X-type tail vane control distribution method based on neural network

Abstract

The invention relates to the technical field of automatic control of aircraft control surfaces, in particular to an X-type tail vane control distribution method based on a neural network. The method comprises the following steps: generating a training sample of an X-type tail vane, obtaining an initial mapping set of a rudder angle combination to a moment of the X-type tail vane, screening the initial mapping set to obtain a mapping set of the rudder angle combination to the moment of the X-type tail vane, training a neural network according to the training sample to obtain a trained neural network model, and designing a neural network structure, and training the neural network structure by using the mapping set from the rudder angle combination to the moment to obtain a trained neural network model, and calculating the rudder angle combination by using the trained neural network model by taking an expected moment as input. According to the method, the neural network with the expected moment as input and the rudder angle as output is designed and trained, and control distribution of multiple control surfaces is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of automatic control of aircraft control surfaces, and in particular to an X-type tail rudder control distribution method based on a neural network. Background Art

[0002] Modern air and underwater vehicles are generally designed with multiple control surfaces to improve the maneuverability, redundancy and fault tolerance of the vehicle. Some fighter jets in the prior art are designed with multiple control surfaces such as ailerons, flaps, elevators, and rudders, and some nuclear submarines in the prior art are designed with an X-rudder composed of four control surfaces. In the control process of multi-control surface vehicles, control allocation is a basic and important component, which aims to calculate the corresponding rudder angle of each control surface for the expected rolling moment, pitching moment and yaw moment.

[0003] For a vehicle with n control surfaces, there are currently two methods for calculating the rudder angle u for the desired torque during its motion control process: the pseudo-inverse method and the optimization-based method. The former has a small amount of calculation, but ignores the physical limitations of the control surfaces, and the generated solution may not be operational; the latter can consider the relevant constraints and can be extended to the case of control surface failure. However, the optimization-based method has a large amount of calculation and is not conducive to real-time response. Therefore, it is urgent to design a new control allocation method with better performance than the pseudo-inverse method and lower calculation than the optimization method. Summary of the invention

[0004] In view of the above-mentioned shortcomings of the prior art, the purpose of the present invention is to provide an X-type tail rudder control allocation method based on a neural network, which solves the problems in the prior art that the pseudo-inverse method cannot take into account the actual constraints and the optimization method has a large amount of calculation.

[0005] To achieve the above-mentioned and other related purposes, the present invention provides an X-type tail rudder control allocation method based on a neural network, comprising:

[0006] S1, generate training samples of X-shaped tail rudder;

[0007] S2. Training a neural network according to the training samples to obtain a trained neural network model;

[0008] S3. Taking the expected torque as input, using the trained neural network model, calculate the steering angle combination.

[0009] In one embodiment of the present invention, generating the training sample of the X-shaped tail rudder in step S1 includes:

[0010] S11, obtaining an initial mapping set S0 of the rudder angle combination to the torque of the X-type tail rudder;

[0011] S12. Screen the initial mapping set S0 to obtain a mapping set S from steering angle combinations to torques.

[0012] In one embodiment of the present invention, obtaining an initial mapping set S0 of the rudder angle combination to the torque of the X-type tail rudder in step S11 includes:

[0013] According to the set step size, in u min <u<u max The rudder angle combination and speed are traversed within the range to obtain the torque τ=(K, M, N) when the rudder angle combination u and the speed v are obtained. After traversal, the initial mapping set S0={s|(u, v, K, M, N)} is obtained.

[0014] In one embodiment of the present invention, in the mapping set S of the rudder angle combination to the torque, under the condition of the speed v, the rudder angle combination u and the torque vector (K, M, N) are mapped one to one, and the inverse mapping of the rudder angle combination u to the speed v and the torque vector (K, M, N) is fitted.

[0015] In one embodiment of the present invention, the step S2 of training a neural network according to the training sample to obtain a trained neural network model includes:

[0016] S21. Design neural network structure;

[0017] S22. Using the mapping set S of the steering angle combination to the torque to train the neural network structure, so as to obtain a trained neural network model.

