Multi-agent adaptive learning cooperative control method, device, electronic device, medium and program product considering channel attenuation

Abstract

The present invention discloses a multi-agent adaptive learning collaborative control method and device considering channel attenuation, the method comprising: constructing a nonlinear multi-agent model of a multi-agent system; constructing an unknown channel attenuation model; constructing an estimation algorithm for a first attenuation gain; obtaining an initial consistency error of the multi-agent; based on the initial consistency error, correcting the attenuation signal according to the unknown channel attenuation model and the estimation algorithm to obtain a target consistency error; constructing an adaptive learning collaborative control model according to the nonlinear multi-agent model and the target consistency error; wherein the adaptive learning collaborative control model is used for adaptive learning collaborative control of multi-agents with unknown channel attenuation. The present invention can improve the performance of collaborative control and can be widely used in the field of multi-agent control technology.

Description

Technical Field

[0001] The present invention relates to the field of multi-agent control technology, and in particular to a multi-agent adaptive learning collaborative control method and device taking into account channel attenuation. Background Art

[0002] A multi-agent system is a group system composed of multiple autonomous individuals, each of which has the ability to make decisions and take actions independently. At the same time, in order to solve more complex problems, agents need to learn to interact and collaborate with other agents. Due to various factors such as distance, obstacles or multipath propagation, multi-agents will inevitably cause channel attenuation when communicating with their neighbors. Channel attenuation may not only seriously reduce the cooperative control performance of the multi-agent system, but may even cause system instability. However, most of the current studies on the impact of channel attenuation on multi-agent systems assume that the statistics of the attenuation gain are known, while the cooperative control performance of non-square nonlinear multi-agent systems under unknown channel attenuation is still low, resulting in system instability. Summary of the invention

[0003] In view of this, the main purpose of the embodiments of the present invention is to provide a multi-agent adaptive learning collaborative control method and device taking channel attenuation into consideration to improve the collaborative control performance.

[0004] To achieve the above object, an embodiment of the present invention provides a multi-agent adaptive learning collaborative control method considering channel attenuation, the method comprising the following steps:

[0005] Construct nonlinear multi-agent models of multi-agent systems;

[0006] Construct unknown channel attenuation model;

[0007] Constructing an estimation algorithm for a first attenuation gain;

[0008] Obtain the initial consistency error of multiple agents;

[0009] Based on the initial consistency error, according to the unknown channel attenuation model and the estimation algorithm, the attenuated signal is corrected to obtain a target consistency error;

[0010] Constructing an adaptive learning collaborative control model according to the nonlinear multi-agent model and the target consistency error;

[0011] Among them, the adaptive learning collaborative control model is used for multi-agent adaptive learning collaborative control of unknown channel attenuation.

[0012] In some embodiments, constructing a nonlinear multi-agent model of a multi-agent system comprises the following steps:

[0013] According to the first system state of the target intelligent agent, a nonlinear function matrix and a control gain matrix are constructed;

[0014] Obtaining control input of the target agent;

[0015] Preset unknown system parameters;

[0016] A nonlinear multi-agent model is obtained according to the nonlinear function matrix, the control gain matrix, the control input and the unknown parameters of the system.

[0017] In some embodiments, constructing the unknown channel attenuation model comprises the following steps:

[0018] Get the second system state of the neighboring agent;

[0019] Obtaining a second attenuation gain of a directional channel from the neighboring agent to the target agent;

[0020] An unknown channel attenuation model is constructed according to the second system state and the second attenuation gain.

[0021] In some embodiments, the constructing the estimation algorithm of the first attenuation gain comprises the following steps:

[0022] Get the unit test signal received by the target agent;

[0023] An estimation algorithm for a first attenuation gain is constructed according to the unit test signal and the attenuation factor.

[0024] In some embodiments, the step of obtaining the initial consistency error of the multi-agent comprises the following steps:

[0025] Obtaining a directed graph of the communication topology of the multi-agent system;

[0026] Obtaining an adjacency matrix of the directed graph;

[0027] Obtaining an indicator matrix of the multi-agent system;

[0028] The initial consistency error of the multi-agent is obtained according to the adjacency matrix, the indicator matrix and the actual tracking error of the multi-agent.

