Cooperative control method and device for uncertain leader
By designing an adaptive observer in a multi-rigid body system, adjusting the system matrix using real and estimated motion parameters, the action consistency problems under unmeasurable disturbances and communication link failures are solved, and the consistency of the action of leaders and followers is achieved. improves the robustness and flexibility of the system.
Patent Information
- Application Number
- CN202410931236.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-11
- Publication Date
- 2024-11-05
AI Technical Summary
In the face of unpredictable external disturbances and communication link failures, it is difficult for the prior art to achieve consistency between the leader and follower actions of multi-rigid body systems, especially because the system matrix contains unknown parameters, which increases the difficulty of the problem.
By designing adaptive observers in followers, using real motion parameters and estimated motion parameters for iterative adjustments, and building distributed observers to achieve leader state estimation and action consistency, relaxing the communication topology and leadership the intellectual requirements of dynamics model.
In the case of communication link failure and uncertain leader dynamics model, the consistency of the leader and follower actions of the multi-rigid body system is achieved, improving the robustness and flexibility of the system.
Smart Images

Figure CN118897574A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computer technology, and in particular to a collaborative control method and device with uncertain leaders. Background Art
[0002] Cooperative control is an important research topic in the control of multi-agent systems (MAS), which has a wide range of applications in many fields, such as attitude alignment of satellites and spacecraft, cooperative control of aircraft, formation flight of unmanned aerial vehicles, etc. The consistency problem of distributed cooperative control is to design communication exchange rules and control laws, specify the information exchange between the two parties, so that all individuals can converge to a unified state.
[0003] In practical applications, the imperfect communication link makes it difficult for the consensus algorithm to cope with the dynamically changing network topology. Although the leader system can assist in tracking and anti-interference, in the face of unpredictable external disturbances, the leader's system matrix should be set to contain unknown parameters. This poses a major challenge to existing designs because the system matrix contains unknown parameters, which increases the difficulty of problem solving. Summary of the invention
[0004] In view of this, an embodiment of the present invention provides a collaborative control method and device with uncertain leader, which can at least solve the problem in the prior art that the communication topology is required to be ideal and relies on the matrix parameters of the leader system.
[0005] To achieve the above object, according to one aspect of an embodiment of the present invention, a collaborative control method with uncertain leader is provided, comprising:
[0006] receiving input motion instructions and sending them to the leader, and when the leader starts to move according to the motion instructions, transmitting the real motion parameters to the observer of the follower in direct communication;
[0007] When each follower imitates the leader's motion, it estimates the leader's motion parameters through the observer and transmits the estimated motion parameters to the observers of other followers in direct communication;
[0008] For the observer of each follower, iteratively adjust the system matrix parameters of the observer according to the real motion parameter, at least one of the motion parameters estimated by other followers, and the motion parameter estimated by each follower;
[0009] The next motion parameter of the leader is estimated using the adjusted system matrix parameters, and the next motion parameter is transmitted to the controller of each follower, so that each follower is controlled by the controller to execute the next motion parameter.
[0010] Optionally, before receiving the input action instruction and issuing it to the leader, the method further includes:
[0011] Receive the rigid body code entered on the rigid body management center registration page, and establish a communication link between the rigid body management center and the registered rigid body;
[0012] Receiving multiple rigid bodies selected from a rigid body collection in a rigid body management center;
[0013] A rigid body selected from the plurality of rigid bodies is received, the rigid body is used as a leader, and the remaining rigid bodies in the plurality of rigid bodies are used as followers.
[0014] Optionally, the receiving the input action instruction and issuing it to the leader includes:
[0015] Receive input action instructions and send them to the leader, so that the leader establishes a communication link with each follower, and each follower establishes a communication link with other followers.
[0016] Optionally, the method further comprises:
[0017] A communication topology is constructed based on the communication links between the leader and followers and between followers. When a direct communication link or an indirect communication link between each follower and the leader in the communication topology is available, an operation of updating the system matrix parameters is performed.
[0018] Optionally, adjusting the system matrix parameters of the observer includes:
[0019] Calculating a global state error between each follower and the leader according to the actual motion parameter and the motion parameter estimated by each follower; or calculating a global state error between each follower and other followers according to the motion parameter estimated by each follower and the motion parameter estimated by other followers;
[0020] According to the global state error calculated in each round, the system matrix parameters of the observer of each follower are updated, and the iterative update operation is repeated until the global state error is less than or equal to a preset threshold, and the updating of the system matrix parameters of the observer of each follower is stopped.
[0021] Optionally, calculating the global state error between each follower and the leader according to the real motion parameter and the motion parameter estimated by each follower includes:
[0022] Subtracting the actual motion parameter from the motion parameter estimated by each follower element by element to obtain an individual error parameter of each follower relative to the leader;
[0023] The attribute of each element is determined, and the individual errors of the elements belonging to the same attribute in the individual error parameters are accumulated to obtain the individual error value of each attribute, and then the global state error between each follower and the leader is obtained by summarizing.
