Game control method for space target flexible capture means

Abstract

The invention discloses a game control method for a flexible capturing means of a space target. The method comprises the following steps: acquiring state information of each satellite of the space flexible capture system; determining an initial state sequence function of each satellite based on the state information of each satellite; designing a weighted DNN network to fit the initial state sequence function, and obtaining a final state sequence function of each satellite; and determining a predicted tension sequence of each satellite based on the final state sequence function of each satellite. According to a traditional flexible capture control method, a formation control scheme is adopted to take tension generated by a flexible net as external interference, a passive suppression method is adopted for compensation, but the control strategy ignores the intrinsic nature of the flexible net, when one moonlet maneuvers, the tension is generated through the net and transmitted to other moonlets, and thus the moonlet can be effectively captured. And the tension generated by the net influences other satellites.

Description

Technical Field

[0001] The present invention relates to the technical fields of flexible capture control and space target capture, and specifically, to a game control method for a flexible capture means for space targets. Background Art

[0002] A space flexible capture system is a new type of flexible capture means for targets such as space debris, failed satellites, and hostile spacecraft. It consists of four small satellites carrying a flexible net, and has the advantages of large envelope, easy capture, and no debris in case of collision, and can simply, safely, and reliably complete the space target capture task. The formation control of each small satellite inside the space flexible capture system is the basis of the capture task. At present, the formation control schemes designed in patents CN105905319A and CN108319135A regard the tension generated by the flexible net as an external interference and adopt a passive suppression method for compensation. However, this control strategy ignores the intrinsic nature of the flexible net. When a small satellite maneuvers, it will generate tension through the net and transmit it to other small satellites. This force not only has a negative effect on the control accuracy, but also has a positive effect. The data-driven method can well describe the complex relationship between satellites caused by the flexible net constraint, and the control method based on game theory is applicable to such scenarios where multiple intelligent agents interact with each other. Therefore, the present invention combines data-driven and game theory to propose a flexible capture cooperative game formation control strategy for the capture task. By depicting the relationship between small satellites interacting with each other through the flexible net, the essence of the space flexible capture system is further explored, and its control ability is greatly improved. Summary of the Invention

[0003] An embodiment of the present invention provides a game control method for a flexible capture means for space targets, so as to at least solve the technical problem that the traditional flexible capture control method regards the tension generated by the flexible net as an external interference and adopts a passive suppression method for compensation, but this control strategy ignores the intrinsic nature of the flexible net. When a small satellite maneuvers, it will generate tension through the net and transmit it to other small satellites, resulting in the tension generated by the net having an impact on other satellites.

[0004] According to one aspect of an embodiment of the present invention, a game control method for a flexible capture means of a space target is provided. The method may include: obtaining the state information of each satellite of the space flexible capture system; determining the initial state sequence function of each satellite based on the state information of each satellite; designing a weighted DNN network to fit the initial state sequence function to obtain the final state sequence function of each satellite; determining the predicted tension sequence of each satellite based on the final state sequence function of each satellite; obtaining the predicted state quantity of each satellite of the space flexible capture system based on the predicted tension sequence, control quantity and state quantity of each satellite; obtaining the tracking error of each satellite based on the predicted state quantity and the desired state quantity of each satellite; defining the cooperation error between different satellites in the space flexible capture system based on the tracking error of each satellite; obtaining the cost function of each satellite based on the cooperation error between different satellites, the control quantity of each satellite and the weight matrix; and solving the Pareto optimal control quantity of each satellite through a model prediction method based on the cost function of each satellite.

[0005] Optionally, determining the initial state sequence function of each satellite based on the state information of each satellite includes: determining the initial state sequence of each satellite based on the state information of each satellite; and obtaining the initial state sequence function of each satellite based on the initial state sequence of each satellite.

[0006] Optionally, after determining the initial state sequence function of each satellite based on the state information of each satellite, the method further includes: determining the initial tension sequence of each satellite based on the initial state sequence function of each satellite.

[0007] Optionally, the expression for obtaining the final state sequence function of each satellite by designing a weighted DNN network to fit the state sequence function is:

[0008] Φ = W L+1 φ(W L (…φ(W 1 X i )…))

[0009] where Φ is the ReLu activation layer, θ = W 1 , …, W L+1 represents a weight, L represents the number of layers, and X i is the state sequence function.

