly improved. The preliminary scheme obtained based on this improved conditional diffusion model has basic feasibility and global consistency, which is intended to lay the foundation for the subsequent refined control and correction of multi-vehicle collaborative handling. After obtaining the preliminary scheme, the method of the present invention optimizes the vehicle speed, acceleration and steering angle based on popular theory and high-dimensional differential geometry constraint mapping constraints to ensure the dynamic consistency of multi-vehicle synchronous motion. After obtaining the corrected control motion instructions, multiple vehicles are controlled based on this motion instruction, and the vehicle's motion status and environmental conditions are detected in real time during the control. Different strategies are formulated based on different detection results to ensure the safety of vehicle operation while ensuring the vehicle's operating efficiency.

[0019] Based on the above, in step 1), the real-time feasible domain information includes load limit, material size and the overall feasible area of ​​multiple AGVs and materials. When the load of multiple AGVs and the material size are within the feasible domain, the materials can be transported in a linked manner under the current geometry and load constraints.

[0020] In the present invention, geometric constraints are used to ensure that the subsequent construction of the environment description model can accurately describe the feasibility of synchronous transportation between large-sized materials and multiple vehicles.

[0021] Based on the above, in step 1), the reverse prediction equation of the reverse denoising stage of the improved conditional diffusion model is:

[0022] ;

[0023] ;

[0024] in, represents the function that makes predictions based on the current state and conditional inputs in the traditional diffusion model, represents the penalty function for large-size material features, is a hyperparameter that balances the basic denoising objective of the model with the suppression strength of the large-size material linkage constraint feature channel. are the weight coefficients of the three penalty terms respectively, They correspond to the collision detection, load verification and channel width limit verification constraint functions respectively.

[0025] The method of the present invention adds a penalty function for the characteristics of large-sized materials to the reverse prediction equation in the reverse denoising stage of the conditional diffusion model. The penalty function includes penalty terms of collision detection, load verification, and channel width limit verification constraint functions, reducing random disturbances that do not meet the requirements of large-sized collaborative handling, ensuring the feasibility and coordination of the generation results of the generated preliminary scheme, and the obtained preliminary scheme has basic feasibility and global consistency.

[0026] Based on the above, in step 2), the process of correcting the obtained overall continuous description result includes: defining a constraint function for describing the coupled motion on the manifold, constructing a high-dimensional differential geometry constraint mapping for differential geometry control, dynamically correcting the preliminary scheme based on the high-dimensional differential geometry method, and solving the motion correction values of each vehicle when satisfying the constraint conditions in the manifold space.

[0027] Based on the above, in step 2), dynamically correcting the preliminary scheme based on the high-dimensional differential geometry method is: making a projection correction along the negative gradient direction of the constraint function within the manifold to cancel the components that violate synchronous coordination.

[0028] Based on the above, in step 2), the constraint function of the coupled motion is: , where is the time-varying curve of the generalized coordinate vector for the overall continuous description of the initial path and action sequence, represents the time derivative of the generalized coordinate;

[0029] The high-dimensional differential geometry constraint mapping is , is for multiple constraints such as AGV speed balance, load distribution balance, and material attitude stability. When satisfying in the manifold space, the coupled motion of multiple vehicles and materials is in an ideal synchronous balance state;

[0030] The formula for dynamically correcting the preliminary scheme based on the high-dimensional differential geometry method is:

[0031]

[0032] where, , , are the corrected speed, acceleration, and steering angular velocity of each AGV respectively, represents the speed vector estimated according to the preliminary scheme, represents the gradient of the constraint function; represents the corresponding Jacobian matrix of the generalized inverse matrix or projection matrix, represents the step size factor, and is an adjustable coefficient used to allocate the correction amplitude of acceleration and steering angle.

[0033] The method of the present invention uses high-dimensional differential geometry dynamic control for multi-vehicle synchronous motion correction. In the multi-vehicle and material handling scenario, a generalized coordinate vector is constructed based on the manifold theory to continuously describe the force and motion states of multi-vehicle linkage. By using coupled motion constraints and dynamic mapping, the vehicle speed, acceleration, and steering angle are optimized to ensure the dynamic consistency of multi-vehicle synchronous motion.

[0034] Based on the above, in step 3), the environmental sensor is a multi-source environmental sensor, and the real-time detection result of the environmental sensor is to fuse the results of the multi-source environmental sensors through time series alignment and weighted fusion to generate globally consistent dynamic environmental information.

[0035] The method of the present invention synthesizes multi-source sensor data, processes the time series and data accuracy differences of different vehicles through time series alignment and weighted fusion, obtains globally consistent real-time environmental information, ensures the accuracy of the obtained real-time environmental information, and provides an accurate real-time environmental basis for subsequent state prediction and interference pre-judgment.

[0036] Based on the above, in step 3), the short-term prediction of the future state includes:

[0037] Combining the dynamic environmental information with the motion instructions obtained in step 2), estimate the interference sources such as the movement of obstacles, channel blockage, and uneven load.

[0038] Combining the dynamic environmental information with the motion instructions obtained in step 2), measure the deviation between the actual state and the ideal reference state of each AGV at the same moment.

[0039] Based on the above, in step 3), the process of evaluating the safety risks during the operation of multi-AGVs according to the short-term prediction results includes:

[0040] When the estimation result of the interference source is within the safe range and the measurement result of the deviation does not exceed the set threshold, the safety risk level is 0 at this time;

[0041] When the estimation result of the interference source reaches the warning level or the measurement result of the deviation exceeds the set threshold, the safety risk level is 1 at this time;

[0042] When the estimation result of the interference source reaches the emergency level of serious obstacle or unexpected situation, the safety risk level is 2 at this time.

