Disclosure of Invention
The invention aims to provide a carrier rocket load shedding control method and system based on LSTM-FCNN attack angle estimation, which improves the estimation accuracy of the attack angle under the condition of uncertain wind fields, reduces the aerodynamic load of the carrier rocket and improves the flight adaptability of the carrier rocket in a high-altitude wind area.
According to a first aspect of an embodiment of the present disclosure, there is provided a launch vehicle load shedding control method based on LSTM-FCNN attack angle estimation, including the steps of:
Generating a data set according to the flying state data of the rocket in the high-altitude wind area and the measured wind speed data, and processing the data set by adopting linear normalization;
training an LSTM-FCNN deep neural network by using the normalized data, and establishing a mapping relation between rocket attitude response and wind speed;
And migrating the trained LSTM-FCNN deep neural network into the actual flight of the rocket, obtaining an attack angle based on the estimated high altitude wind speed, and further adopting a load shedding control method based on attack angle feedback to reduce the aerodynamic load of the carrier rocket.
According to a second aspect of embodiments of the present disclosure, there is provided a launch vehicle load shedding control system based on LSTM-FCNN angle of attack estimation, comprising:
the data acquisition module is used for generating a data set according to the flight state data of the rocket in the high-altitude area and the measured wind speed data, and processing the data set by adopting linear normalization;
The mapping relation establishing module trains an LSTM-FCNN depth neural network by using the normalized data and establishes a mapping relation between rocket attitude response and wind speed;
And the load shedding control module migrates the trained LSTM-FCNN deep neural network into the actual flight of the rocket, obtains an attack angle based on the estimated high altitude wind speed, and further adopts a load shedding control method based on attack angle feedback to reduce the aerodynamic load of the carrier rocket.
According to a third aspect of the embodiments of the present disclosure, there is provided an electronic device, including a memory, a processor and a computer program running on the memory, where the processor implements the launch vehicle load shedding control method based on LSTM-FCNN attack angle estimation when executing the program.
According to a fourth aspect of embodiments of the present disclosure, there is provided a computer readable storage medium having stored thereon a computer program which when executed by a processor implements the method of launch vehicle load shedding control based on LSTM-FCNN angle of attack estimation.
Compared with the prior art, the technical scheme adopted by the invention has the advantages that: (1) Aiming at the situation that the estimation precision of the attack angle of the carrier rocket is limited, the invention utilizes the LSTM-FCNN depth neural network to establish the mapping relation between rocket attitude response and wind speed, and estimates the attack angle by estimating the wind speed, thereby effectively improving the estimation precision of the uncertainty attack angle of the wind field.
(2) Aiming at the situation that the aerodynamic load is overlarge in the carrier rocket flight process, the invention acquires the attack angle information and then introduces the attack angle information into a control system, so that the aerodynamic load of the carrier rocket can be effectively reduced.
Detailed description of the preferred embodiments
The disclosure is further described below with reference to the drawings and examples.
It should be noted that the following detailed description is illustrative and is intended to provide further explanation of the application. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the present application. As used herein, the singular is also intended to include the plural unless the context clearly indicates otherwise, and furthermore, it is to be understood that the terms "comprises" and/or "comprising" when used in this specification are taken to specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof.
It is noted that the flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of methods and systems according to various embodiments of the present disclosure. It should be noted that each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the logical functions specified in the various embodiments. It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by special purpose hardware-based systems which perform the specified functions or operations, or combinations of special purpose hardware and computer instructions.
Embodiment one:
As shown in fig. 1, the embodiment provides a launch vehicle load shedding control method based on LSTM-FCNN attack angle estimation, which includes the following steps:
the first step: generating a data set according to the flying state data of the rocket in the high-altitude wind area and the measured wind speed data, and processing the data set by adopting linear normalization;
Specifically, in order to avoid larger network prediction error caused by larger order-of-magnitude difference, and to accelerate the training speed of the network and improve the prediction precision, the data normalization method adopted by the invention is linear normalization, and the formula is as follows:
Wherein x' is normalized data, the normalized data range is [0,1], and max (x) and min (x) are the maximum value and the minimum value of the data respectively.