[0010] Optionally, the expression for obtaining the predicted state quantity of each satellite of the space flexible capture system based on the predicted tension sequence, control quantity and state quantity of each satellite is:

[0011]

[0012] where represents the first predicted state quantity of the \(i\)-th satellite, \(r_0\) represents the radius of the orbit of the \(i\)-th satellite, and \(\omega\) represents the angular velocity of the orbit of the \(i\)-th satellite. represents the acceleration of the orbit of the \(i\)-th satellite, \(u\) i represents the control quantity of the \(i\)-th satellite. represents the predicted tension sequence of the \(i\)-th satellite, \(x\) i represents the state quantity of the \(i\)-th satellite.

[0013] Optionally, based on the predicted state quantity of each satellite and the desired state quantity of each satellite, the expression for the tracking error of each satellite is obtained as:

[0014]

[0015] where represents the tracking error of the \(i\)-th satellite, represents the predicted state quantity of the \(i\)-th satellite, represents the desired state quantity of the \(i\)-th satellite.

[0016] Optionally, based on the cooperative error between different satellites, the control quantity of each satellite, and the weight matrix, the expression for the cost function of each satellite is obtained as:

[0017]

[0018] where \(J\) i represents the cost function of each satellite, \(t\) represents time, and the value range of \(t\) is \(t\) o ~\(t\) f , \(t_0\) and represent the initial and terminal times, \(T\) represents the transpose, \(e\) i represents the cooperative error between the \(i\)-th satellite and the other satellites except the \(i\)-th satellite, \(Q\) i , \(R\) i and \(R\) ij both represent the weight matrix, \(u\) i represents the control quantity of the \(i\)-th satellite, represents the predicted control quantity of the \(i\)-th satellite, \(u\) j represents the control quantity of the \(j\)-th satellite, the predicted control quantity of the \(j\)-th satellite, \(N\) i represents the number of other satellites except the \(i\)-th satellite in the space flexible capture system, represents the transpose of the predicted control quantity of the \(j\)-th satellite.

[0019] Advantages of the present invention:

[0020] The present invention proposes a game control method for a flexible capture means of space targets. First, a neural network tension observer is designed to estimate and predict the flexible net tension based on DNN (Deep Neural Network), revealing the interaction relationship between satellites through the flexible net. On this basis, a distributed cooperative game formation control strategy is proposed, enabling each small satellite to actively control the influence of the flexible net tension caused by its own maneuver on other satellites through interaction and optimization, solving the technical problem that the traditional flexible capture control method treats the tension generated by the flexible net as an external interference in the formation control scheme and compensates it by a passive suppression method. However, this control strategy ignores the intrinsic nature of the flexible net. When a small satellite maneuvers, it generates tension through the net and transmits it to other small satellites, resulting in the tension generated by the net affecting other satellites. The technical effect is achieved by combining data-driven, game theory, and formation control methods, enabling the satellites in the flexible net to reach an equilibrium state where the movement of any satellite in the flexible net does not affect other satellites. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0022] Figure 1 is a flowchart of a game control method for a flexible capture means of space targets according to an embodiment of the present invention;

[0023] Figure 2 is a schematic diagram of a space flexible capture system according to an embodiment of the present invention;

[0024] Figure 3 is a schematic diagram of the process of a space flexible capture system capturing a target according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0026] It should be noted that the terms "first", "second", etc. in the specification, claims and above-mentioned drawings of the present invention are used to distinguish similar objects and describe a specific order or sequence. It should be understood that such terms can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order different from those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0027] Embodiment 1

[0028] According to an embodiment of the present invention, a game control method for a flexible capture means of a space target is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system including at least one set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that here.

[0029] Figure 1 is a flowchart of a game control method for a flexible capture means of a space target according to an embodiment of the present invention. As Figure 1 shown, the method may include the following steps:

[0030] Step S101, obtaining the state information of each satellite of the space flexible capture system.

[0031] In the technical solution provided in step S101 of the present invention above, before obtaining the state information of each satellite of the space flexible capture system, the dynamic model of the space flexible capture system is:

[0032]

[0033] Wherein, represents the derivative of the state quantity of node mn in the space flexible capture system, x mn represents the state quantity of node mn in the space flexible capture system, u mn represents the control quantity of node mn in the space flexible capture system, T mn represents the tension of node mn in the space flexible capture system.