[0024] Optionally, the global state error includes an angular velocity error and a posture error, and the global state error being less than or equal to a preset threshold includes: the angular velocity error being less than or equal to a first preset threshold, and the posture error being less than or equal to a second preset threshold.
[0025] Optionally, the method further comprises:
[0026] According to the global state error of the current round, the update gain of the system matrix parameters is calculated to adjust the training convergence time of the observer through the update gain. Next step motion parameter Next step motion parameter Next step motion parameter
[0027] To achieve the above object, according to another aspect of an embodiment of the present invention, a collaborative control device with an uncertain leader is provided, comprising:
[0028] A receiving module, used for receiving an input motion instruction and sending it to the leader, and transmitting the real motion parameters to the observer of the follower in direct communication when the leader starts to move according to the motion instruction;
[0029] An estimation module, which is used for each follower to estimate the motion parameters of the leader through an observer when imitating the motion of the leader, and transmit the estimated motion parameters to the observers of other followers in direct communication;
[0030] A training module, configured to iteratively adjust a system matrix parameter of the observer for each follower according to the real motion parameter, at least one of the motion parameters estimated by other followers, and the motion parameter estimated by each follower;
[0031] An application module is used to use the adjusted system matrix parameters to estimate the next motion parameters of the leader, and transmit the next motion parameters to the controller of each follower, so as to control each follower to execute the next motion parameters through the controller.
[0032] Optionally, the device further comprises an operation module, configured to:
[0033] Receive the rigid body code entered on the rigid body management center registration page, and establish a communication link between the rigid body management center and the registered rigid body;
[0034] Receiving multiple rigid bodies selected from a rigid body collection in a rigid body management center;
[0035] A rigid body selected from the plurality of rigid bodies is received, the rigid body is used as a leader, and the remaining rigid bodies in the plurality of rigid bodies are used as followers.
[0036] Optionally, the receiving module is used to:
[0037] Receive input action instructions and send them to the leader, so that the leader establishes a communication link with each follower, and each follower establishes a communication link with other followers.
[0038] Optionally, the device further includes a judging module, configured to:
[0039] A communication topology is constructed based on the communication links between the leader and followers and between followers. When a direct communication link or an indirect communication link between each follower and the leader in the communication topology is available, an operation of updating the system matrix parameters is performed.
[0040] Optionally, the training module is used to:
[0041] Calculating a global state error between each follower and the leader according to the actual motion parameter and the motion parameter estimated by each follower; or calculating a global state error between each follower and other followers according to the motion parameter estimated by each follower and the motion parameter estimated by other followers;
[0042] According to the global state error calculated in each round, the system matrix parameters of the observer of each follower are updated, and the iterative update operation is repeated until the global state error is less than or equal to a preset threshold, and the updating of the system matrix parameters of the observer of each follower is stopped.
[0043] Optionally, the training module is used to:
[0044] Subtracting the actual motion parameter from the motion parameter estimated by each follower element by element to obtain an individual error parameter of each follower relative to the leader;
[0045] The attribute of each element is determined, and the individual errors of the elements belonging to the same attribute in the individual error parameters are accumulated to obtain the individual error value of each attribute, and then the global state error between each follower and the leader is obtained by summarizing.
[0046] Optionally, the global state error includes an angular velocity error and a posture error, and the global state error being less than or equal to a preset threshold includes: the angular velocity error being less than or equal to a first preset threshold, and the posture error being less than or equal to a second preset threshold.
[0047] Optionally, the training module is further used to:
[0048] According to the global state error of the current round, the update gain of the system matrix parameters is calculated to adjust the training convergence time of the observer through the update gain.
[0049] To achieve the above objective, according to another aspect of an embodiment of the present invention, a collaborative control electronic device with an uncertain leader is provided.
[0050] An electronic device according to an embodiment of the present invention includes: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement any of the above-mentioned leader-uncertain collaborative control methods.
[0051] To achieve the above objective, according to another aspect of an embodiment of the present invention, a computer-readable medium is provided, on which a computer program is stored, and when the program is executed by a processor, any of the above-mentioned collaborative control methods with uncertain leaders is implemented.
[0052] To achieve the above object, according to another aspect of an embodiment of the present invention, a computer program product is provided. A computer program product of an embodiment of the present invention includes a computer program, and when the program is executed by a processor, the leader-uncertain cooperative control method provided by an embodiment of the present invention is implemented.
[0053] According to the solution provided by the present invention, one embodiment of the above invention has the following advantages or beneficial effects: in the case of communication link failure and unknown leader dynamics model, based on an adaptive strategy, a fully distributed observer is designed in each follower, which can estimate the leader state and output information based on motion parameters, so as to achieve consistency in the actions of the leader and followers of the multi-rigid body system.