[0029] In some embodiments, based on the initial consistency error, the attenuated signal is corrected according to the unknown channel attenuation model and the estimation algorithm to obtain a target consistency error, comprising the following steps:

[0030] Correcting the attenuated signal according to the unknown channel attenuation model and the estimation algorithm;

[0031] A target consistency error is obtained according to the corrected attenuation signal and the initial consistency error.

[0032] To achieve the above object, another aspect of an embodiment of the present invention provides a multi-agent adaptive learning collaborative control device considering channel attenuation, the device comprising:

[0033] The first module is used to construct a nonlinear multi-agent model of a multi-agent system;

[0034] The second module is used to construct an unknown channel attenuation model;

[0035] A third module is used to construct an estimation algorithm for the first attenuation gain;

[0036] The fourth module is used to obtain the initial consistency error of multiple agents;

[0037] A fifth module is used to correct the attenuated signal based on the initial consistency error, according to the unknown channel attenuation model and the estimation algorithm to obtain a target consistency error;

[0038] A sixth module is used to construct an adaptive learning collaborative control model according to the nonlinear multi-agent model and the target consistency error;

[0039] Among them, the adaptive learning collaborative control model is used for multi-agent adaptive learning collaborative control of unknown channel attenuation.

[0040] To achieve the above-mentioned purpose, another aspect of an embodiment of the present invention provides an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the multi-agent adaptive learning collaborative control method considering channel attenuation as described above.

[0041] To achieve the above-mentioned purpose, another aspect of an embodiment of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the aforementioned multi-agent adaptive learning collaborative control method considering channel attenuation.

[0042] To achieve the above-mentioned purpose, another aspect of an embodiment of the present invention provides a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. A processor of a computer device can read the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the aforementioned multi-agent adaptive learning cooperative control method considering channel attenuation.

[0043] The embodiments of the present invention include at least the following beneficial effects: The present invention provides a multi-agent adaptive learning collaborative control method considering channel attenuation, which constructs a nonlinear multi-agent model of the multi-agent system; constructs an unknown channel attenuation model; constructs an estimation algorithm for the first attenuation gain; obtains the initial consistency error of the multi-agent; based on the initial consistency error, according to the unknown channel attenuation model and the estimation algorithm, corrects the attenuation signal to obtain the target consistency error; constructs an adaptive learning collaborative control model according to the nonlinear multi-agent model and the target consistency error. The present invention proposes a new learning-based statistical estimation method to estimate the unknown attenuation gain, and corrects the data to compensate for the impact of the unknown channel attenuation. According to the repeated operation characteristics of the multi-agent system, the adaptive learning collaborative control model is introduced into the multi-agent collaborative control, and the collaborative control performance is gradually improved by using the repeated learning of the multi-agent. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 It is a flow chart of a multi-agent adaptive learning collaborative control method considering channel attenuation provided by an embodiment of the present invention;

[0046] Figure 2 is a schematic diagram of a communication network of a multi-agent system under the influence of channel fading provided by an embodiment of the present invention;

[0047] Figure 3 It is a schematic diagram of a design framework for adaptive iterative learning collaborative control of a multi-intelligence system provided by an embodiment of the present invention;

[0048] Figure 4 is a structural block diagram of an adaptive iterative learning control algorithm provided by an embodiment of the present invention;

[0049] Figure 5 It is a schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of the present invention, and they are only examples of devices and methods consistent with some aspects of the embodiments of the present invention as detailed in the attached claims.

[0051] It should be noted that, although the functional modules are divided in the system schematic diagram and the logical order is shown in the flow chart, in some cases, the steps shown or described may be performed in a different order than the module division in the system or the flow chart. The terms "first / S100" and "second / S200" in the specification and claims and the above-mentioned drawings may be used to describe various concepts in this article, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiment of the present invention, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the words "if" and "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determination".

[0052] The terms "at least one", "multiple", "each", "any", etc. used in the present invention, at least one includes one, two or more, multiple includes two or more, each refers to each of the corresponding multiple, and any refers to any one of the multiple.

[0053] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein are only for the purpose of describing the embodiments of the present invention and are not intended to limit the present invention.