[0043] Based on the above, in step 3), the different adjustment strategies according to different safety risk assessment results are:

[0044] When the safety risk level is 0, only routine online monitoring needs to be maintained;

[0045] When the safety risk level is 1, a moderate correction of the speed or steering angle adjustment is triggered;

[0046] When the safety risk level is 2, a more coercive response strategy of immediate local path replanning or forced docking is carried out.

[0047] In the process of multi-vehicle operation, the present invention conducts real-time detection, adds an interference evaluation process to the prediction model, quantifies interference sources such as obstacles and unbalanced load distribution, and outputs risk indicators to detect in real time whether there are potential safety hazards during the operation of the vehicle. According to the level of the risk situation, no intervention is made when there is no risk, simple intervention adjustment is carried out when there is a small risk situation, and path replanning or control stop strategy is carried out when there are major safety risks, so as to ensure the safe operation of the vehicle while ensuring the operation efficiency of the vehicle. Description of the Drawings

[0048] Figure 1 is a flowchart of the collaborative control method for multi-AGVs that cooperate to carry heavy objects in the present invention.

[0049] Figure 2 is a schematic flowchart of the path and action sequence generation process in the present invention.

[0050] Figure 3 is a flowchart of using high-dimensional differential geometry dynamic control for multi-vehicle synchronous motion correction in the present invention. Detailed Embodiments

[0051] The following will clearly and completely describe the technical solutions of the present invention in combination with specific implementation schemes. However, those skilled in the art should understand that the implementation schemes described below are only used to illustrate the present invention and should not be regarded as limiting the scope of the present invention. Based on the implementation schemes in the present invention, all other implementation schemes obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0052] Embodiment of the Collaborative Control Method for Multi-AGVs that Cooperate to Carry Heavy Objects

[0053] This embodiment proposes a collaborative optimization method for multiple heavy-load automatic guided vehicles for handling large-size and large-weight materials based on a conditional diffusion model and high-dimensional differential geometry dynamic control, to optimize in the handling scenario of large-size and large-weight materials, through multi-vehicle collaborative optimization, to achieve efficient and safe path planning, synchronous control, real-time dynamic adjustment and error feedback mechanism, and meet the handling requirements in complex dynamic environments.

[0054] As Figure 1As shown in the figure, the collaborative control method of this embodiment includes the following steps:

[0055] Step S1: Construct a high-dimensional environment description model for the collaborative handling scenario of multiple heavy-load automated guided vehicles.

[0056] S11: Define the basic coordinates and entity set of the collaborative handling environment for multiple heavy loads.

[0057] This step is mainly to determine the basic reference system and key entities for high-dimensional environment description, providing a coordinate basis and entity recognition basis for subsequent model construction.

[0058] It is established that within the factory operation area, a global coordinate system is established, and the information of all automated guided vehicles to be involved in handling and the materials to be handled is identified and recorded.

[0059] 1. Set a fixed global coordinate system , where is the coordinate origin, and respectively represent the reference axes of the global coordinate system in the plane and vertical directions.

[0060] 2. Let represent the coordinate transformation matrix corresponding to the th automated guided vehicle, and let represent the main geometric center of the large-sized material and its attitude reference coordinates .

[0061] 3. Collect the load capacity limit of each automated guided vehicle, the overall size of the material and the pre-designed plant layout parameters , and store them in the initial environment entity set .

[0062] It is established that through the global coordinate system and entity set, the relative positions and attitudes between different vehicles and materials can be uniformly calculated in subsequent steps, avoiding interference from multiple coordinate definitions.

[0063] S12: Establish the geometric constraint relationship between the large-sized material and multi-vehicle linkage.

[0064] This step is mainly to construct a geometric basis for high-dimensional environment description by considering the size boundary, load limit of the large-sized material and the geometric linkage constraints between multiple vehicles on the premise of the above basic coordinate system and entity set.

[0065] That is, a unified geometric description of the moving boundary of each automated guided vehicle and the outer shape boundary of the material is formed to form a linkage relationship constraint function, and on this basis, a geometric representation of the environmental feasible space is defined.

[0066] 1. Let represent the geometric area occupied by the th automated guided vehicle at time . Let represent the geometric range of the large-sized material at time .

[0067] 2. Denote the overall feasible area of the multi-vehicle and the material as :

[0068]

[0069] where represents the coordinate transformation operation of the th automated guided vehicle relative to the global coordinate system , and represents the coordinate transformation operation of the material relative to the global coordinate system.

[0070] 3. Incorporate the load capacity limit and the material size into the linkage constraint, and define the constraint function :

[0071]

[0072] When , it means that the large-sized material can be transported in a linked manner under the current geometric and load constraints;

[0073] When , it means that the large-sized material can be transported in a linked manner under the current geometric and load constraints;

[0074] When , it means that the collaborative transportation requirements cannot be met.

[0075] Through the geometric constraint relationship, ensure that the subsequent constructed environment description model can accurately depict the synchronous transportation feasibility between the large-sized material and the multi-vehicle.

[0076] S13. Integrate the vehicle kinematic characteristics and online sensor data.

[0077] Based on the vehicle kinematic characteristics and the information collected by the sensors, incorporate the dynamics and real-time environment perception into the environment description to form a high-dimensional scene representation that can reflect the motion states of the vehicle and the material at the current moment.

[0078] For the kinematic characteristics of the speed, acceleration, and steering angle of each automated guided vehicle, combined with the outputs of the environmental sensors and the load sensors, generate high-dimensional data that can be updated in real time:

[0079] 1. Let and respectively represent the velocity and acceleration vectors of the th automated guided vehicle at a certain moment. Let represent the steering angle.