[0018] In one embodiment of the present invention, designing a neural network structure in step S21 includes:

[0019] S211, the input layer of the neural network structure includes 4 nodes, corresponding to K, M, N and speed v in the moment vector respectively;

[0020] S212, the output layer of the neural network structure has n nodes, corresponding to n steering angles u1, u2, u3, ..., u n ;

[0021] S213, the number of hidden layers of the neural network structure ranges from 1 to 5.

[0022] In one embodiment of the present invention, the step S22 of training the neural network structure using the mapping set S of the steering angle combination to the torque to obtain the trained neural network model includes:

[0023] S221, randomly divide the mapping set S into 3 subsets, the proportion of the number of subset elements is The value range of a is 0.5 to 0.8, and the three subsets are used as training set, test set, and validation set respectively;

[0024] S222. Randomly set the initial parameters of the neural network structure, use the training set, test set, and validation set to perform training, testing, and validation respectively. If the mean square errors of training, testing, and validation are all less than a threshold value t, the training is terminated, and the corresponding neural network structure is used to control the allocation.

[0025] In one embodiment of the present invention, the step S3 uses the desired torque as input and the trained neural network model to calculate the steering angle combination, including:

[0026] When the speed is v, the expected torque τ is (K, M, N), then (K, M, N, v) is used as input, and the output of the neural network model is calculated using the trained neural network model, and the output of the neural network model u = (u1, u2, u3, ..., u n ) as the final control rudder angle.

[0027] In one embodiment of the present invention, the threshold t is in the range of 10 -10 ~10 -6 .

[0028] The present invention also provides an electronic device, comprising a processor and a memory, wherein the memory stores program instructions, and the processor runs the program instructions to implement the above-mentioned X-type tail rudder control allocation method based on neural network.

[0029] As described above, the X-type tail rudder control allocation method based on neural network of the present invention has the following beneficial effects:

[0030] (1) The neural network-based X-type tail rudder control allocation method of the present invention designs and trains a neural network with a desired torque as input and a rudder angle as output, thereby realizing control allocation of multiple control surfaces.

[0031] (2) The X-type tail rudder control allocation method based on neural network of the present invention has a control surface angle range of u when generating training samples. min <u<u max , taking into full account the physical limitations of the control surface, thus overcoming the problem that the solution generated by the pseudo-inverse method may exceed u min <u<u max scope, resulting in inoperability issues.

[0032] (3) Compared with the optimization method, the neural network-based X-type tail rudder control allocation method of the present invention only needs to use the trained neural network model for regression calculation when performing control allocation. There is no loop and iteration process during calculation, and the amount of calculation is small, which is conducive to improving the real-time response capability of the aircraft. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A schematic diagram of multiple control surfaces of an X-type rudder control allocation method based on a neural network provided in an embodiment of the present application.

[0034] Figure 2 A flowchart of a neural network-based X-type tail rudder control allocation method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0035] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0036] It should be noted that the illustrations provided in the following embodiments are only used to illustrate the basic concept of the present invention in a schematic manner, and thus the illustrations only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0037] Terms such as first or second can be used to describe various components, but these components are not limited by the above terms. The above terms are used to distinguish one component from another component, for example, without departing from the scope of the concept according to the present disclosure, a first component can be referred to as a second component, and similarly, a second component can be referred to as a first component.

[0038] In addition, "connected / coupled" means that one component is directly electrically coupled to another component or is indirectly electrically coupled through another component. The singular form may include the plural form unless explicitly stated in the sentence. In addition, "include / comprise" or "includes / comprises" used in this specification indicates that one or more components, steps, operations, and elements exist or have been added. The specific structural or functional descriptions of the examples of the implementation of the concepts disclosed in this specification are only illustrated to describe the examples of the implementation of the concepts, and the examples of the implementation of the concepts can be implemented in various forms, but these descriptions are not limited to the examples of the implementation described in this specification.

[0039] According to the concept, various modifications and changes can be applied to the examples of the embodiments, so that the examples of the embodiments will be illustrated in the drawings and described in the specification. However, the examples of the embodiments according to the concept are not limited to the specific embodiments, but include all changes, equivalents or replacements included in the spirit and technical scope of the present disclosure.