[0054] Before describing the embodiments of the present invention in detail, some nouns and terms involved in the embodiments of the present invention are first described. The nouns and terms involved in the embodiments of the present invention are subject to the following explanations.

[0055] Adaptive iterative learning control (AILC),Adaptive iterative learning control combines the advantages of adaptive control and iterative learning control.It gradually reduces the error between the system output and the desired trajectory by adjusting the control input in each iteration.

[0056] Iterative learning control (ILC) is a control strategy for repetitive tasks that uses historical iteration data to improve system performance. This control method is particularly suitable for tasks that require high-precision tracking, such as robotics, manufacturing process control, etc.

[0057] The global Lipschitz condition (GLC) is a concept often used in mathematics and engineering to describe the global stability and continuity properties of a function or mapping.

[0058] The composite energy function (CEF) is a versatile analysis and design tool that provides a method to evaluate and optimize system performance by integrating multiple energy terms into a unified framework.

[0059] Multi-agent systems (MASs) are computing systems composed of multiple interacting agents. Each agent can be regarded as an entity that can perceive its environment and make decisions based on the perceived information.

[0060] Due to the limitations of the capabilities of a single agent and the lack of its own resources, the results of multi-target and multi-task are often unsatisfactory. Current research has only discussed the asymptotic consistency problem of multi-agent systems in the time domain, and has failed to use past historical information for learning to enhance the collaborative tracking performance of multi-agent systems that need to perform repetitive or periodic tasks. Although some studies have proposed collaborative tracking control methods for multi-agent systems based on iterative learning control, these results usually require that the multi-agent dynamics are linear or meet the global Lipschitz condition, which seriously limits the application of iterative learning control in actual multi-agent systems.

[0061] In addition, due to various factors such as distance, obstacles or multipath propagation, channel attenuation is inevitable when multi-agents communicate with their neighbors. Channel attenuation may not only seriously reduce the cooperative control performance of the multi-agent system, but may even cause system instability. However, most of the current studies on the impact of channel attenuation on multi-agent systems assume that the statistics of the attenuation gain are known. Although there are also some studies aimed at solving the problem of unknown channel attenuation for single systems, there is still the problem of cooperative tracking of non-square nonlinear multi-agent systems under unknown channel attenuation.

[0062] In view of this, if Figure 1As shown, an embodiment of the present invention provides a multi-agent adaptive learning collaborative control method considering channel attenuation, the method includes but is not limited to steps S100 to S600:

[0063] Step S100, constructing a nonlinear multi-agent model of the multi-agent system;

[0064] Step S200, constructing an unknown channel attenuation model;

[0065] Step S300, constructing an estimation algorithm for a first attenuation gain;

[0066] Step S400, obtaining the initial consistency error of multiple agents;

[0067] Step S500, based on the initial consistency error, according to the unknown channel attenuation model and the estimation algorithm, correcting the attenuated signal to obtain a target consistency error;

[0068] Step S600, constructing an adaptive learning collaborative control model according to the nonlinear multi-agent model and the target consistency error;

[0069] Among them, the adaptive learning collaborative control model is used for multi-agent adaptive learning collaborative control of unknown channel attenuation.

[0070] In some embodiments, step S100 includes but is not limited to steps S110 to S140:

[0071] Step S110, constructing a nonlinear function matrix and a control gain matrix according to the first system state of the target agent;

[0072] Step S120, obtaining the control input of the target agent;

[0073] Step S130, presetting unknown system parameters;

[0074] Step S140, obtaining a nonlinear multi-agent model according to the nonlinear function matrix, the control gain matrix, the control input and the unknown parameters of the system.

[0075] Exemplarily, the following nonlinear multi-agent model can be established:

[0076]

[0077] Where k represents the iteration index, k = 1, 2, ...; i represents the i-th agent, i = 1, 2, ..., N; t represents the time, T represents the trial length, t∈[0,T], T>0; x k,i represents the system state that can be measured by the ith agent, that is, the first system state, xk,i ∈R m ;u k,i represents the control input of the ith agent, u k,i ∈R m ;f(x k,i ,t) represents the nonlinear function matrix, f(x k,i ,t)∈R m×m ; Θ i represents the unknown parameters of the system, Θ i ∈R m ; B(x k,i ,t) represents the control gain matrix, B(x k,i ,t)∈R m ×m ; Represents the first system state x k,i The first derivative of .