[0034] The biggest difference from other space vehicles lies in the tension T mn generated by the flexible net; inside the flexible net, each grid edge between two adjacent nodes is simplified into a spring-damping model; therefore, the tension calculation formula of node mn is:

[0035]

[0036] Among them, T mn represents the tension of node mn, N mn is the set of adjacent nodes of node mn; T mn-kh is the tension between node mn and node kh; l z and l mn-kh respectively represent the initial length and the actual length of the grid, Δl mn-kh = l mn-kh - l z is the elongation of the grid; E is the Young's modulus of the net material; A is the cross-sectional area of the grid; is the elastic modulus; is the unit vector pointing from node mn to node kh.

[0037] According to dynamics, the tension T i of satellite i is related to the nearby nodes, and these nodes are in turn affected by their respective neighbors; extending outwards, it can be seen that the four satellites affect each other by transmitting forces through the net; therefore, T i is caused by the maneuvers of adjacent satellites and can be expressed as:

[0038] T i = f(x i , x -i , u i , u -i )

[0039] Among them, x -i and u -i are the states and control variables of the adjacent units of the i-th satellite; x i , u i are respectively the state variable and the control variable of the i-th satellite itself, and f is a function representing the mapping from the state and control to the tension, which is highly non-linear. Although the analytical solution can be obtained by calculating the tension of each node inside the net, in practical applications, it is unrealistic to obtain the state of each node; at the same time, this function has very obvious time delays. The vibration caused by one satellite will be transmitted to adjacent satellites like a wave through the net; therefore, the current force T i，k is related to the past states, and the current states will generate a new sequence of tensions in the future.

[0040] Step S102, based on the state information of each satellite, determine the initial state sequence function of each satellite.

[0041] In the technical solution provided in step S102 of the present invention above, according to the state information of each satellite, the initial state sequence function of each satellite is obtained.

[0042] Step S103: Design a DNN network with weights to fit the initial state sequence function, and obtain the final state sequence function of each satellite.

[0043] In the technical solution provided in step S103 of the present invention, a DNN with weights θ = W 1 , …, W L+1 is designed to fit the initial state sequence function, and the final state sequence function of each satellite is obtained.

[0044] Step S104: Based on the final state sequence function of each satellite, determine the predicted tension sequence of each satellite.

[0045] In the technical solution provided in step S104 of the present invention, according to the final state sequence function of each satellite, the predicted tension sequence of each satellite is determined.

[0046] Step S105: Based on the predicted tension sequence, control quantity, and state quantity of each satellite, obtain the predicted state quantity of each satellite in the space flexible capture system.

[0047] In the technical solution provided in step S105 of the present invention, according to the predicted tension sequence, control quantity, and state quantity of each satellite, the predicted state quantity of each satellite in the space flexible capture system is obtained.

[0048] Step S106: Based on the predicted state quantity of each satellite and the desired state quantity of each satellite, obtain the tracking error of each satellite.

[0049] In the technical solution provided in step S106 of the present invention, the difference between the predicted state quantity of each satellite and the desired state quantity of each satellite is determined as the tracking error of each satellite.

[0050] Step S107: Based on the tracking error of each satellite, define the collaborative error between different satellites in the space flexible capture system.

[0051] In the technical solution provided in step S107 of the present invention, according to the tracking error, the expression of the collaborative error is defined as:

[0052]

[0053] where a ij and b i are constants.

[0054] Define the global error The derivative of the collaborative error is: where n is the number of satellites, T is the transpose,

[0055]

[0056] Among them, l i is the i-th column of the Laplacian matrix ; similarly, b is the i-th column of the matrix i . As can be seen from the above formula, there is a cooperation relationship between satellites. n is the number of satellites, and T is the transpose. ; Step S108: Based on the collaborative error between different satellites, the control quantity of each satellite, and the weight matrix, obtain the cost function of each satellite.

[0057] In the technical solution provided in step S108 of the present invention, according to the collaborative error between different satellites, the control quantity of each satellite, and the weight matrix, determine the cost function of each satellite.

[0058] ;

[0059] Step S109: Based on the cost function of each satellite, solve the Pareto optimal control quantity of each satellite through the model prediction method.