[0040] It should be understood that when an element is described as being "coupled" or "connected" to another element, the element may be directly coupled or directly connected to the other element, or may be coupled or connected to the other element through a third element. Conversely, it should be understood that when an element is referred to as being "directly coupled to" or "directly coupled to" another element, no other element is placed between them. Other expressions describing the relationship between components (i.e., "between" and "directly between" or "adjacent to" and "directly adjacent to") need to be interpreted in the same manner.

[0041] The terms used in this specification are only used to describe specific examples of the embodiments and are not intended to limit the present disclosure. If there is no clear opposite meaning in the context, the singular form may include the plural form. In this specification, it should be understood that the term "including" or "having" indicates the presence of the features, quantities, steps, operations, components, parts or combinations thereof described in the specification, but the possibility of the presence or addition of one or more other features, quantities, steps, operations, components, parts or combinations thereof cannot be precluded.

[0042] If there is no contrary definition, all terms (including technical terms or scientific terms) used herein have the same meaning as those generally understood by ordinary technicians in the field. If the terms defined in the commonly used dictionary are not clearly defined in this specification, they should be interpreted as having the same meaning as in the context of the relevant technology, and not interpreted as ideal or overly formal meanings.

[0043] Descriptions of well-known components and processing techniques may be omitted so as not to unnecessarily obscure the embodiments of the disclosure.

[0044] Throughout the specification, the same reference numerals refer to the same elements. Thus, even if a reference numeral is not mentioned or described with reference to one figure, it may be mentioned or described with reference to another figure. Furthermore, even if a reference numeral is not shown in one figure, it may be mentioned or described with reference to another figure.

[0045] In addition, the logic level of the signal may be different or opposite to the described logic level. For example, a signal described as having a logic "high" level may alternatively have a logic "low" level, and a signal described as having a logic "low" level may alternatively have a logic "high" level.

[0046] The following will describe the various embodiments of the present disclosure in detail with reference to the accompanying drawings. However, it will be appreciated by those skilled in the art that in the various embodiments of the present disclosure, many technical details are provided in order to enable the reader to better understand the present disclosure. However, even without these technical details and various changes and modifications based on the following embodiments, the technical solutions claimed in the present disclosure can be implemented.

[0047] At present, for a vehicle with n control surfaces, during its motion control process, the virtual control command (i.e., the desired rolling moment K, pitching moment M, and yaw moment N) given by the control law is τ∈R 3 , the angle of the control surface (i.e., the rudder angle) is u∈R n , the control allocation efficiency matrix is ​​B, then: There are currently two methods for calculating the rudder angle u for the desired torque τ: the pseudo-inverse method and the optimization-based method. The former has a small amount of calculation, but ignores the physical limitations of the control surface, and the generated solution may not be operational; the latter can consider the relevant constraints and can be extended to the case of control surface failure. However, the optimization-based method has a large amount of calculation and is not conducive to real-time response.

[0048] In the pseudo-inverse method, u is generally ignored. min <u<u max The constraints of matrix B are used to directly calculate the rudder angle u, which is:

[0049] The pseudo-inverse method has a small amount of computation, but ignores u min <u<u max Constraints, so in practical applications, the generated solution may exceed u min <u<u max range and thus not operable.

[0050] In the optimization method, the target optimization function, equality constraints and inequality constraints are generally established, namely:

[0051]

[0052] Then, the solution is performed based on an optimization solution algorithm, such as a sequential quadratic programming algorithm (SQP). The optimization method generally requires multiple iterations and has a relatively high amount of calculation, so it is not conducive to application in an environment with high real-time requirements.

[0053] In order to simultaneously solve the problems that the pseudo-inverse method cannot take actual constraints into account and the optimization method has a large amount of calculation, the present invention proposes an X-type tail rudder control allocation method based on a neural network.

[0054] See also Figure 1 , Figure 1 A schematic diagram of multiple control surfaces of an X-type rudder control allocation method based on a neural network provided in an embodiment of the present application.