[0078] In some embodiments, step S200 may include but is not limited to steps S210 to S230:

[0079] Step S210, obtaining the second system state of the neighboring agent;

[0080] Step S220, obtaining a second attenuation gain of a directional channel from the neighboring agent to the target agent;

[0081] Step S230: construct an unknown channel attenuation model according to the second system state and the second attenuation gain.

[0082] In some embodiments, channel attenuation can be modeled by multiplying the original signal by a random variable. Exemplarily, the following unknown channel attenuation model can be established:

[0083] x k,i,j (t) = μ k,i,j (t)x k,j (t) (2)

[0084] In the formula, x k,i,j represents the signal received by the ith agent from its neighbor agent j; x k,j represents the system state measurable by the jth agent, i.e., the second system state; μ k,i,j (t) represents the attenuation gain of the directional channel from neighbor agent j to the i-th agent at time t, that is, the second attenuation gain, μ k,i,j (t)∈(0,1), obeys continuous distribution. And, for all iteration indexes k and time t, the second attenuation gain μ k,i,j (t) are independent and identically distributed. In addition, the second attenuation gain μ k,i,j The mathematical expectation μ of (t) i,jand variance are all unknown. The mathematical expectation expression of the second attenuation gain is: in, Represents mathematical expectation; the variance expression of the second attenuation gain is: like Figure 2 The communication network of a multi-agent system shown is affected by channel fading, where the fading phenomenon only occurs in the communication channel between follower agents.

[0085] In some embodiments, step S300 may include but is not limited to steps S310 to S320:

[0086] Step S310, obtaining a unit test signal received by the target agent;

[0087] Step S320: construct an estimation algorithm for the first attenuation gain according to the unit test signal and the attenuation factor.

[0088] In some embodiments, at the beginning of each iteration, the unknown attenuation gain mean of the channel is experimentally estimated by sending and collecting unit test signals from agent j to agent i. When the unit test signal received by agent i is The expression is:

[0089]

[0090] Among them, μ k,i,j (τ) represents the attenuation gain suffered by the directional channel of the unit test signal from agent j to agent i at time τ, μ k,i,j The distribution of (τ) and μ k,i,j (t) is the same. Combined with the adjustable attenuation factor, the following learning-based estimation algorithm for the unknown channel attenuation gain can be constructed:

[0091]

[0092] in, represents the estimated unknown attenuation gain; β represents the attenuation factor, β∈(0,1). By adjusting the attenuation factor, the estimation requirements in different environments can be met. This estimation algorithm can use historical information to improve the estimation accuracy.

[0093] In some embodiments, step S400 may include but is not limited to steps S410 to S440:

[0094] Step S410, obtaining a directed graph of the communication topology of the multi-agent system;

[0095] Step S420, obtaining an adjacency matrix of the directed graph;

[0096] Step S430, obtaining an indicator matrix of the multi-agent system;

[0097] Step S440, obtaining the initial consistency error of the multi-agent according to the adjacency matrix, the indicator matrix and the actual tracking error of the multi-agent.

[0098] In some embodiments, the communication topology of the multi-agent system can be represented by a directed graph Description. Where v represents a set of N agents (i.e., a directed graph Node set), v=(v1,v2,…,v N ), ε represents the communication connection between different agents (i.e., the directed graph , If (i,j)∈ε, then the i-th agent can receive information from the j-th agent; if Represents that the i-th agent is not allowed to connect to itself. In addition, the directed graph The adjacency matrix of It can be expressed as: And a i,j ∈{0,1}. If (i,j)∈ε, then a i,j =1; if Then a i,j =0; due to Then a i,i = 0. The neighbor set of the ith agent can be expressed as: N i = {v j ∈v:(i,j)∈ε}. The in-degree d of the i-th node i It can be expressed as: In addition, the indicator matrix It can be expressed as: If the i-th agent can receive information from the leader, then p i =1; if the i-th agent cannot receive the message from the leader, then p i = 0. Directed graph The Laplace matrix of can be represented by the positive semidefinite matrix Definition. Among them, D is a directed graph The degree matrix of Assuming a directed graph If a spanning tree is included, the leader serves as the root of the spanning tree.