[0060] In the technical solution provided in step S109 of the present invention, Figure 2 is a schematic diagram of the space flexible capture system according to an embodiment of the present invention. As Figure 2 shown, in a cooperative game with 4 small satellites, a flexible net, and n participants, for any two sets of game strategies and u = [u1,..., u n T , if the following conditions exist, then u * is superior to u in the Pareto sense: Among them, the participants are each small satellite:

[0061]

[0062] If there is no other strategy superior to u * , then u * is the Pareto optimal solution; the Pareto optimality can be regarded as a consensus among the participants, that is, it is impossible to improve the benefit of a certain participant without causing losses to other participants; the Pareto optimality can be obtained by minimizing the cost functions of all participants:

[0063]

[0064] where λ i ∈(0, 1) is the cost function weight of participant i,

[0065] ​To obtain the Pareto optimal solution of the cooperative game of the design, a model predictive solution strategy is proposed. First, the game model is discretely iterated. First, the dynamics of the satellite are discretized:

[0066]

[0067] where A k and B k are the discrete dynamics matrices.

[0068] Furthermore, the discrete cooperative consistency error is obtained:

[0069]

[0070] where, The same applies to other matrices.

[0071] Then, this formula is iterated within the prediction time domain and the error sequence and control sequence are defined: Then the game model becomes:

[0072]

[0073]

[0074] Next, the same treatment is done for the cost function, and its discretely iterated form is:

[0075]

[0076] where,

[0077] Due to the coupling of the game model and the cost function, a single satellite cannot calculate its optimal game strategy alone. To achieve distributed formation control, the cost function is further decoupled:

[0078]

[0079] Satellite i can only optimize its own strategy, so the terms not including U i，k are discarded, and thus the game controller can be transformed into the following quadratic programming (QP) problem:

[0080]

[0081] During the game process, satellite i at each moment t k solves the above QP problem based on its own information, as well as the control strategy U j and the state information X j of adjacent units to obtain the Pareto optimal strategy. After obtaining their respective optimal strategy sequences U i , the first item is extracted and through calculation, the true control quantity is obtained. Figure 3 is a schematic diagram of the process of a space flexible capture system capturing a target according to an embodiment of the present invention.

[0082] The above method of this embodiment will be further introduced below.

[0083] As an alternative embodiment, in step S102, based on the state information of each satellite, determining the initial state sequence function of each satellite includes: based on the state information of each satellite, determining the initial state sequence of each satellite; based on the initial state sequence of each satellite, obtaining the initial state sequence function of each satellite.

[0084] In this embodiment, according to the state information of each satellite the initial state sequence of each satellite is obtained According to the initial state sequence of each satellite, the initial state sequence function of each satellite (Φ(X i (t k -N p Δtt k ))) is obtained.

[0085] As an alternative embodiment, in step S102, after determining the initial state sequence function of each satellite based on the state information of each satellite, the method further includes: based on the initial state sequence function of each satellite, determining the initial tension sequence of each satellite.

[0086] In this embodiment, the expression for determining the initial tension sequence of each satellite based on the state sequence function of each satellite is:

[0087] T i (t k +N p Δt|t k ) = Φ(X i (t k -N p Δt|t k ))

[0088] where, Φ(X i (t k -N P Δt|t k )) is the state sequence function of the i-th satellite, and X i (t k -N p Δt|t k ) represents the initial state sequence of the i-th satellite The expression for the initial state is Np is the prediction time domain, represents the transpose of the state quantity of the adjacent satellites of the i-th satellite at the k-th moment, represents the transpose of the control quantity of the i-th satellite at the k-th moment, represents the transpose of the control quantity of the adjacent satellites of the i-th satellite at the k-th moment, represents the transpose of the state quantity of the i-th satellite at the k-th moment, T i (t k +N p Δt|t k ) represents the initial tension sequence

[0089] As an alternative embodiment, in step S103, a weighted DNN network is designed to fit the state sequence function, and the expression of the final state sequence function of each satellite is obtained as:

[0090] Φ = W L+1 φ(W L (…φ(W 1 X i ))…))

[0091] where Φ is the ReLu activation layer, θ = W 1 , …, W L+1 represents a weight, L represents the number of layers, and X i is the state sequence function.