[0055] The present invention converts the control allocation problem of multiple control surfaces into a fitting problem of the mapping relationship from torque to control surface angle, and provides an X-type tail rudder control allocation method based on a neural network. The method uses data obtained from theoretical calculations or fluid tests to design and train a neural network with a desired torque as input and a rudder angle as output, thereby realizing control allocation of multiple control surfaces.

[0056] In one embodiment of the present invention, the present invention includes but is not limited to a common X-rudder aircraft composed of four control surfaces. For the convenience of description, the conventional symbols are described as follows: Figure 1 As shown, for a spacecraft with n control surfaces, the installation angle of the i-th control surface is α i , the corresponding rudder angle is u i , rudder angle combination u=(u1,u2,u3,…,u n )The torque τ generated at speed v is (K, M, N).

[0057] See also Figure 2 , Figure 2 The present invention provides a neural network-based X-type rudder control allocation method, comprising:

[0058] Step S1: Generate training samples of X-type tail rudder.

[0059] Specifically, the training samples for generating the X-shaped tail rudder in step S1 include:

[0060] Step S11, obtaining an initial mapping set S0 of the rudder angle combination to the torque of the X-type tail rudder.

[0061] Specifically, the initial mapping set S0 of the rudder angle combination to the torque of the X-type tail rudder in step S11 includes: according to the set step size,min <u<u max The rudder angle combination and speed are traversed within the range to obtain the torque τ=(K, M, N) when the rudder angle combination u and the speed v are obtained. After traversal, the initial mapping set S0={s|(u, v, K, M, N)} is obtained.

[0062] Step S12: Screen the initial mapping set S0 to obtain a mapping set S from steering angle combinations to torques.

[0063] The basic elements of the training samples in the present invention are torque τ = (K, M, N), speed v, and corresponding steering angle combination u. The generation of training samples includes two steps: obtaining an initial mapping set S0 of steering angle combination to torque; screening the initial mapping set to obtain a mapping set S of steering angle combination to torque.

[0064] In one embodiment of the present invention, the basic elements of the training samples in the present invention are torque τ = (K, M, N), speed v, and corresponding steering angle combination u. The generation of training samples includes two steps: obtaining an initial mapping set S0 of steering angle combination to torque; screening the initial mapping set S0 to obtain a mapping set S of steering angle combination to torque.

[0065] Step 1: Get the initial mapping set S0 from rudder angle combination to torque:

[0066] Using theoretical calculations, or wind tunnel or towing tank tests, at a certain step length, min <u<u max The rudder angle combination and speed are traversed within the range to obtain the torque τ=(K, M, N) when the rudder angle combination u and the speed v are obtained. After traversal, the initial mapping set S0={s|(u, v, K, M, N)} is obtained.

[0067] Step 2: Screen the initial mapping set S0 to obtain a mapping set S from rudder angle combination to torque:

[0068] Due to the redundancy of multiple control surfaces, different rudder angle combinations can produce the same torque at the same speed v. Therefore, the initial mapping set S0 of rudder angle combination to torque is not a one-to-one mapping relationship. For this reason, the initial mapping set S0 needs to be screened to obtain a one-to-one mapping set S of torque to rudder angle combination.

[0069] The pseudo code for the screening algorithm is as follows:

[0070]

[0071] The criterion used when selecting the best m from Q can be determined according to specific needs. If the overall rudder angle is expected to be as small as possible, the criterion can be set as follows:

[0072] min(u1·u1+u2·u2+…+u n ·u n )

[0073] If we want the ith rudder angle u i As small as possible, the criterion can be set as:

[0074] min(u i ·u i ),

[0075] In the mapping set S obtained after screening, under the condition of a specific speed v, the rudder angle combination u and (K, M, N) are mapped one to one. Therefore, the neural network can be used to fit the inverse mapping of the rudder angle combination u to the speed v and the torque vector (K, M, N), that is, the control allocation is completed.

[0076] Step S2: training a neural network according to the training samples to obtain a trained neural network model.