[0054] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS
[0055] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention.
[0056] Figure 1 is a main flow chart of a collaborative control method with uncertain leaders according to an embodiment of the present invention;
[0057] Figure 2 It is a schematic diagram of the communication topology constructed by the leader and followers;
[0058] Figure 3 is a flow chart of an optional collaborative control method with uncertain leader according to an embodiment of the present invention;
[0059] Figure 4 It is a schematic diagram of the meaning of the letters involved in this scheme;
[0060] Figure 5 is a schematic diagram of main modules of a collaborative control device with uncertain leader according to an embodiment of the present invention;
[0061] Figure 6 is an exemplary system architecture diagram to which embodiments of the present invention may be applied;
[0062] Figure 7 It is a schematic diagram of the structure of a computer system of a mobile device or a server suitable for implementing an embodiment of the present invention. DETAILED DESCRIPTION
[0063] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.
[0064] It should be noted that the collection, collection, updating, analysis, processing, use, transmission, storage and other aspects of user personal information involved in the technical solution of this disclosure are in compliance with the provisions of relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. Necessary measures are taken for user personal information to prevent illegal access to user personal information data and maintain the security of user personal information, network security and national security.
[0065] Conventional consistency control algorithms usually require that the communication topology of the entire system has a directed spanning tree with the leader as the root node, that the topology does not change during the entire communication process, and that the communication weight is usually assumed to be a constant. In practical applications, due to the existence of communication link failures (delay, quantization error, communication noise, etc.), the existence of the above problems increases the difficulty of specifying communication interaction algorithms. At the same time, under the leader-follower consistency model, there is often an implicit assumption in previous algorithms: the system matrix parameters of the leader are known to its neighboring nodes. This assumption is a trivial assumption in the observer-based multi-rigid body system attitude consistency problem algorithm. In summary, the current mainstream algorithm design is based on an ideal communication link topology and a completely known leader dynamics model.
[0066] The consensus algorithm is to mathematically model the real-world multi-body system (drone formation, unmanned boat formation, or any entity composed of multiple monomers that can interact with the outside world). Based on the theory of automatic control, a control algorithm is designed for each monomer so that the entire group can achieve coordinated behavior, including but not limited to: drone formation navigation, robotic arm collaborative handling of heavy objects, and voltage consistency regulation of distributed microgrids. After mathematically modeling the multi-agent, the entire multi-body system forms a graph structure: each monomer is a node in the graph, and the edges between the nodes represent the communication status between the monomers. For example, a_{ij} represents the communication status between monomer i and monomer j. At present, a_{ij}=0or 1 is generally used to represent the communication between i and j. In other words, a_{ij}=0 means that there is no communication between i and j, and a_{ij}=1 means that there is communication between i and j. Therefore, it is usually assumed that the communication weight is a constant.
[0067] This project involves the field of consistent collaborative control, and studies multi-rigid body systems. Multi-rigid body systems are a specific field of multi-agent systems. The consistency research of multi-agent systems started with the clustering phenomenon of certain organisms in reality, such as flocks of birds migrating in formation, schools of fish swimming in pairs, and ants working together to carry food. They all produce complex collaborative behaviors through simple communication. The purpose of multi-agent consistency research is to enable some simple monomer components (such as drones, etc.) to act collaboratively through consistency algorithms to achieve a complex behavior as a whole.
[0068] The rigid body model is a mathematical modeling of some objects in reality, including but not limited to: drones, unmanned boats, unmanned satellites, etc. This solution mainly involves the consistency of multi-rigid body systems, such as drone formations forming various text patterns. For example, the process of StarLink arranging in a long queue in space is the process of achieving consistency in multi-rigid body systems. Another example is the robotic arms on the factory production line, which are also multi-rigid body systems. The process of them coordinating to move workpieces is a consistency process.
[0069] See also Figure 1 , which shows a main flow chart of a collaborative control method with uncertain leader provided by an embodiment of the present invention, including the following steps:
[0070] S101: receiving an input motion instruction and sending it to the leader, and when the leader starts to move according to the motion instruction, transmitting the real motion parameter to the observer of the follower in direct communication;
[0071] S102: when each follower imitates the movement of the leader, it estimates the movement parameters of the leader through the observer, and transmits the estimated movement parameters to the observers of other followers in direct communication;
[0072] S103: for the observer of each follower, iteratively adjusting the system matrix parameters of the observer according to the real motion parameter, at least one of the motion parameters estimated by other followers, and the motion parameter estimated by each follower;
[0073] S104: using the adjusted system matrix parameters to estimate the next motion parameter of the leader, and transmitting the next motion parameter to the controller of each follower, so that the controller controls each follower to execute the next motion parameter.