[0092] After determining the predicted tension sequence of each satellite based on the final state sequence function of each satellite, it further includes: determining the target loss function based on the predicted tension sequence of each satellite and the initial tension sequence of each satellite.

[0093] The target loss function L is defined as follows:

[0094]

[0095] where, is the estimated value of the tension T i ; α and β are positive hyperparameters.

[0096] The loss function includes two parts: the estimation part L e and the prediction part L p ; The purpose of the first part L e is to estimate the tension value at each moment:

[0097]

[0098] Among them, γ is a positive number. Since excessive tension will cause damage to the net or even instability of the entire system, overestimation of the tension is allowed, while underestimation is not allowed. Therefore, γ takes values between (0, 0.5) to penalize underestimation.

[0099] The second part L p is aimed at predicting the change of tension over a period of time, so it has the following form:

[0100]

[0101] Among them, η n ∈(0, 1) is a discount factor. Due to the time delay effect between satellites, the current state X i (t k ) has little influence on the tension T i (t k +Δt) in the next moment. On the contrary, it will have a greater impact on the tension in the near future. Therefore, the parameter η n is strongly related to time.

[0102] As an alternative embodiment, in step S105, based on the predicted tension sequence, control quantity, and state quantity of each satellite, the expression for the predicted state quantity of each satellite of the space flexible capture system is:

[0103]

[0104] Among them, represents the first predicted state quantity of the i-th satellite, r0 represents the radius of the orbit of the i-th satellite, ω represents the angular velocity of the orbit of the i-th satellite, represents the acceleration of the orbit of the i-th satellite, u i represents the control quantity of the i-th satellite, represents the predicted tension sequence of the i-th satellite, x i represents the state quantity of the i-th satellite.

[0105] In this embodiment, it is further defined that the predicted state quantity of each satellite of the space flexible capture system can also be:

[0106]

[0107] As an alternative embodiment, in step S107, based on the predicted state quantity of each satellite and the desired state quantity of each satellite, the expression for the tracking error of each satellite is:

[0108]

[0109] Among them, represents the tracking error of the i-th satellite, represents the predicted state quantity of the i-th satellite, represents the desired state quantity of the i-th satellite.

[0110] As an alternative embodiment, in step S109, based on the cooperative error between different satellites, the control quantity of each satellite, and the weight matrix, the expression of the cost function for each satellite is:

[0111]

[0112] where, J i represents the cost function of each satellite, t represents time, and the value range of t is t o ~t f , t0 and represent the initial and terminal times, T represents transpose, e i represents the cooperative error between the i-th satellite and other satellites except the i-th satellite, Q i , R i and R ij both represent weight matrices, u i represents the control quantity of the i-th satellite, represents the predicted control quantity of the i-th satellite, u j represents the control quantity of the j-th satellite, the predicted control quantity of the j-th satellite, N i represents the number of satellites other than the i-th satellite in the space flexible capture system, represents the transpose of the predicted control quantity of the j-th satellite.

[0113] In the embodiments of the present invention, the state information of each satellite in the space flexible capture system is obtained; based on the state information of each satellite, the initial state sequence function of each satellite is determined; a weighted DNN network is designed to fit the initial state sequence function to obtain the final state sequence function of each satellite; based on the final state sequence function of each satellite, the predicted tension sequence of each satellite is determined; based on the predicted tension sequence, control quantity, and state quantity of each satellite, the predicted state quantity of each satellite in the space flexible capture system is obtained; based on the predicted state quantity of each satellite and the desired state quantity of each satellite, the tracking error of each satellite is obtained; based on the tracking error of each satellite, the collaborative error between different satellites in the space flexible capture system is defined; based on the collaborative error between different satellites, the control quantity of each satellite, and the weight matrix, the cost function of each satellite is obtained; based on the cost function of each satellite, the Pareto optimal control quantity of each satellite is solved by the model prediction method, solving the technical problem that in the traditional flexible capture control method, the formation control scheme is adopted to regard the tension generated by the flexible net as an external interference and the passive suppression method is used for compensation. However, this control strategy ignores the intrinsic nature of the flexible net. When a small satellite maneuvers, it will generate tension through the net and transmit it to other small satellites, resulting in the tension generated by the net affecting other satellites, achieving the technical effect of combining data-driven, game theory, and formation control methods to enable the satellites in the flexible net to reach an equilibrium state and the movement of any satellite in the flexible net not to affect other satellites.