[0099] In some embodiments, in order to realize multi-agent collaborative tasks, according to the adjacency matrix, the indicator matrix and the actual tracking error of the multi-agents, the initial consistency error ξ of the target agent can be obtained as follows: k,i The expression of (t) is:

[0100]

[0101] In the formula, e k,i (t) is the actual tracking error of the ith agent, e k,j (t) is the actual tracking error of the jth agent, Among them, x d (t) is the desired reference trajectory.

[0102] In some embodiments, step S500 may include but is not limited to steps S510 to S520:

[0103] Step S510, correcting the attenuated signal according to the unknown channel attenuation model and the estimation algorithm;

[0104] Step S520: acquiring a target consistency error according to the corrected attenuation signal and the initial consistency error.

[0105] In step S510 of some embodiments, according to the unknown channel attenuation model, the attenuated signal received by the ith agent from its neighboring agent j is corrected using the unknown channel attenuation gain estimation algorithm based on learning, and the corrected attenuated signal x′ is k,i,j (t) can be expressed as:

[0106]

[0107] In the formula, represents the first intermediate parameter,

[0108] In step S520 of some embodiments, considering the signal x k,j (t) is unavailable due to channel attenuation. The corrected attenuation signal can be used, combined with the definition of the initial consistency error, to obtain the corrected target consistency error of the i-th agent as follows: The expression is:

[0109]

[0110] In some embodiments, Figure 3As shown in the figure, in the multi-agent adaptive learning cooperative control method considering channel attenuation, the nonlinear multi-agent model, the unknown channel attenuation model, the estimation algorithm of the first attenuation gain, the initial consistency error and the target consistency error, etc. can be obtained. In order to simplify the symbols, the nonlinear multi-agent model of the multi-agent system can be converted into the following compact form by using the lifting technique:

[0111]

[0112] In the formula,

[0113]

[0114] In addition, define the diagonal matrix Ω i ∈R N×N , where the i-th element is 1 and the rest are 0. Similarly, the initial consistency error of the multi-agent and the target consistency error after correction can be converted into the following compact forms:

[0115]

[0116] In the formula,

[0117]

[0118] In the formula, Represents the second intermediate parameter; I m is the m*m identity matrix; represents a positive semidefinite matrix; represents the indicator matrix; Ω L N diagonal matrices Ω i Composition, Ω L ∈R N×NN ; Represents the third intermediate parameter; I N is the N*N identity matrix; A represents the adjacency matrix of the directed graph.

[0119] In step S600 of some embodiments, in order to realize the coordinated control of MASs under unknown channel attenuation, an AILC scheme is proposed, and the new adaptive iterative learning coordinated control model is:

[0120]

[0121] To further simplify the notation, we use f k , B k represents f(x k ,t),B(x k ,t);K∈R mN×mNis the control gain matrix to be designed, which can be assumed to be a positive definite matrix; is the expected trajectory; u k is the control input; is an intermediate parameter; the constant θ satisfies:

[0122]

[0123] in, Similarly, In the formula, sup(·) is the upper limit function, is an estimate of the unknown parameters Θ of the system in compact form, The update rules are as follows:

[0124]

[0125] In the formula, The learning gain matrix Γ to be determined is a diagonal matrix with positive diagonal elements. In addition, when implementing the proposed cooperative controller, the initial input u0 can be directly set to zero or generated by any pre-specified feedback controller to produce a bounded tracking error e0. Then we can get Figure 4 The adaptive iterative learning cooperative controller structure frame is shown. For the nonlinear multi-agent model with unknown parameters under the influence of unknown channel attenuation (i.e., equation (1)), the proposed adaptive learning cooperative control model (i.e., equation (11)), parameters The learning update rate (i.e., Equation (12)) and the learning-based estimation algorithm for the unknown channel attenuation gain (i.e., Equation (4)) can ensure that the consistency error ξ in a compact form decreases as the number of iterations k→∞. k (t),t∈[0,T] converges to zero in the mean square sense.