[0080] 2. Denote the multi-sensor data as , which includes the lidar ranging values and the load distribution values collected by the force sensor. Here, it is directly represented in matrix form as:

[0081]

[0082] where represents the environmental distance data vector output by the lidar, represents the load distribution data vector output by the force sensor, and are the corresponding data dimensions.

[0083] 3. Combining the vehicle kinematic parameters and sensor information, define the comprehensive state vector at moment :

[0084]

[0085] Specifically, let represent the comprehensive vector of the real-time sensing data, position, and attitude information corresponding to the material.

[0086] 4. Stack and splice the comprehensive state vectors of all automated guided vehicles and materials to obtain the high-dimensional state matrix :

[0087]

[0088] The matrix contains the kinematic and sensor information of all vehicles and materials at the current moment.

[0089] S14. Generate a high-dimensional environmental description vector.

[0090] Based on geometric constraints, vehicle-material kinematic characteristics, and real-time sensing data, output a high-dimensional environmental description vector of the overall collaborative handling scenario, providing a good and comprehensive feature representation for the next-stage strategy generation and control.

[0091] That is, by fusing the feasible region obtained from S12 and the state matrix obtained from S13 in the global coordinate system, define the environmental description function and output the final vector available for learning or planning.

[0092] 1. Define the environmental description function , whose output is a vector with a dimension of :

[0093]

[0094] 2. Inside the function weighted combination of the geometric attributes of the feasible region and the multi-source information of vehicles and materials in the state matrix is performed to form a mapping:

[0095]

[0096] where, represents the geometric parameter extraction mapping, represents the kinematics and sensor fusion feature mapping, represents the weighting coefficient in the dimension, and

[0097] 3. Stack the above -dimensional output to obtain the high-dimensional environment description vector . This vector can characterize the collaborative handling state of multiple heavy-load automated guided vehicles and large-sized materials at time , external environmental interference, and geometric feasible region constraints.

[0098] That is, through the above steps, the geometric constraints, vehicle kinematic characteristics, and sensor data of large-sized and heavy-weight materials in the multi-vehicle linkage scenario are comprehensively expressed to form a high-dimensional environment description vector.

[0099] Step S2: Generate an initial multi-vehicle collaborative handling plan based on an improved conditional diffusion model, as Figure 2 shown.

[0100] S21: Construct the conditional input and initial random distribution for collaborative handling.

[0101] This sub-step mainly clarifies the conditional input and initial random distribution required for the improved conditional diffusion model to provide a feasible search space for the diffusion process.

[0102] Based on the high-dimensional environment description vector generated in step S1, it is used as the model conditional input, and an initial random noise distribution is set in the path and action sequence space.

[0103] 1. Let represent the high-dimensional environment description vector output at time . This vector synthesizes the real-time feasible region, kinematic characteristics, and sensor information of multiple heavy-load automated guided vehicles and large-sized materials.

[0104] 2. Define the multi-vehicle collaborative handling path and action sequence to be generated as a high-dimensional variable , where represents the total dimension of all position, velocity, attitude, and action execution parameter requirements within a discrete time period or step.

[0105] 3. Let the initial random noise distribution be , adopt the Gaussian distribution probability distribution form, and serve as the starting point of the diffusion process:

[0106]

[0107] 4. Write as conditional information into the conditional channel of the improved diffusion model to form the prior constraint of :

[0108]

[0109] S22. Design a large-size material linkage constraint feature channel and establish an improved diffusion equation.

[0110] In response to the special requirements of large-size material handling, this application proposes to introduce a linkage constraint channel during the denoising process of the traditional diffusion model, so that the generated path and action sequence meet the requirements of material size, load distribution, and plant channel width.

[0111] During the forward diffusion and backward denoising processes, embed a large-size material linkage constraint feature channel, and reduce random disturbances that do not meet the requirements of large-size collaborative handling by introducing penalty functions and constraint functions.

[0112] 1. Let the forward stochastic evolution equation of the standard diffusion process be:

[0113]

[0114] where represents the retention weight of the previous moment state at the th iteration, represents the noise vector sampled from the standard Gaussian distribution, is the total number of steps for forward diffusion.

[0115] 2. In the backward denoising stage, introduce an improved linkage constraint feature channel .

[0116] Then the backward prediction equation is:

[0117]

[0118] where:

[0119] A function that makes predictions based on the current state and conditional input in a traditional diffusion model;

[0120] A penalty function for the characteristics of large-sized materials, reflecting the constraints of material size, load distribution, and plant passage width on the path and action sequence;

[0121] is a hyperparameter that balances the basic denoising objective of the model and the suppression intensity of the characteristic channels of the large-sized material linkage constraints.

[0122] 3. In For large-sized materials, multiple sub-items of geometric collision penalty, overloading penalty, and plant aisle limit penalty are set:

[0123]

[0124] Among them are the weight coefficients of the three penalty terms respectively, corresponding to the collision detection, load verification, and channel width limit verification constraint functions respectively.

[0125] 4. By backpropagating the gradient of in the reverse denoising process, the random perturbations that do not conform to the large-sized collaborative handling are suppressed, so as to gradually approximate the generated result to a realistic and feasible solution that meets the material linkage requirements.

[0126] S23. Execute the diffusion denoising iteration and output the initial handling path and action sequence.

[0127] After introducing the characteristic channels of large-sized material linkage constraints, complete the denoising iteration of the improved diffusion model to generate an initial multi-vehicle collaborative solution that meets the constraints.

[0128] Execute the forward diffusion and reverse denoising steps in sequence, and incorporate the geometric and load penalty terms into the gradient calculation of the prediction function in the reverse denoising stage to obtain the final path and action sequence.