[0077] Specifically, the step S2 of training a neural network according to the training sample to obtain a trained neural network model includes:

[0078] Step S21, designing a neural network structure.

[0079] Specifically, the designing of the neural network structure in step S21 includes:

[0080] Step S211: The input layer of the neural network structure includes 4 nodes, corresponding to K, M, N and speed v in the moment vector respectively.

[0081] Step S212: The output layer of the neural network structure has n nodes, corresponding to n steering angles u1, u2, u3, ..., u n .

[0082] Step S213: The number of hidden layers of the neural network structure ranges from 1 to 5.

[0083] Step S22: training the neural network structure using the mapping set S of the steering angle combination to the torque to obtain a trained neural network model.

[0084] Specifically, in step S22, the neural network structure is trained using the mapping set S of the steering angle combination to the torque to obtain a trained neural network model, which includes:

[0085] Step S221: randomly divide the mapping set S into three subsets, with the proportion of the number of subset elements being The value range of a is 0.5-0.8, and the three subsets are used as training set, test set, and validation set respectively.

[0086] Step S222, randomly set the initial parameters of the neural network structure, use the training set, test set, and validation set to perform training, testing, and validation respectively, if the mean square errors of training, testing, and validation are all less than a threshold value t, then terminate the training, and the corresponding neural network structure is used to control the allocation.

[0087] In one embodiment of the present invention, the input layer of the neural network structure has 4 nodes, corresponding to K, M, N and speed v in the moment vector respectively; the output layer has n nodes, corresponding to the steering angles u1, u2, u3, ..., u n The number of hidden layers ranges from 1 to 5, and can be selected according to the number of control surfaces and the complexity of the layout. The general principle is that the more control surfaces and the more complex the layout, the larger the number of hidden layers can be set to. The activation function can be set to the sigmoid function.

[0088] The mapping set S is randomly divided into 3 subsets, and the proportion of the number of subset elements is (the value range of a is 0.5-0.8), and the obtained three subsets are used as training set, test set and validation set.

[0089] The initial parameters of the neural network are randomly set, and training, testing and verification are performed using the training set, test set and verification set. If the mean square error of training, testing and verification is less than the threshold t, the training is completed and the corresponding neural network can be used to control the allocation. The value range of the threshold t is 10 -10 ~10 -6 .

[0090] Step S3: Taking the expected torque as input, using the trained neural network model to calculate the steering angle combination.

[0091] In one embodiment of the present invention, when the speed is v, if the desired torque τ is (K, M, N), (K, M, N) is used as input, and the output of the neural network is calculated using the trained neural network model, and the output u = (u1, u2, u3, ..., u n ) as the final control rudder angle.

[0092] The present invention also proposes an electronic device, which includes a processor and a memory, wherein the memory stores program instructions, and the processor runs the program instructions to implement the above-mentioned X-type tail rudder control allocation method based on neural network. The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components; the memory may include a random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. The memory can also be an internal memory of the random access memory (RAM) type, and the processor and the memory can be integrated into one or more independent circuits or hardware, such as an application-specific integrated circuit (ASIC). It should be noted that the computer program in the above-mentioned memory can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention is essentially 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, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, an electronic device, or a network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention.

[0093] The present invention also proposes a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, wherein the computer instructions are used to enable the computer to execute the above-mentioned X-type tail rudder control allocation method based on neural network. The computer-readable storage medium can be an electronic medium, a magnetic medium, an optical medium, an electromagnetic medium, an infrared medium or a semiconductor system or a propagation medium. The computer-readable storage medium can also include a semiconductor or solid-state memory, a magnetic tape, a removable computer disk, a random access memory (RAM), a read-only memory (ROM), a hard disk and an optical disk. The optical disk can include a compact disk-read only memory (CD-ROM), a compact disk-read / write (CD-RW) and a DVD.