[0074] Next step motion parameters Next step motion parameters Next step motion parameters
[0075] In the above implementation, this solution first assumes that the communication failure between rigid bodies meets the model requirements, such as time delay and noise interference, and does not require the communication topology to be ideal. It only requires that both the direct communication link and the indirect communication link established between the leader and the follower are available. This solution is applied to the situation where the leader is uncertain and the communication link failure exists. Leader uncertainty refers to the uncertainty of the leader's dynamic model.
[0076] A dynamic model is a mathematical model of the behavior of a real object. A simple example: for a sine function sin(2x), its complete motion trajectory can be directly obtained, but for sin(ax), a is an unknown number, so its motion trajectory at all times cannot be obtained. A certain leader means that the follower can know the motion trajectory of the leader at any time, because the leader's motion function is already clear. Conversely, an uncertain leader means that the follower can only know the current state of the leader, and cannot obtain the leader's state at other times based on the current state.
[0077] For steps S101 to S104, a rigid body management center is set up, and rigid bodies can be registered in the rigid body management center, such as entering the drone code on the drone management center registration page to register, so as to obtain a drone collection and establish a communication link between the drone and the drone management center. Users can select multiple rigid bodies from the rigid body management center, such as selecting 50 drones from 100 drones, and select one of the 50 drones as a leader, and the remaining 49 drones as followers. It should be noted that this solution does not consider the formation, and the purpose is to keep all rigid bodies consistent.
[0078] The user can also enter actions in the rigid body management center, and then click the send button to send this set of action instructions to the leader, such as a set of radio gymnastics, such as moving east as a whole, moving west as a whole. After receiving the action instructions, the leader triggers an attempt to establish a communication link with the remaining 49 followers, and these 49 followers can also try to establish a communication link with each other. It should be noted that the communication link between the leader and the follower is not required to be a direct communication link, such as AB form (assuming A is the leader), it can be an indirect communication link form, such as ABC, A and C communicate indirectly through B. In the communication topology constructed by the last 50 rigid bodies, when a direct communication link or indirect communication link between each follower and the leader is available, such as AB and ABC mentioned above, the system matrix parameter update operation is performed.
[0079] A distributed adaptive leader state estimator, or observer, is designed in each follower to estimate the leader’s motion parameters, such as angular velocity and attitude. Figure 2 As shown in the figure, A is the leader, BCD are followers, and when leader A starts to move according to the action command, it will send its real action parameters to the observer of follower B and the observer of follower C in direct communication. Since it communicates indirectly with follower D, the real action parameters will not be transmitted to follower D.
[0080] When follower B receives the real motion parameters transmitted by leader A and starts to move, it will also transmit the motion parameters of leader A estimated by its observer to the observers of follower C and follower D that are in direct communication with follower B. Similarly, when follower C receives the real motion parameters transmitted by leader A and starts to move, it will also transmit the motion parameters of leader A estimated by its observer to the observers of follower B and follower D that are in direct communication with follower C. For follower D, after it receives the motion parameters of leader A estimated by followers B and C and starts to move, its observer will also simulate the motion parameters of leader A, and its motion parameters will also be transmitted to the observers of follower B and follower C that are in direct communication.
[0081] Therefore, the final result is:
[0082] B---A, C, D
[0083] C---A, B, D
[0084] D---B, C
[0085] For follower B, it is necessary to continuously adjust the system matrix parameters in the observer of follower B based on the real motion parameters of leader A and the motion parameters estimated by followers BCD. For follower C, it is also necessary to continuously adjust the system matrix parameters in the observer of follower C based on the real motion parameters of leader A and the motion parameters estimated by followers BCD. For follower D, the system matrix parameters in the observer of follower D are adjusted only based on the motion parameters estimated by followers BCD.
[0086] It should be noted that the system matrix parameters here need to be adjusted continuously. For example, if a set of actions includes 12345678, the system matrix parameters are updated once in action 1 and once in action 2. Or, considering that each action has a moving distance, the convergence time of the system matrix parameters can be adjusted according to the step length of the moving distance. For example, the convergence time of 100m is shorter than that of 50m, so that the system matrix parameters can be quickly adjusted when the error range is large. For details, see Figure 3 The description of step S303 is shown.
[0087] After stopping updating the system matrix parameters, it is no longer necessary to receive the real motion parameters transmitted by the leader and the estimated motion parameters transmitted by other followers, or no longer need to be processed after receiving, preferably no longer receive. Directly predict the next motion parameters of the leader based on them. For example, during the process of leader A performing the above 123 actions, follower BCD completes updating the system matrix parameters, then follower BCD can directly simulate the next action of leader A as 4. After performing action 4, it simulates the next action of leader A as 5, that is, there is no need to imitate leader A.
[0088] Taking follower B as an example, after simulating action 4 using the finally updated system matrix parameters, action 4 is sent to the controller of follower B, so that the controller controls the angular velocity and posture of follower B to perform action 4.