And a second step of: training an LSTM-FCNN deep neural network by using the normalized data, and establishing a mapping relation between rocket attitude response and wind speed; the method specifically comprises the following steps:
Analyzing the mass center movement and the motion around the mass center of the carrier rocket, and determining the input quantity of the LSTM-FCNN depth neural network as follows:
output is as follows
Wherein t0 is the time starting point corresponding to each state quantity in the sequence, tf is the time ending point,Is the pitch angle deviation, delta phi is the yaw angle deviation,As the apparent acceleration deviation in the y-direction,The apparent acceleration deviation in the z direction is Vfx, the x direction wind speed, Vfz, the z direction wind speed, H (t0) the rocket flight level at time t0, and H (tf) the rocket flight level at time tf.
The long-short period neural network LSTM is adopted to process the flight state data and the wind speed data sequence, and the multi-layer fully connected neural network FCNN is combined to conduct feature extraction on the output quantity of the long-short period neural network LSTM, so that a mapping relation which is more accurate and close to the actual situation is obtained, and an activation function between each two layers of neural networks is Relu.
Training the LSTM-FCNN deep neural network by using the normalized data, wherein the training algorithm adopts an Adam algorithm, and continuously adjusts the network weight according to the loss function and the evaluation index until the network converges, thereby completing the training. The loss function during training is a mean square error function MSE, and the evaluation index is a root mean square error function RMSE, which is defined as:
Wherein the method comprises the steps ofFor the neural network predicted value, yi is the true value and n is the number of training samples.
And a third step of: and migrating the trained LSTM-FCNN deep neural network into the actual flight of the rocket, obtaining an attack angle based on the estimated high altitude wind speed, and further adopting a load shedding control method based on attack angle feedback to reduce the aerodynamic load of the carrier rocket.
The trained LSTM-FCNN deep neural network is utilized to estimate wind speed on line, and an attack angle is obtained based on the wind speed, specifically:
converting rocket speed from a launch system to an rocket system:
Wherein the method comprises the steps ofThe method is characterized in that the method is a coordinate transformation matrix from a rocket launching system to an rocket system, Vx、Vy、Vz is the ground speed of the rocket in three directions under the launching coordinate system, and VqxB、VqyB、VqzB is the air speed of the rocket in three directions under the rocket body coordinate system;
Under the rocket system, the attack angle and sideslip angle of the rocket are obtained based on the air velocity, the following are provided:
VqB in the above formula is the empty velocity of the rocket in the rocket body coordinate system, alpha1、β1 is the total attack angle and the total sideslip angle respectively, and the specific expression is as follows:
Where α is the response angle of attack, αw is the airflow angle of attack, β is the response sideslip angle, and βw is the airflow sideslip angle.
After the attack angle is estimated, the attack angle information is introduced into a control system, and a rocket pitch channel control equation is as follows:
Wherein the method comprises the steps ofThe gain factor is fed back for the attitude angle,The gain factor is fed back for the angular velocity of the attitude,For the angle of attack feedback relief gain factor, deltaalpha is the response angle of attack deviation,Is an engine swing angle command.
When considering the effects of the wind, the rocket pitch attitude motion equation can be expressed as:
wherein alphawq is the angle of attack of the airflow generated by the shear wind, the coefficientThe expression is as follows:
Wherein the method comprises the steps ofFor the moment of inertia of the pitch channel,The derivative of the normal aerodynamic coefficient to alpha, q is dynamic pressure, SM is rocket reference cross-sectional area, Xd is distance from the theoretical tip of the rocket body to the pneumatic pressure center, Xz is distance from the theoretical tip of the rocket body to the mass center, P is rocket thrust, XR is distance from the engine swinging point to the theoretical tip of the rocket body, mR is mass of the swinging part of a single engine,For arrow longitudinal apparent acceleration, lR is the distance from the engine centroid to the pendulum shaft.