[0129] 1. Utilize the initial distribution in S21 and the conditional input to gradually perform forward diffusion to obtain the intermediate state

[0130] 2. Starting from execute the reverse denoising iteration:

[0131]

[0132] 3. After iterations, obtain the final handling path and action sequence , and rearrange or decode it into path coordinates and action instructions for multiple vehicles at discrete time or critical state moments:

[0133]

[0134] where represents the reference trajectory point sequence of each automated guided vehicle during the entire handling process, represents its corresponding execution action sequence, and the function is used to map the high-dimensional vector into an executable control instruction format.

[0135] S24. Complete the output of the initial cooperation plan and prepare for subsequent dynamic control.

[0136] Output and integrate the preliminary plan after being corrected by the large-size material linkage constraint to form the input basis for subsequent high-dimensional differential geometry dynamic control.

[0137] That is, structurally encapsulate the path and action sequence obtained in the previous step and output it to the control algorithm of the next stage.

[0138] 1. Index the position, attitude, speed, acceleration, and steering information of multiple vehicles included in and record it in the multi-vehicle collaborative handling plan set:

[0139]

[0140] where and respectively represent the path points and action instruction sequences of the th automated guided vehicle within the handling cycle .

[0141] 2. Output the preliminary plan for multi-vehicle cooperation as the input for subsequent high-dimensional differential geometry dynamic control and subsequent dynamic environment perception and replanning.

[0142] 3. The preliminary plan already has basic feasibility and global consistency under the large-size material linkage constraint, laying a foundation for the subsequent refined control and correction of multi-vehicle collaborative handling.

[0143] Step S3. Use high-dimensional differential geometry dynamic control to perform multi-vehicle synchronous motion correction, as shown in Figure 3 .

[0144] S31. Establish a manifold space model of multi-vehicles and materials.

[0145] This step mainly aims to express the coupled motion characteristics of multi-vehicles carrying large-size materials in a high-dimensional space and form a manifold structure for differential geometry control.

[0146] Regarding the preliminary coordination plan output in step S2 as a discretized approximate solution of the system state, based on this, define the generalized coordinates of the system, and use manifold theory to continuously describe the overall feasible states of multiple vehicles and materials.

[0147] 1. Let the generalized coordinate vector of the system be:

[0148]

[0149] where represents the pose (position and orientation) and additional dynamic parameters of the th automatic guided vehicle at time , and represents the key states such as the central pose and force distribution of the material.

[0150] 2. Define the feasible region where is located as the manifold , where:

[0151]

[0152] is the number of automatic guided vehicles, is the dimension of the single vehicle state, and is the dimension of the material state.

[0153] 3. Based on the path and action sequence in the preliminary plan generated in step S2, construct a preliminary discretized mapping of as the input for the next coupled dynamics correction:

[0154]

[0155] The function is used to decode the discrete instructions of multi-vehicle pose, speed, and acceleration into a time-varying curve of the generalized coordinate vector.

[0156] S32. Construct a coupled motion constraint function and introduce a high-dimensional differential geometry constraint mapping.

[0157] Constrain the force and motion coupling relationship during the multi-vehicle coordinated handling of large-sized materials to ensure the coordination consistency between vehicles and between vehicles and materials.

[0158] Define a constraint function describing the coupled motion on the manifold , construct a dynamic mapping for differential geometry control to achieve the correction of speed, acceleration, and steering angle.

[0159] 1. On the manifold Define the coupled motion constraint function within:

[0160]

[0161] Reflect the force balance, geometric linkage, and motion consistency between the vehicle and the material, Denote the time derivative of the generalized coordinates.

[0162] 2. Unify the kinematic equations of each automated guided vehicle and the dynamic equations of the material:

[0163]

[0164] Where Denotes the control input of the th automated guided vehicle, Denotes the external adjustment of the material's attitude or mechanical state (such as synchronous lifting at multiple contact points, force distribution during turning, etc.).

[0165] 3. Combine all the subsystem equations in the manifold to obtain the overall evolution equation of the system:

[0166]

[0167] Where:

[0168]

[0169] Is the overall dynamic mapping of the coupling of multiple vehicles and the material.

[0170] 4. Define the high-dimensional differential geometry constraint mapping , which includes multiple constraints on vehicle speed balance, load distribution balance, and material attitude stability:

[0171]

[0172] Such that when is satisfied within the manifold, the coupled motion of multiple vehicles and the material is in an ideal synchronous equilibrium state.

[0173] S33. Perform differential geometry dynamic correction and output the corrected speed, acceleration, and steering angle.

[0174] Based on the high-dimensional differential geometry method, perform dynamic correction on the preliminary scheme, solve the motion correction values of each vehicle when satisfying the constraint conditions in the manifold space, and ensure the motion consistency of multi-vehicle synchronous handling.

[0175] Propose and adopt the motion constraint projection within the manifold, correct the speed, acceleration, and steering angle of each vehicle, and output the correction solution that satisfies .

[0176] For the coupled constraint equations in step S32 and the differential geometry constraints , at the current moment correct the generalized coordinates and velocities under the preliminary scheme:

[0177]

[0178] where: represents the velocity vector deduced according to the preliminary scheme ; represents the gradient of the constraint function; represents the generalized inverse matrix or projection matrix of the corresponding Jacobian matrix ; represents the step size factor.

[0179] That is to say, it is equivalent to making a projection correction along the negative gradient direction of the constraint function within the manifold to cancel the components that violate the synchronization and coordination.