[0094] In summary, the X-type tail rudder control allocation method based on neural network of the present invention designs and trains a neural network with the desired torque as input and the rudder angle as output, thereby realizing the control allocation of multiple control surfaces. min <u<u max , taking into full account the physical limitations of the control surface, thus overcoming the problem that the solution generated by the pseudo-inverse method may exceed u min <u<u max Compared with the optimization-based method, when performing control allocation, only the trained neural network model needs to be used for regression calculation. There is no loop and iteration process during calculation, and the amount of calculation is small, which is conducive to improving the real-time response capability of the spacecraft.

[0095] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A method for allocating X-type tail rudder control based on neural network, characterized in that: include: S1, generate training samples of X-shaped tail rudder; S2. Training a neural network according to the training samples to obtain a trained neural network model; S3. Taking the expected torque as input, using the trained neural network model, calculate the steering angle combination.

2. The X-type tail rudder control allocation method based on neural network according to claim 1, characterized in that: The training samples for generating the X-shaped tail rudder in step S1 include: S11, obtaining an initial mapping set S0 of the rudder angle combination to the torque of the X-type tail rudder; S12. Screen the initial mapping set S0 to obtain a mapping set S from steering angle combinations to torques.

3. The X-type tail rudder control allocation method based on neural network according to claim 2, characterized in that: The step S11 of obtaining the initial mapping set S0 of the rudder angle combination to the torque of the X-type tail rudder includes: According to the set step size, in u min <u<u max The rudder angle combination and speed are traversed within the range to obtain the torque τ=(K, M, N) when the rudder angle combination u and the speed v are obtained. After traversal, the initial mapping set S0={s|(u, v, K, M, N)} is obtained.

4. The X-type tail rudder control allocation method based on neural network according to claim 3 is characterized by: In the mapping set S of the rudder angle combination to the torque, under the condition of the speed v, the rudder angle combination u and the torque vector (K, M, N) are mapped one to one, and the inverse mapping of the rudder angle combination u to the speed v and the torque vector (K, M, N) is fitted.

5. The X-type tail rudder control allocation method based on neural network according to claim 2, characterized in that: The step S2 of training a neural network according to the training sample to obtain a trained neural network model includes: S21. Design neural network structure; S22. Using the mapping set S of the steering angle combination to the torque to train the neural network structure, so as to obtain a trained neural network model.

6. The method for allocating control of an X-type tail rudder based on a neural network according to claim 5, characterized in that: The design of the neural network structure in step S21 includes: S211, the input layer of the neural network structure includes 4 nodes, corresponding to K, M, N and speed v in the moment vector respectively; S212, the output layer of the neural network structure has n nodes, corresponding to n steering angles u1, u2, u3, ..., u n ; S213, the number of hidden layers of the neural network structure ranges from 1 to 5.

7. The X-type tail rudder control allocation method based on neural network according to claim 6, characterized in that: The step S22 uses the mapping set S of the steering angle combination to the torque to train the neural network structure to obtain a trained neural network model, including: S221, randomly divide the mapping set S into 3 subsets, the proportion of the number of subset elements is The value range of a is 0.5 to 0.8, and the three subsets are used as training set, test set, and validation set respectively; S222. Randomly set the initial parameters of the neural network structure, use the training set, test set, and validation set to perform training, testing, and validation respectively. If the mean square errors of training, testing, and validation are all less than a threshold value t, the training is terminated, and the corresponding neural network structure is used to control the allocation.

8. The method for allocating control of an X-type tail rudder based on a neural network according to claim 1, characterized in that: In step S3, taking the desired torque as input and using the trained neural network model to calculate the steering angle combination includes: When the speed is v, the expected torque τ is (K, M, N), then (K, M, N, v) is used as input, and the trained neural network model is used to calculate the output of the neural network model, and the output u of the neural network model is u = (u1, u2, u3, ..., u n ) as the final control rudder angle.

9. The method for allocating control of an X-type tail rudder based on a neural network according to claim 7, characterized in that: The threshold t ranges from 10 -10 ~10 -6 .

10. An electronic device, comprising a processor and a memory, wherein the memory stores program instructions, characterized in that: The processor runs program instructions to implement the neural network-based X-type tail rudder control allocation method as described in any one of claims 1 to 9.

Images