Further, the total attack angle deviation caused by the tangential wind is obtained by combining the formula (8) and the formula (9):
equation (12) shows that after the estimated attack angle information of the neural network is introduced, the appropriate attack angle information is selectedThe value can reduce the total attack angle, further reduce the aerodynamic load of the rocket body and improve the flight adaptability of the carrier rocket in a high-altitude area.
The method can resist the influence of structural interference force and measurement noise under the condition of uncertain wind fields, and improves the accuracy of attack angle estimation; and the device does not need an additional measuring device, is easy to realize, can effectively reduce the pneumatic load of the rocket, and improves the flight adaptability of the carrier rocket in a high-altitude area.
Embodiment two:
The embodiment provides a carrier rocket load shedding control system based on LSTM-FCNN attack angle estimation, which comprises the following steps:
the data acquisition module is used for generating a data set according to the flight state data of the rocket in the high-altitude area and the measured wind speed data, and processing the data set by adopting linear normalization;
The mapping relation establishing module trains an LSTM-FCNN depth neural network by using the normalized data and establishes a mapping relation between rocket attitude response and wind speed;
And the load shedding control module migrates the trained LSTM-FCNN deep neural network into the actual flight of the rocket, obtains an attack angle based on the estimated high altitude wind speed, and further adopts a load shedding control method based on attack angle feedback to reduce the aerodynamic load of the carrier rocket.
Embodiment III:
an electronic device comprising a memory, a processor and a computer program running on the memory, wherein the processor implements a launch vehicle load shedding control method based on LSTM-FCNN attack angle estimation as described above when executing the program, comprising:
Generating a data set according to the flying state data of the rocket in the high-altitude wind area and the measured wind speed data, and processing the data set by adopting linear normalization;
training an LSTM-FCNN deep neural network by using the normalized data, and establishing a mapping relation between rocket attitude response and wind speed;
And migrating the trained LSTM-FCNN deep neural network into the actual flight of the rocket, obtaining an attack angle based on the estimated high altitude wind speed, and further adopting a load shedding control method based on attack angle feedback to reduce the aerodynamic load of the carrier rocket.
Embodiment four:
a computer readable storage medium having stored thereon a computer program which when executed by a processor implements a launch vehicle load shedding control method based on LSTM-FCNN angle of attack estimation as described above, comprising:
Generating a data set according to the flying state data of the rocket in the high-altitude wind area and the measured wind speed data, and processing the data set by adopting linear normalization;
training an LSTM-FCNN deep neural network by using the normalized data, and establishing a mapping relation between rocket attitude response and wind speed;
And migrating the trained LSTM-FCNN deep neural network into the actual flight of the rocket, obtaining an attack angle based on the estimated high altitude wind speed, and further adopting a load shedding control method based on attack angle feedback to reduce the aerodynamic load of the carrier rocket.
It will be appreciated by those skilled in the art that the modules or steps of the disclosure described above may be implemented in a general-purpose computer device, alternatively they may be implemented in program code executable by a computing device, such that they may be stored in a memory device for execution by the computing device, or they may be separately fabricated into individual integrated circuit modules, or a plurality of modules or steps in them may be fabricated into a single integrated circuit module. The present disclosure is not limited to any specific combination of hardware and software.
The above description is only of the preferred embodiments of the present application and is not intended to limit the present application, but various modifications and variations can be made to the present application by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.
While the specific embodiments of the present disclosure have been described above with reference to the drawings, it should be understood that the present disclosure is not limited to the embodiments, and that various modifications and changes can be made by one skilled in the art without inventive effort on the basis of the technical solutions of the present disclosure while remaining within the scope of the present disclosure.