[0180] Similarly correct the acceleration and steering angle of each vehicle:

[0181]

[0182] where and are adjustable coefficients used to allocate the correction amplitudes of the acceleration and steering angle.

[0183] After the above dynamic correction, at time the velocity, acceleration and steering angle correction amounts that satisfy the high-dimensional differential geometry coupling constraints for multiple vehicles can be obtained, further ensuring the vehicle motion consistency.

[0184] S34. Generate the multi-vehicle synchronous motion instructions and output the correction results.

[0185] Output the velocity, acceleration and steering angle control quantities obtained after the differential geometry dynamic correction as the multi-vehicle actual execution synchronous motion instructions, providing a more stable attitude basis for the dynamic environment prediction and replanning in the next stage.

[0186] Based on the corrected kinematic parameters, uniformly construct the synchronous instruction set for multiple vehicles and encapsulate it into a format that can be executed in the actual controller.

[0187] 1. According to the correction results of step S33, define the synchronous motion instructions of each automatic guided vehicle at time :

[0188]

[0189] where Convert the corrected speed, acceleration, steering angle and other quantities into the torque, rotational speed or steering commands required for the underlying execution of the vehicle.

[0190] 2. Aggregate the correction instruction sets of all vehicles during the entire handling cycle :

[0191]

[0192] 3. Output the corrected multi-vehicle synchronous movement commands and the updated system status , ensure the movement consistency of the vehicles during coordinated handling, meet the force and attitude stability requirements of large-sized materials, and provide a stable attitude basis for the dynamic environment prediction and replanning in the next stage.

[0193] Step S4: Perform coordinated handling status prediction and interference pre-judgment based on dynamic environment perception.

[0194] S41: Obtain and fuse multi-source sensor information.

[0195] During the coordinated handling of large-sized materials, collect and fuse the dynamic data of each automated guided vehicle and environmental sensors in real time to provide a real-time environment basis for subsequent status prediction and interference pre-judgment.

[0196] That is, comprehensively fuse the output data of various sensors such as lidar, cameras, force sensors, and ultrasonic rangefinders to perform multi-dimensional fusion on interference sources (such as temporary obstacles, road congestion) and material force information.

[0197] 1. Let the multi-source sensor data collected by each vehicle at time constitute a matrix:

[0198]

[0199] where represents the environmental information vector collected by the th automated guided vehicle, including elements such as obstacle position, channel congestion degree, and mechanical load change.

[0200] 2. For the time synchronization and network delay between automated guided vehicles, adopt the time series alignment and weighted fusion method, and record the fusion result as:

[0201]

[0202] to process the time series and data accuracy differences of different vehicles and output globally consistent environmental perception information.

[0203] S42: Perform coordinated handling status prediction and interference pre-judgment.

[0204] On the basis of real-time fusion of environmental information, short-term prediction is carried out on the future states of vehicles and materials during collaborative handling, potential interference sources are identified, and their impacts on the movement of multiple vehicles are estimated.

[0205] Based on the synchronized motion command corrected in step S3 and the environmental perception data obtained by fusing in S41, the recent state is predicted by a dynamic model, and the interference sources are estimated and calibrated.

[0206] 1. Define the rolling prediction window , and predict the system state at future time based on the dynamic equations of vehicles and materials :

[0207]

[0208] where represents the prediction function based on the multi-vehicle and material coupling model, comprehensively considering the dynamic effects of the corrected control input and environmental information.

[0209] 2. Estimate the interference sources of obstacle movement, passage blockage, and uneven load, and define the interference evaluation function:

[0210]

[0211] which is used to quantitatively evaluate the risks of each interference source based on the current environmental perception and future state prediction, and output the impact index on the linked handling .

[0212] 3. Feed back the predicted state and the interference evaluation results to the subsequent online replanning mechanism and dynamic control link, so that the system can trigger local path adjustment or control parameter correction in a timely manner when potential risks are detected.

[0213] Step S5: Construct an online feedback correction mechanism for multi-vehicle collaboration and perform local path adjustment.

[0214] S51: Continuously monitor the vehicle state deviation and construct an online feedback error function.

[0215] In this sub-step, the system continuously obtains the multi-vehicle synchronized motion command output in step S3 and the environmental perception data generated by fusing in step S4 . Measure the deviation between the actual state of each vehicle at time and the ideal reference state , and define the feedback error function:

[0216]

[0217] wherein represents the norm operation in the appropriately weighted space (combining multi-dimensional information of vehicle speed, acceleration, and steering angle).

[0218] If any exceeds the set threshold , or obvious load imbalance is detected, or the speed limit is exceeded, the system will trigger the online feedback correction process, thereby fine-tuning the motion parameters of the vehicle during operation and bringing the error back within the tolerance range.

[0219] S52. Dynamically trigger and determine the anomaly level based on the interference prediction result.

[0220] In this sub-step, the system receives the interference evaluation and the predicted state given in step S4, and comprehensively evaluates them with the real-time error . A grading function can be defined:

[0221]

[0222] When , the system only needs to maintain regular online monitoring;

[0223] When , a moderate correction is triggered, such as a small-range speed or steering angle adjustment;

[0224] When , it indicates the existence of serious interference or potential danger, and more forceful response strategies such as immediate local path replanning or forced docking are required.

[0225] S53. Constraints and optimization solution for local path replanning.

[0226] Once it is determined that the interference reaches the level that requires path adjustment ( or 2 , the local path replanning algorithm is started in this sub-step to ensure that the vehicle can avoid or reduce the current interference impact as much as possible while maintaining the established global goal. The optimization objective function for local path search is proposed and defined as:

[0227]

[0228] wherein represents the spatio-temporal trajectory of multiple vehicles in the local interval ; is used to measure the obstacle avoidance cost; measures the deviation from the global reference path; represents costs such as energy consumption or mechanical load; is the weight coefficient. By performing minimization within the local feasible region of the optimization algorithm (such as those based on fast search trees, improved algorithms or non - linear optimizers, etc. are all acceptable), the optimal local driving route of the vehicle under the current interference conditions can be obtained.