[0089] The method provided in the above embodiment mathematically models the communication link failure, estimates the leader state, and then transforms the multi-rigid body consistency problem into a single rigid body tracking problem. The observer is responsible for estimating the motion parameters of the leader in real time, and the controller is responsible for controlling the follower motion, ultimately achieving the consistency of the leader-follower motion parameters of the multi-rigid body system in the presence of leader uncertainty and communication link failure.
[0090] See also Figure 3 , shows a schematic flow chart of an optional leader-uncertain collaborative control method according to an embodiment of the present invention, comprising the following steps:
[0091] S301: Calculate the global state error between each follower and the leader based on the actual motion parameters and the motion parameters estimated by each follower;
[0092] S302: Calculating a global state error between each follower and other followers according to the motion parameters estimated by each follower and the motion parameters estimated by other followers;
[0093] S303: Calculate the update gain of the system matrix parameters according to the current round global state error, so as to adjust the training convergence time of the observer by updating the gain;
[0094] S304: updating the system matrix parameters of the observer of each follower according to the global state error calculated in each round, repeating the iterative update operation until the global state error is less than or equal to a preset threshold, and stopping updating the system matrix parameters of the observer of each follower.
[0095] In the above implementation, for steps S301 and S302, the real motion parameters of the leader are first defined, and the leader of the multi-rigid body system is defined as a typical linear system with uncertain system matrix parameters, which has the following form:
[0096]
[0097] y 0 =E 0 v 0 ,
[0098]
[0099] in, and For the specific meaning of each letter, see Figure 4 In addition, is an unknown parameter, E 0 is an unknown output matrix.
[0100] S(θ) is the system matrix. The system matrix of the leader is an antisymmetric matrix, which is generally expressed as:
[0101]
[0102] in, θ=[θ 01 ,θ 02 ,...,θ 0l ] T , Represents the zero matrix, 2l+m 0 =m,θ 0i >0,i=1,....,l. Further assume that m 0= 0, S(θ) follows the following form:
[0103] The follower's observer sets two calculation formulas to estimate the leader's angular velocity and attitude respectively:
[0104] Angular velocity estimation: in, represents the system matrix parameters, represents the output matrix, represents the auxiliary variable of angular velocity, represents the adaptive parameter of angular velocity. i and ξ i is the leader's true angular velocity, is the angular velocity of the leader estimated by the follower's observer.
[0105] Pose Estimation: in, represents the adaptive parameters of the posture, Represents attitude auxiliary variables.
[0106] The internal state error variable of the above local observer is:
[0107]
[0108] in,
[0109]
[0110] a ij and g i is a fixed communication weight, and is an unknown time-varying function and Assuming that CLFs are unknown corrupted communication weights, the CLFs between i and j, such as time delay, capacity constraints, and transmission noise, can be obtained by and To accurately describe and analyze CLF, the following assumptions are proposed to determine sufficient conditions that include communication topology and communication link failures, thereby giving sufficient conditions that include topology information and communication link failures to ensure the existence of the observer:
[0111] (1) The leader’s system matrix is an antisymmetric matrix.
[0112] (2) Fixed undirected topology is strongly connected.
[0113] (3) Communication link failures and their derivatives are bounded.
[0114] (4) The communication weight after loss has the same sign as the fixed communication weight.
[0115] The global form for defining internal state and unknown parameters is as follows:
[0116] e v =col(e v1 ,…,e vN )
[0117]
[0118] The global form of the internal state observation error is obtained as follows:
[0119]
[0120] In addition to the above method to calculate the global state error, you can also use the following method to calculate it. Here we take the calculation of the global state error between each follower and the leader as an example:
[0121] 1. First, define the motion parameters, which should include key parameters that describe the motion state, such as attitude, speed, angular velocity, etc. This involves sensor data processing, such as GPS (Global Positioning System) positioning to obtain position information, accelerometers and gyroscopes to obtain speed and attitude information, etc.
[0122] 2. For each individual (leader, follower), there is a corresponding motion parameter. Assume that the real motion parameter of the leader is (X_{L}=[x_{L},y_{L},v_{x,L},v_{y,L},w_{x,L},w_{y,L},...]^T), where (x,y) represents the posture coordinates, (v_x,v_y) represents the velocity component, (w_x,w_y) represents the angular velocity component, etc. For the follower, assume that the estimated motion parameter state of the leader is (X_{F}^{est}=[x_{F}^{est},y_{F}^{est},v_{x,F}^{est},v_{y,F}^{est},...,]^T). In order to simplify the description, only two-dimensional situations are shown here, and it can actually be set to three-dimensional situations or multi-dimensional situations.