[0126] In some embodiments, the convergence of the learning-based unknown channel attenuation gain estimation algorithm can be proved by the Lyapunov function. First, the Lyapunov function is established as follows:

[0127]

[0128] For t∈[0,T], V k,1 The difference can be expressed as:

[0129]

[0130] Among them, since β∈(0,1) and k>0, Therefore, from equation (14), we can see that ΔV k,1 <0, which means that as the number of iterations k increases, V k,1 The values ​​are monotonically decreasing.

[0131] From formula (14), we can deduce:

[0132]

[0133] It should be noted that V k,1 is positive, and since μ i,j is finite, so V 0,1 is bounded. Therefore, it can be deduced from equation (15) that when the number of iterations k tends to infinity, converges to zero, that is

[0134] In some embodiments, the composite energy function CEF can be used for analysis to prove the convergence of the new adaptive iterative learning cooperative control model. First, CEF is established as follows:

[0135]

[0136] In the formula, is the parameter estimation error,

[0137] Convergence analysis can be divided into two parts, including the convergence of difference and tracking errors between two consecutive iterations:

[0138] 1) Difference of CEF between two iterations

[0139] The difference of CEF in two iterations can be expressed as:

[0140]

[0141] It can be seen from the above formula that formula (17) is divided into terms. The first term is is to prove the convergence of the tracking error. In each iteration, the initial state of the system can be reset to the same position as the reference state, that is, e k (0) = 0, from equation (9) we can get ξ k (0)=0, then we can get:

[0142]

[0143] By applying the definition of the compact form of the consistency error (i.e., Equation (9)) and the new adaptive iterative learning cooperative control model (i.e., Equation (11)), we can obtain:

[0144]

[0145] In addition, the following inequality holds:

[0146]

[0147] Similarly, we can get the following equation:

[0148]

[0149] Therefore, substituting equations (20) and (21) into equation (19), we can obtain:

[0150]

[0151] For the second item of CEF According to the iterative learning law, the following inequality can be obtained:

[0152] Substituting equations (22) and (23) into the CEF difference equation (17), we can get the following conclusions:

[0153] In addition, by combining equations (9) and (10), we can obtain:

[0154]

[0155] In the above transformation, the inequality ab≥-1 / 2a is used 2 -1 / 2b 2 Substituting (25) into (24) we obtain:

[0156]

[0157] Using ψ k The boundedness of the second-order moment of (t) can be deduced as follows:

[0158]

[0159] The constant Taking the mathematical expectation on both sides of (26) and substituting (27) into (26) yields:

[0160]

[0161] because And K is positive definite, the above equation holds. From equation (28), we can see that the difference between the two iterations of CEF is less than 0, indicating that The value of decreases monotonically with the increase of the number of iterations. According to equation (28), the convergence of the tracking error can be analyzed as follows.

[0162] 2) Convergence of tracking error

[0163] According to equations (26) and (28), it is not difficult to draw the following conclusion: the difference accumulation can be written as:

[0164]

[0165] because

[0166] It should be noted that is positive definite, and since both ξ0 and Θ0 are finite, V 0,2 (t) is bounded. Therefore, it can be deduced from equation (29) that ||ξ k-1 (t)||, t∈[0,T] is convergent in the mean square sense.

[0167] Further, by k and e k The definition of can be obtained, in the presence of unknown channel attenuation, the tracking error e k (t),t∈[0,T] converges in the mean square sense as the number of iterations increases.

[0168] The embodiment of the present invention further provides a multi-agent adaptive learning collaborative control device considering channel attenuation, which can implement the above-mentioned multi-agent adaptive learning collaborative control method considering channel attenuation, and the device includes:

[0169] The first module is used to construct a nonlinear multi-agent model of a multi-agent system;

[0170] The second module is used to construct an unknown channel attenuation model;

[0171] A third module is used to construct an estimation algorithm for the first attenuation gain;

[0172] The fourth module is used to obtain the initial consistency error of multiple agents;

[0173] A fifth module is used to correct the attenuated signal based on the initial consistency error, according to the unknown channel attenuation model and the estimation algorithm to obtain a target consistency error;

[0174] A sixth module is used to construct an adaptive learning collaborative control model according to the nonlinear multi-agent model and the target consistency error;

[0175] Among them, the adaptive learning collaborative control model is used for multi-agent adaptive learning collaborative control of unknown channel attenuation.