[0229] S54. Output the re - calibrated motion instruction and form a closed - loop with the online feedback.

[0230] Finally, this sub - step performs differential - geometric dynamic correction on the local replanning result (similarly to the coupling constraints in step S3), and obtains the corrected speed, acceleration, and steering angle control quantities . At the same time, send the new reference state to the underlying controller of each vehicle:

[0231]

[0232] where is the discrete point of the optimal trajectory after local path optimization, is the control signal mapping function, which is used to convert pose and kinematic information into vehicle execution instructions.

[0233] Through this instruction update, the system forms a closed - loop with the aforementioned online feedback correction mechanism: if deviation or interference is detected again, continue to execute the loop process of S51 - S54, so as to achieve timely response and flexible adjustment to various interferences on the entire handling line.

[0234] Step S6. Execute high - precision handling operations based on the multi - overload synchronous control strategy.

[0235] S61. Establish the coupled dynamic equation of multi - overload synchronous control.

[0236] In this sub - step, to uniformly describe the coupled dynamic relationship between multiple vehicles and materials, the overall dynamic equation of the system is defined.

[0237] Let represent the generalized coordinates of multiple vehicles and materials (refer to the manifold definition in step S3), then the dynamic equation with coupling terms can be written as:

[0238]

[0239] where is the mass - inertia matrix, is the Coriolis force and centrifugal force term, is the gravity and other generalized force vectors related to attitude, It represents the total control force / moment vector (including driving, braking, steering moment, etc.) output by multiple vehicles respectively. Through this equation, the coupled motion of the vehicle when carrying large-sized materials can be described as a whole, laying a foundation for the design data of the multi-overload synchronous control law in the next step.

[0240] S62. Design the multi-overload synchronous control law.

[0241] In this sub-step, for the aforementioned coupled dynamics equation, a synchronous control strategy is adopted to ensure the consistency of the forces, speeds, and attitude adjustments of each vehicle to the material. Let represent the desired system trajectory (combining the local path adjustment and online feedback results in step S5), while the actual system state is . Then the synchronization error is defined as:

[0242]

[0243] To achieve fast and stable error convergence, a multi-overload synchronous control law is proposed and designed:

[0244]

[0245] where and are positive definite gain matrices, corresponding to the feedback terms of the position / attitude error and the speed error respectively; and are used to compensate for the coupling and gravity terms in the dynamics.

[0246] Through the above control law, each vehicle will tend to follow the desired trajectory under the action of the resultant force / moment, achieving synchronous load bearing and precise attitude maintenance.

[0247] S63. Allocate the vehicle output in real time and maintain the multi-overload force balance.

[0248] In this sub-step, to prevent extreme load distribution or inconsistent motion among vehicles, the total control force is decomposed in real time to obtain the individual output of each vehicle:

[0249]

[0250] where is the current load ratio coefficient of the th vehicle, satisfying .

[0251] The coefficient can be dynamically adjusted according to the load measurement of the real-time sensor, the vehicle dynamic performance, the steering demand, etc., so that the actual output total resultant force is still the same as that calculated by the synchronous control law Keep consistent, thereby maintaining force balance between multiple loads and avoiding overloading or unbalanced dragging of the bicycle.

[0252] S64. Issue high-precision synchronous action instructions and execute the transport action.

[0253] In this sub-step, the aforementioned multiple load control laws and vehicle output allocation results are mapped to the underlying control system to form directly executable high-precision action instructions.

[0254] make Indicates The actual execution command vector of the vehicle (including the drive motor speed and steering wheel steering angle) is:

[0255]

[0256] in represents the transformation function that converts the target force / torque into actual motor speed, steering angle and possible braking command.

[0257] That is, in the end, each vehicle The handling actions are performed synchronously, and under the constraints of multiple load synchronization control strategies, the precise loading and stable movement of large-sized materials are continuously maintained, thereby achieving the overall high-precision handling goal.

[0258] Compared with the prior art, the method of this embodiment has the following advantages:

[0259] 1. This embodiment is based on the construction method of the global coordinate system and entity set, defines the geometric constraint relationship and dynamic state representation between multiple vehicles and large-sized materials, and forms a unified high-dimensional environment description vector. A real-time fusion strategy of multi-source sensor data (such as lidar, force sensor) is introduced to generate a comprehensive state matrix in conjunction with the vehicle kinematic characteristics. Compared with the prior art, the environmental modeling in multi-vehicle collaborative handling is usually limited to geometric constraints or simple kinematic features, and lacks comprehensive integration of dynamic sensor information. The method of this embodiment constructs a real-time updated high-dimensional environment description vector through the global coordinate system, multi-source sensor data fusion (such as lidar, force sensor) and vehicle kinematic characteristics, which can dynamically reflect the collaborative handling status of multiple vehicles and large-sized materials, and ensure the comprehensiveness and real-time nature of the high-dimensional environment description.

[0260] 2. This embodiment adopts an improved traditional diffusion model, introduces a linkage constraint feature channel, and dynamically adjusts the diffusion process through penalty functions (geometric collision, overload, channel width limitation). A diffusion denoising mechanism with constrained optimization is defined to gradually guide the random perturbation to a path and action sequence that meets the requirements of collaborative handling of large-sized materials. Compared with the existing technology where the generation of path and action sequences often adopts static methods based on optimization or planning and is difficult to handle complex linkage constraints, especially in the scenario of handling large-sized materials, the method of this embodiment significantly improves the feasibility and coordination of the generation results by proposing an improved conditional diffusion model, introducing a linkage constraint feature channel, and embedding geometric constraints, load limitations, and channel width limitations.