[0123] 3. The state of the leader is used as the reference state. For each follower, the actual motion parameters of the leader are subtracted from the estimated motion parameters of the follower element by element to obtain the individual error parameters of the follower relative to the leader. For the global state error, the individual errors need to be summarized, which are determined by the attributes of the elements. For example, (x, y) represents the posture coordinates, (v_x, v_y) represents the velocity component, and (w_x, w_y) represents the angular velocity component. The individual errors representing the posture are summarized as one value, the individual errors representing the velocity are summarized as one value, and the individual errors representing the angular velocity are summarized as one value.
[0124] In addition, the errors can be summed directly or calculated as the mean, variance, and other statistics of the errors, depending on whether the overall consistency or dispersion is desired to be evaluated. The errors can be weighted according to the importance of the motion parameters. For example, the posture error may have a greater impact than the velocity error. Different weights can be assigned to different attribute elements and then weighted summed.
[0125] 4. The final global state error is an indicator that comprehensively considers the state difference between the follower and the leader, which reflects the synchronization and accuracy of the entire following system.
[0126] This solution preferably considers angular velocity and attitude, Figure 2 Taking follower B as an example, the angular velocity error and attitude error between the motion parameters estimated by follower B and the motion parameters of leader A, the angular velocity error and attitude error between the motion parameters estimated by follower B and the motion parameters estimated by follower C, and the angular velocity error and attitude error between the motion parameters estimated by follower B and the motion parameters estimated by follower D are calculated in each round. For follower D, it is only indirectly connected to leader A, so only the angular velocity error and attitude error between the motion parameters estimated by follower D and the motion parameters estimated by follower B, and the angular velocity error and attitude error between the motion parameters estimated by follower D and the motion parameters estimated by follower C are calculated.
[0127] For step S303, after calculating the global state error in each round, the update gain of the system matrix parameters can be calculated based on it, that is, as above and Two adaptive parameters are used to adjust the training convergence time of the observer by updating the gain, such as the training convergence time with an error of 10m is less than the training convergence time with an error of 50m.
[0128] In automatic control theory, the most basic is negative feedback control u=-kv, where u is the control input, v is the error, and k is the control gain. For example, when the electric kettle is heated to 70 degrees, the meaning of negative feedback control is that if the actual temperature is low, it will be heated, and if the actual temperature is high, it will be cooled. The k in the above formula represents the size of the heating power. The larger the k, the faster the heating. In summary: k is a theoretical value that can determine whether it converges or not and the speed of convergence (the concept of convergence is equivalent to the electric kettle reaching its set temperature). In non-adaptive control algorithms, the control gain k is generally fixed. On the contrary: adaptive control means that the control gain k is changeable, not a fixed value, and after the initial value of k is set, the subsequent changes do not need to be manually set by the controller, and k will adjust itself according to the algorithm.
[0129] For step S304, after calculating the global state error of the leader's motion parameters of the follower and other rigid bodies, the system matrix parameters of the observer in the follower are updated based on the global state error, and the above steps are repeated and iterated until the angular velocity error is less than or equal to the first threshold and the posture error is less than or equal to the second threshold, then stop updating the system matrix parameters of the observer.
[0130] The method provided in the above embodiment introduces adaptive parameters so that each rigid body can adaptively adjust the training convergence time of the system matrix parameters in the presence of communication uncertainty. The training convergence time of different followers can be different, thereby achieving the purpose of adaptive training of different followers.
[0131] The method provided in the embodiment of the present invention mathematically models the uncertain leader and communication link failure, and models it as a typical linear system. It only restricts the form of the system matrix parameters, and does not impose more restrictions on the parameters of the system matrix and the output matrix. Therefore, it can be used to generate many classic function trajectories and simulate the impact of external disturbances on the system, making the studied model universal.
[0132] The method provided by the embodiment of the present invention assumes that the communication topology of the multi-rigid body system is fully connected and the root node contains the leader, thereby relaxing the requirements for the communication weight of the attitude consistency of the multi-rigid body system and eliminating the dependence of the communication network on global information. On this basis, a fully distributed observer using the leader's state and output information is designed, so the observer does not need to know the eigenvalues of the Laplace matrix of the communication topology and other related information, so that each rigid body can track the uncertain system matrix parameters and outputs of the leader, thereby adjusting the system matrix parameters. In addition, the present invention relaxes the requirements for the knowability of the leader's dynamic model, eliminates the dependence of the observer design on the leader's system matrix parameters, and makes the observer of each follower more robust and more flexible.