[0176] It can be understood that the contents of the above method embodiments are all applicable to the present device embodiments, the functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0177] The embodiment of the present invention further provides an electronic device, which includes a processor and a memory, wherein the memory stores a computer program, and when the processor executes the computer program, the above-mentioned multi-agent adaptive learning cooperative control method considering channel attenuation is implemented. The electronic device can be any intelligent terminal including a tablet computer, a car computer, etc.

[0178] It can be understood that the contents of the above method embodiments are all applicable to the present device embodiments, the functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0179] refer to Figure 5 , Figure 5 The hardware structure of an electronic device of another embodiment is illustrated, and the electronic device includes:

[0180] The processor 701 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present invention.

[0181] The memory 702 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 702 can store an operating system and other applications. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 702, and the processor 701 calls and executes a multi-agent adaptive learning collaborative control method considering channel attenuation in an embodiment of the present invention;

[0182] Input / output interface 703, used to implement information input and output;

[0183] Communication interface 704, used to realize communication interaction between the device and other devices, which can be realized through wired mode (such as USB, network cable, etc.) or wireless mode (such as mobile network, WIFI, Bluetooth, etc.);

[0184] A bus 705 that transmits information between the various components of the device (e.g., the processor 701, the memory 702, the input / output interface 703, and the communication interface 704);

[0185] The processor 701 , the memory 702 , the input / output interface 703 and the communication interface 704 are connected to each other in communication within the device via a bus 705 .

[0186] An embodiment of the present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the above-mentioned multi-agent adaptive learning collaborative control method considering channel attenuation.

[0187] It can be understood that the contents of the above method embodiments are all applicable to the present storage medium embodiments, the functions specifically implemented by the present storage medium embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0188] The embodiment of the present invention also provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the aforementioned multi-agent adaptive learning collaborative control method considering channel attenuation.

[0189] In summary, the multi-agent adaptive learning collaborative control method and device considering channel attenuation in the embodiment of the present invention has the following advantages:

[0190] 1. Due to the influence of unknown channel attenuation, the neighbor signals received by the agent usually deviate randomly from their true values, which will seriously reduce the collaborative control performance of the multi-agent system and cause the multi-agent system to be unstable. The embodiment of the present invention estimates and corrects the statistical information of the unknown attenuation gain of each neighborhood channel through a new learning-based estimation method, thereby solving the problem that the received signal randomly deviates from the true value due to the influence of unknown channel attenuation on the communication between agents.

[0191] 2. The embodiment of the present invention proposes a novel adaptive iterative learning control scheme to realize the collaborative tasks of nonlinear multi-agent systems with repetitive motion characteristics. It can fully utilize the repeatability of the multi-agent system for learning while ensuring the accuracy of collaborative control, thereby achieving the purpose of improving the adaptability and robust performance of the multi-agent system.

[0192] In some selectable embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided by way of example, for the purpose of providing a more comprehensive understanding of technology. The disclosed method is not limited to the operation and logic flow presented herein. Selectable embodiments are expected, wherein the order of various operations is changed and the sub-operation of a part for which is described as a larger operation is performed independently.

[0193] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise specified, one or more of the functions and / or features described may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in separate physical devices or software modules. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the present invention. More specifically, in view of the properties, functions, and internal relationships of the various functional modules in the device disclosed herein, the actual implementation of the module will be understood within the conventional skills of the engineer. Therefore, those skilled in the art can implement the present invention set forth in the claims without excessive experimentation using ordinary techniques. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.

[0194] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0195] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0196] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0197] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0198] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0199] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

[0200] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the described embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of the present invention.