[0114] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0115] In the above embodiments of the present invention, the descriptions of the respective embodiments have their own emphases. For the parts not detailed in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0116] In several embodiments provided by the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the division of units can be a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of units or modules can be in an electrical or other form.

[0117] The unit described as a separating component may or may not be physically separated, and the component displayed as a unit may or may not be a physical unit, that is, it may be located in one place or may be distributed over multiple units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0118] In addition, each functional unit in various embodiments of the present invention may be integrated in a first processing unit, may also exist physically separately for each unit, or two or more units may be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0119] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and modifications can be made, and these improvements and modifications should also be regarded as the protection scope of the present invention.

Claims

1. A game control method for flexible capture of space targets, characterized in that: include: Obtain status information of each satellite in the space flexible capture system; Based on the state information of each satellite, determining an initial state sequence function of each satellite; Design a weighted DNN network to fit the initial state sequence function and obtain the final state sequence function of each satellite; Determining a predicted tension sequence for each satellite based on the final state sequence function for each satellite; Based on the predicted tension sequence, control quantity and state quantity of each satellite, the predicted state quantity of each satellite of the space flexible capture system is obtained; Based on the predicted state quantity of each satellite and the expected state quantity of each satellite, the tracking error of each satellite is obtained; Based on the tracking error of each satellite, the coordination error between different satellites in the space flexible capture system is defined; Based on the coordination errors between different satellites, the control amount and weight matrix of each satellite, the cost function of each satellite is obtained; Based on the cost function of each satellite, the Pareto optimal control quantity of each satellite is solved by the model prediction method.

2. The method according to claim 1, characterized in that: The step of determining the initial state sequence function of each satellite based on the state information of each satellite comprises: Based on the state information of each satellite, determining an initial state sequence of each satellite; Based on the initial state sequence of each satellite, an initial state sequence function of each satellite is obtained.

3. The method according to claim 1, characterized in that After determining the initial state sequence function of each satellite based on the state information of each satellite, the method further includes: Based on the initial state sequence function of each satellite, the initial tension sequence of each satellite is determined.

4. The method according to claim 1, characterized in that: The design of a weighted DNN network to fit the state sequence function obtains the expression of the final state sequence function of each satellite as follows: Φ=W L+1 φ(W L (…φ(W 1 X i )…)) Where Φ is the ReLu activation layer, θ = W 1 , …, W L+1 represents a weight, L represents the number of layers, X i is the state sequence function.

5. The method according to claim 1, characterized in that Based on the predicted tension sequence, control quantity and state quantity of each satellite, the expression of the predicted state quantity of each satellite of the space flexible capture system is obtained as follows: in, Indicates i The first predicted state of the satellite, r 0 Indicates i The radius of the satellite orbit, ω represents the i The angular velocity of the satellite orbit, Indicates i The acceleration of the satellite orbit, u i Indicates i The control quantity of each satellite, Indicates i The predicted tension series of satellites, x i Indicates i The state of a satellite.

6. The method according to claim 1, characterized in that The expression of the tracking error of each satellite obtained based on the predicted state quantity of each satellite and the expected state quantity of each satellite is: in, Indicates i The tracking error of the satellites, Indicates i The predicted state of each satellite is Indicates i The expected state of a satellite.

7. The method according to claim 1, characterized in that The expression of the cost function of each satellite is obtained based on the coordination error between different satellites, the control amount and weight matrix of each satellite: Among them, J i represents the cost function of each satellite, t represents time, and the value range of t is t o ~t f , t0 and t0 represent the initial and terminal moments, T represents the transpose, e i Representative i Satellites and except i The coordination error between other satellites of the satellite, Q i , R i and R ij Both represent weight matrices, u i Indicates i The control quantity of each satellite, Indicates i The predicted control quantity of a satellite, u j represents the control amount of the jth satellite, The predicted control quantity of the jth satellite, N i represents the number of satellites other than the i-th satellite in the space flexible capture system, represents the transpose of the predicted control quantity for the jth satellite.

8. A processor, characterized in that: The processor is used to run a program, wherein the program executes the method according to any one of claims 1 to 7 when running.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

Images