[0133] See also Figure 5 , showing a schematic diagram of main modules of a collaborative control device 500 with uncertain leader provided by an embodiment of the present invention, including:
[0134] The receiving module 501 is used to receive the input motion instruction and send it to the leader, and when the leader starts to move according to the motion instruction, transmit the real motion parameter to the observer of the follower in direct communication;
[0135] An estimation module 502, for each follower to estimate the motion parameters of the leader through an observer when imitating the motion of the leader, and transmit the estimated motion parameters to the observers of other followers in direct communication;
[0136] A training module 503, configured to iteratively adjust the system matrix parameters of the observer for each follower according to the real motion parameter, at least one of the motion parameters estimated by other followers, and the motion parameter estimated by each follower;
[0137] The application module 504 is used to estimate the next motion parameter of the leader using the adjusted system matrix parameters, and transmit the next motion parameter to the controller of each follower, so that each follower is controlled by the controller to execute the next motion parameter.
[0138] The implementation device of the present invention also includes an operation module, which is used to:
[0139] Receive the rigid body code entered on the rigid body management center registration page, and establish a communication link between the rigid body management center and the registered rigid body;
[0140] Receiving multiple rigid bodies selected from a rigid body collection in a rigid body management center;
[0141] A rigid body selected from the plurality of rigid bodies is received, the rigid body is used as a leader, and the remaining rigid bodies in the plurality of rigid bodies are used as followers.
[0142] In the implementation device of the present invention, the receiving module 501 is used to:
[0143] Receive input action instructions and send them to the leader, so that the leader establishes a communication link with each follower, and each follower establishes a communication link with other followers.
[0144] The implementation device of the present invention also includes a judgment module, which is used to:
[0145] A communication topology is constructed based on the communication links between the leader and followers and between followers. When a direct communication link or an indirect communication link between each follower and the leader in the communication topology is available, an operation of updating the system matrix parameters is performed.
[0146] In the implementation device of the present invention, the training module 503 is used to:
[0147] Calculating a global state error between each follower and the leader according to the actual motion parameter and the motion parameter estimated by each follower; or calculating a global state error between each follower and other followers according to the motion parameter estimated by each follower and the motion parameter estimated by other followers;
[0148] According to the global state error calculated in each round, the system matrix parameters of the observer of each follower are updated, and the iterative update operation is repeated until the global state error is less than or equal to a preset threshold, and the updating of the system matrix parameters of the observer of each follower is stopped.
[0149] In the implementation device of the present invention, the training module 503 is used to:
[0150] Subtracting the actual motion parameter from the motion parameter estimated by each follower element by element to obtain an individual error parameter of each follower relative to the leader;
[0151] The attribute of each element is determined, and the individual errors of the elements belonging to the same attribute in the individual error parameters are accumulated to obtain the individual error value of each attribute, and then the global state error between each follower and the leader is obtained by summarizing.
[0152] In the implementation device of the present invention, the global state error includes an angular velocity error and a posture error, and the global state error being less than or equal to a preset threshold includes: the angular velocity error being less than or equal to a first preset threshold, and the posture error being less than or equal to a second preset threshold.
[0153] In the implementation device of the present invention, the training module 503 is also used for:
[0154] According to the global state error of the current round, the update gain of the system matrix parameters is calculated to adjust the training convergence time of the observer through the update gain.
[0155] In addition, the specific implementation content of the device described in the embodiment of the present invention has been described in detail in the method described above, so the repeated content will not be described again here.
[0156] Figure 6An exemplary system architecture 600 to which embodiments of the present invention may be applied is shown, including terminal devices 601 , 602 , 603 , a network 604 and a server 605 (only an example).
[0157] Terminal devices 601, 602, 603 can be various electronic devices with display screens and supporting web browsing, and various communication client applications are installed. Users can use terminal devices 601, 602, 603 to interact with server 605 through network 604 to receive or send messages, etc.
[0158] The network 604 is used to provide a medium for communication links between the terminal devices 601, 602, 603 and the server 605. The network 604 may include various connection types, such as wired, wireless communication links or optical fiber cables.
[0159] The server 605 may be a server that provides various services. It should be noted that the method provided in the embodiment of the present invention is generally executed by the server 605 , and accordingly, the device is generally set in the server 605 .
[0160] It should be understood that Figure 6 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.
[0161] Reference below Figure 7 , which shows a schematic diagram of the structure of a computer system 700 of a terminal device suitable for implementing an embodiment of the present invention. Figure 7 The terminal device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.
[0162] like Figure 7 As shown, the computer system 700 includes a central processing unit (CPU) 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage part 708 into a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the system 700 are also stored. The CPU 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.
[0163] The following components are connected to the I / O interface 705: an input section 706 including a keyboard, a mouse, etc.; an output section 707 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, a modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 710 as needed, so that a computer program read therefrom is installed into the storage section 708 as needed.
[0164] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 709, and / or installed from the removable medium 711. When the computer program is executed by the central processing unit (CPU) 701, the above-mentioned functions defined in the system of the present invention are executed.
[0165] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.
[0166] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0167] The modules involved in the embodiments of the present invention may be implemented by software or hardware. The modules described may also be set in a processor. For example, they may be described as: a processor includes a receiving module, an estimation module, a training module, and an application module. The names of these modules do not, in some cases, constitute limitations on the modules themselves. For example, an application module may also be described as a "use module".