Claims

1. A multi-agent adaptive learning cooperative control method considering channel attenuation, characterized in that: The following steps are involved: Construct nonlinear multi-agent models of multi-agent systems; Construct unknown channel attenuation model; Constructing an estimation algorithm for a first attenuation gain; Obtain the initial consistency error of multiple agents; Based on the initial consistency error, according to the unknown channel attenuation model and the estimation algorithm, the attenuated signal is corrected to obtain a target consistency error; Constructing an adaptive learning collaborative control model according to the nonlinear multi-agent model and the target consistency error; Wherein, the adaptive learning collaborative control model is used for multi-agent adaptive learning collaborative control with unknown channel attenuation; The method of constructing a nonlinear multi-agent model of a multi-agent system comprises the following steps: According to the first system state of the target intelligent agent, a nonlinear function matrix and a control gain matrix are constructed; Obtaining control input of the target agent; Preset unknown system parameters; Obtaining a nonlinear multi-agent model according to the nonlinear function matrix, the control gain matrix, the control input and the unknown parameters of the system; The method of constructing an unknown channel attenuation model comprises the following steps: Get the second system state of the neighboring agent; Obtaining a second attenuation gain of a directional channel from the neighboring agent to the target agent; constructing an unknown channel attenuation model according to the second system state and the second attenuation gain; The estimation algorithm for constructing the first attenuation gain comprises the following steps: Get the unit test signal received by the target agent; Constructing an estimation algorithm of a first attenuation gain according to the unit test signal and the attenuation factor; The method of obtaining the initial consistency error of the multi-agent comprises the following steps: Obtaining a directed graph of the communication topology of the multi-agent system; Obtaining an adjacency matrix of the directed graph; Obtaining an indicator matrix of the multi-agent system; Obtaining an initial consistency error of the multi-agent according to the adjacency matrix, the indicator matrix and the actual tracking error of the multi-agent; The method of correcting the attenuated signal based on the initial consistency error and according to the unknown channel attenuation model and the estimation algorithm to obtain a target consistency error includes the following steps: Correcting the attenuated signal according to the unknown channel attenuation model and the estimation algorithm; A target consistency error is obtained according to the corrected attenuation signal and the initial consistency error.

2. A multi-agent adaptive learning cooperative control device considering channel attenuation, characterized in that: include: The first module is used to construct a nonlinear multi-agent model of a multi-agent system; The second module is used to construct an unknown channel attenuation model; A third module is used to construct an estimation algorithm for the first attenuation gain; The fourth module is used to obtain the initial consistency error of multiple agents; A fifth module is used to correct the attenuated signal based on the initial consistency error, according to the unknown channel attenuation model and the estimation algorithm to obtain a target consistency error; A sixth module is used to construct an adaptive learning collaborative control model according to the nonlinear multi-agent model and the target consistency error; Wherein, the adaptive learning collaborative control model is used for multi-agent adaptive learning collaborative control with unknown channel attenuation; The first module is specifically used for: According to the first system state of the target intelligent agent, a nonlinear function matrix and a control gain matrix are constructed; Obtaining control input of the target agent; Preset unknown system parameters; Obtaining a nonlinear multi-agent model according to the nonlinear function matrix, the control gain matrix, the control input and the unknown parameters of the system; The second module is specifically used for: Get the second system state of the neighboring agent; Obtaining a second attenuation gain of a directional channel from the neighboring agent to the target agent; constructing an unknown channel attenuation model according to the second system state and the second attenuation gain; The third module is specifically used for: Get the unit test signal received by the target agent; Constructing an estimation algorithm of a first attenuation gain according to the unit test signal and the attenuation factor; The fourth module is specifically used for: Obtaining a directed graph of the communication topology of the multi-agent system; Obtaining an adjacency matrix of the directed graph; Obtaining an indicator matrix of the multi-agent system; Obtaining an initial consistency error of the multi-agent according to the adjacency matrix, the indicator matrix and the actual tracking error of the multi-agent; The fifth module is specifically used for: Correcting the attenuated signal according to the unknown channel attenuation model and the estimation algorithm; A target consistency error is obtained according to the corrected attenuation signal and the initial consistency error.

3. An electronic device, characterized in that: including a processor and a memory; The memory is used to store programs; The processor executes the program to implement the method according to claim 1.

4. A computer-readable storage medium, characterized in that: The storage medium stores a program, and the program is executed by a processor to implement the method according to claim 1.

5. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to claim 1 is implemented.

Images