[0261] 3. In the scenario of multi-vehicle and material handling, this embodiment constructs a generalized coordinate vector based on manifold theory to continuously describe the force and motion states of multi-vehicle linkage. Using coupled motion constraints and dynamic mapping, the vehicle speed, acceleration, and steering angle are optimized to ensure the dynamic consistency of multi-vehicle synchronous motion. Compared with the existing technology where multi-vehicle cooperative control mostly adopts simple feedback control strategies and is difficult to comprehensively describe the complex force coupling relationship between multi-vehicles, the method of this embodiment continuously describes the motion states of multi-vehicles and materials through manifold theory, and combines coupled constraints and dynamic mapping to achieve precise correction of speed, acceleration, and steering angle, ensuring the dynamic consistency of multi-vehicle synchronous motion.

[0262] 4. This embodiment synthesizes multi-source sensor data, obtains globally consistent real-time environment information through time series alignment and weighted fusion. An interference evaluation function is added to the rolling prediction model to quantify interference sources such as obstacles and unbalanced load distribution, and output risk indicators. Compared with the existing technology where environmental perception is mostly limited to single-sensor data and the ability to identify interference sources and predict states is limited, the method of this embodiment fuses multi-source sensor data, generates globally consistent dynamic environment information through time series alignment and weighted fusion, and cooperates with the rolling prediction model and interference evaluation function to achieve accurate quantification and prediction of interference sources such as obstacles and load imbalance.

[0263] 5. This embodiment defines a feedback error function by monitoring the error between the vehicle and the ideal state in real time, and dynamically triggers multi-level interference responses. A local path optimization algorithm is introduced to perform local path replanning based on space-time constraints while ensuring global goal consistency. Compared with the existing technology where path adjustment is mostly offline, with slow response speed and difficult to dynamically adapt to environmental changes, the method of this embodiment dynamically triggers local path adjustment according to the interference level through real-time error monitoring and hierarchical feedback mechanism, and uses an optimization algorithm to quickly generate a locally optimal path to ensure that the system has high flexibility and robustness.

[0264] 6. This embodiment proposes a coupled dynamics model for multi-vehicle collaboration, dynamically coupling the motion between vehicles and the forces on large-sized materials. A synchronous control law and an output allocation mechanism are designed to ensure the synchronization of vehicles and the dynamic balance of load distribution during motion. Compared with the prior art where mechanical distribution between multiple vehicles mostly uses empirical rules or equal distribution strategies and is difficult to handle complex dynamic force scenarios, the method of this embodiment designs a multi-load synchronous control law based on coupled dynamics equations and realizes the dynamic balance of load distribution by allocating vehicle outputs in real time, avoiding problems such as single-vehicle overload or towing imbalance.

[0265] 7. This embodiment takes into account the special optimization for large-sized material handling. Compared with the prior art that fails to fully consider special constraint conditions such as the size boundary, load distribution, and plant passage width of large-sized materials, the method of this embodiment introduces multiple constraints in the diffusion model and differential geometry control, especially geometric collision penalty, load limit verification, and plant passage limit verification, and optimizes path generation and synchronous control specifically.

[0266] This method also has the following commercial value:

[0267] 1. Wide application scope and large market demand:

[0268] It can be widely applied to multiple scenarios such as intelligent manufacturing, logistics warehousing, port loading and unloading, especially scenarios that require collaborative handling of large-sized and heavy materials, such as automobile manufacturing, aerospace assembly, semiconductor equipment handling, etc.

[0269] With the deep promotion of Industry 4.0, the market demand for multi-vehicle collaborative handling and automated logistics systems is growing continuously. The method of this invention empowers traditional logistics and manufacturing industries through intelligent and precise collaborative control.

[0270] 2. Significantly reduce operating costs:

[0271] Through efficient path planning, dynamic control, and real-time feedback, this technology reduces energy waste and the risk of misoperation in multi-vehicle collaborative handling.

[0272] Improve equipment utilization rate, reduce dependence on manual intervention, and greatly reduce the labor and time costs of enterprises in the process of warehousing and handling.

[0273] 3. Improve safety and reliability:

[0274] Introduce multi-source data fusion and dynamic interference prediction, perceive environmental changes in real time, actively avoid potential risks, and improve the safety of the handling process.

[0275] Specific optimization for large-sized materials reduces the risk of accidents such as dropping and collision during material handling, and improves the overall reliability of the handling system.

[0276] 4. Meet the requirements of complex scenarios and improve system flexibility:

[0277] Through the linkage constraint feature channel and high-dimensional differential geometry dynamic control, it supports the refined operation of multi-vehicle collaborative handling in complex scenarios, especially the high-precision handling requirements of large-size materials.

[0278] Dynamically adjust the path and control strategy to quickly adapt to dynamic requirements such as production line changes and environmental changes, and enhance the flexibility of the logistics and handling system.

[0279] Although the present invention has been described in detail with general descriptions and specific embodiments above, based on the present invention, some modifications or improvements can be made, which are obvious to those skilled in the art. Therefore, these modifications or improvements made without departing from the spirit of the present invention all fall within the scope of protection required by the present invention.