[0168] As another aspect, the present invention further provides a computer-readable medium, which may be included in the device described in the above embodiment; or may exist independently without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a device, the device executes any of the above leader-uncertain collaborative control methods.
[0169] The computer program product of the present invention comprises a computer program, and when the computer program is executed by a processor, the leader-uncertain collaborative control method in the embodiment of the present invention is implemented.
[0170] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A collaborative control method with uncertain leader, characterized in that: include: receiving input motion instructions and sending them to the leader, and when the leader starts to move according to the motion instructions, transmitting the real motion parameters to the observer of the follower in direct communication; When each follower imitates the leader's motion, it estimates the leader's motion parameters through the observer and transmits the estimated motion parameters to the observers of other followers in direct communication; For the observer of each follower, iteratively adjust the system matrix parameters of the observer according to the real motion parameter, at least one of the motion parameters estimated by other followers, and the motion parameter estimated by each follower; The next motion parameter of the leader is estimated using the adjusted system matrix parameters, and the next motion parameter is transmitted to the controller of each follower, so that each follower is controlled by the controller to execute the next motion parameter.
2. The method according to claim 1, characterized in that Before receiving the input action instruction and issuing it to the leader, the method further includes: Receive the rigid body code entered on the rigid body management center registration page, and establish a communication link between the rigid body management center and the registered rigid body; Receiving multiple rigid bodies selected from a rigid body collection in a rigid body management center; A rigid body selected from the plurality of rigid bodies is received, the rigid body is used as a leader, and the remaining rigid bodies in the plurality of rigid bodies are used as followers.
3. The method according to claim 1, characterized in that The step of receiving the input action instruction and sending it to the leader includes: Receive input action instructions and send them to the leader, so that the leader establishes a communication link with each follower, and each follower establishes a communication link with other followers.
4. The method according to claim 1 or 3, characterized in that: The method further comprises: A communication topology is constructed based on the communication links between the leader and followers and between followers. When a direct communication link or an indirect communication link between each follower and the leader in the communication topology is available, an operation of updating system matrix parameters is performed.
5. The method according to claim 1, characterized in that The adjusting the system matrix parameters of the observer comprises: Calculating a global state error between each follower and the leader according to the actual motion parameter and the motion parameter estimated by each follower; or calculating a global state error between each follower and other followers according to the motion parameter estimated by each follower and the motion parameter estimated by other followers; According to the global state error calculated in each round, the system matrix parameters of the observer of each follower are updated, and the iterative update operation is repeated until the global state error is less than or equal to a preset threshold, and the updating of the system matrix parameters of the observer of each follower is stopped.
6. The method according to claim 5, characterized in that The calculating the global state error between each follower and the leader according to the real motion parameter and the motion parameter estimated by each follower comprises: Subtracting the actual motion parameter from the motion parameter estimated by each follower element by element to obtain an individual error parameter of each follower relative to the leader; The attribute of each element is determined, and the individual errors of the elements belonging to the same attribute in the individual error parameters are accumulated to obtain the individual error value of each attribute, and then the global state error between each follower and the leader is obtained by summarizing.
7. The method according to claim 5 or 6, characterized in that: The global state error includes an angular velocity error and a posture error. The global state error being less than or equal to a preset threshold includes: the angular velocity error being less than or equal to a first preset threshold, and the posture error being less than or equal to a second preset threshold.
8. The method according to claim 5 or 6, characterized in that: The method further comprises: According to the global state error of the current round, the update gain of the system matrix parameters is calculated to adjust the training convergence time of the observer through the update gain.
9. A collaborative control device with uncertain leader, characterized in that: include: A receiving module, used for receiving an input motion instruction and sending it to the leader, and transmitting the real motion parameters to the observer of the follower in direct communication when the leader starts to move according to the motion instruction; An estimation module, which is used for each follower to estimate the motion parameters of the leader through an observer when imitating the motion of the leader, and transmit the estimated motion parameters to the observers of other followers in direct communication; A training module, configured to iteratively adjust a system matrix parameter of the observer for each follower according to the real motion parameter, at least one of the motion parameters estimated by other followers, and the motion parameter estimated by each follower; The application module is used to estimate the next motion parameter of the leader using the adjusted system matrix parameters, and transmit the next motion parameter to the controller of each follower, so as to control each follower to execute the next motion parameter through the controller. Next motion parameter Next motion parameter Next motion parameter Next motion parameter 10. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 8.
11. A computer readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.
12. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.
Citation Information
Cited By
Underwater glider formation control method based on multi-agent reinforcement learning
CN122111051A
Underwater glider formation control method based on multi-agent reinforcement learning
CN122111051B
Port double-AGV layered cooperative control method and system based on rigid body consistency constraint
CN122525959A