Claims

1. A method for cooperatively controlling multiple AGVs for cooperatively transporting heavy objects, characterized in that: The steps include: 1) The kinematic characteristic information of multiple AGVs to be handled, the real-time feasible domain information of the multiple AGVs and the materials they jointly handle, and the high-dimensional environment description vector of the real-time environment information collected by sensors and the real-time load information of multiple AGVs are input as model conditions into the improved conditional diffusion model to obtain a preliminary plan including the initial path and action sequence of the multiple AGVs to be jointly handled; The improved conditional diffusion model is to add geometric collision penalty, overload penalty and factory aisle restriction penalty items for materials to the reverse prediction equation of the reverse denoising stage of the traditional conditional diffusion model; 2) Using the popular theory to describe the initial path and action sequence as a whole, and correcting the result of the whole continuous description, the correction value of the speed, acceleration and steering angle of each AGV when the constraint conditions are met in the manifold space is obtained, and the motion command of each AGV is determined according to the correction value to perform collaborative motion control on multiple AGVs; The constraint conditions are that the constraint function of the coupled motion of the force balance between the multiple AGVs and the materials is established, and the high-dimensional differential geometry constraint mapping constraint including multiple constraints on the AGV speed balance, load distribution balance and material posture stability is established; 3) In the collaborative motion control of multiple AGVs, the future state is predicted in the short term based on the real-time detection results of the environmental sensors, and the safety risks in the operation of multiple AGVs are evaluated based on the short-term prediction results. Different adjustment strategies are implemented according to the different safety risk assessment results to ensure the safety of multiple AGV operations; The multi-load synchronization control law is designed based on the coupled dynamic equations, and the dynamic balance of load distribution is achieved by allocating vehicle outputs in real time.

2. The collaborative control method of multiple AGVs according to claim 1, characterized in that: In step 1), the real-time feasible domain information includes load limit, material size, and the overall feasible area of ​​multiple AGVs and materials. When the load of multiple AGVs and the material size are within the feasible domain, the materials can be transported in a linked manner under the current geometry and load constraints.

3. The collaborative control method of multiple AGVs according to claim 1, characterized in that: In step 1), the reverse prediction equation of the reverse denoising stage of the improved conditional diffusion model is: ； ； in, represents the function that makes predictions based on the current state and conditional inputs in the traditional diffusion model, represents the penalty function for large-size material features, is a hyperparameter that balances the basic denoising objective of the model with the suppression strength of the large-size material linkage constraint feature channel. are the weight coefficients of the three penalty terms respectively, They correspond to the collision detection, load verification and channel width limit verification constraint functions respectively.

4. The collaborative control method of multiple AGVs according to claim 1, characterized in that: In step 2), the process of correcting the obtained overall continuous description result includes: defining a constraint function describing the coupled motion on the manifold, constructing a high-dimensional differential geometry constraint mapping for differential geometry control, dynamically correcting the preliminary plan based on the high-dimensional differential geometry method, and solving the motion correction value of each vehicle when the constraint conditions are met on the manifold space.

5. The collaborative control method of multiple AGVs according to claim 4, characterized in that: In step 2), the preliminary plan is dynamically corrected based on the high-dimensional differential geometry method: a projection correction is made along the negative gradient direction of the constraint function within the manifold to offset the components that violate synchronization and coordination.

6. The collaborative control method of multiple AGVs according to claim 5, characterized in that: In step 2), the constraint function of the coupled motion is: ,in is a time-varying curve of a generalized coordinate vector that describes the initial path and action sequence in an overall continuous manner, represents the time derivative of generalized coordinates; High-dimensional differential geometry constraint mapping for , In order to balance the AGV speed, load distribution and material posture, multiple constraints are satisfied in the manifold space. When , the coupled motion of multiple vehicles and materials is in an ideal synchronous equilibrium state; The dynamic correction formula for the preliminary scheme based on high-dimensional differential geometry method is: ； in, , , They are the corrected speed, acceleration and steering angular velocity of each AGV. represents the velocity vector calculated according to the preliminary scheme, represents the gradient of the constraint function; represents the corresponding Jacobian matrix The generalized inverse matrix or projection matrix of represents the step size factor, and It is an adjustable coefficient used to distribute the correction amplitude of acceleration and steering angle.

7. The collaborative control method of multiple AGVs according to claim 1, characterized in that: In step 3), the environmental sensor is a multi-source environmental sensor, and the real-time detection result of the environmental sensor is generated by fusing the results of the multi-source environmental sensors through time alignment and weighted fusion to generate globally consistent dynamic environmental information.

8. The collaborative control method of multiple AGVs according to claim 7, characterized in that: In step 3), short-term predictions of future states include: Combine the dynamic environment information with the motion instructions obtained in step 2) to estimate the interference sources of obstacle movement, channel blockage and uneven load; Combining the dynamic environment information with the motion instructions obtained in step 2), the deviation between the actual state of each AGV at the same time and the ideal reference state is measured.

9. The method for cooperative control of multiple AGVs according to claim 8, characterized in that: In step 3), the process of evaluating the safety risks during the operation of multiple AGVs based on the short-term prediction results includes: When the estimated result of the interference source is within the safe range and the measurement result of the deviation does not exceed the set threshold, the security risk level is 0; When the estimated result of the interference source reaches the warning level or the measurement result of the deviation exceeds the set threshold, the security risk level is 1; When the estimated result of the interference source reaches the emergency level of severe obstacles or emergencies, the security risk level is 2.

10. The collaborative control method of multiple AGVs according to claim 9, characterized in that: In step 3), different adjustment strategies are implemented according to different security risk assessment results: When the security risk level is 0, only regular online monitoring needs to be maintained; When the safety risk level is 1, a moderate correction of the speed or steering angle adjustment is triggered; When the safety risk level is 2, a more stringent response strategy of local route replanning or forced stopping is immediately implemented.

Images