阅读说明：本技术 超声速混合层混合增强闭环控制方法及系统 (Supersonic mixed layer mixing enhancement closed-loop control method and system ) 是由 谭建国 高政旺 刘瑶 张冬冬 姚霄 肖犇 侯廙 于 2021-06-21 设计创作，主要内容包括：超声速混合层混合增强闭环控制方法及系统,包括第一超声速流道、第二超声速流道、控制器、控制台、速度场测量系统以及计算机,第一超声速流道和第二超声速流道之间设置有混合增强板,混合增强板将第一超声速流道和第二超声速流道分隔开,第一股超声速气流和第二股超声速气流在混合增强板的末端外实现增强混合；控制器通过控制台将激励函数信号作用于混合增强板；速度场测量系统对混合增强板末端的流场速度进行实时测量并将测量到的数据传输给计算机,计算机中预先加载有机器学习算法,通过运行机器学习算法产生激励函数信号给控制器。本发明具有可适应宽范围工作状况、鲁棒性强、混合增强效果显著等特点。(The system comprises a first supersonic flow channel, a second supersonic flow channel, a controller, a console, a speed field measuring system and a computer, wherein a mixing reinforcing plate is arranged between the first supersonic flow channel and the second supersonic flow channel, the mixing reinforcing plate separates the first supersonic flow channel from the second supersonic flow channel, and the first supersonic air flow and the second supersonic air flow realize reinforced mixing outside the tail end of the mixing reinforcing plate; the controller acts the excitation function signal on the hybrid enhancement board through the console; the speed field measurement system measures the flow field speed at the tail end of the hybrid reinforced plate in real time and transmits the measured data to the computer, the computer is pre-loaded with a machine learning algorithm, and an excitation function signal is generated to the controller by running the machine learning algorithm. The invention has the characteristics of adaptability to wide-range working conditions, strong robustness, obvious mixing enhancement effect and the like.)

1. The supersonic mixed layer mixing enhancement closed-loop control method is characterized by comprising the following steps:

s1, constructing a supersonic mixed layer hybrid enhanced closed-loop control system;

the supersonic mixed layer hybrid enhanced closed-loop control system comprises a first supersonic flow channel, a second supersonic flow channel, a controller, a control console, a speed field measurement system and a computer, wherein a hybrid enhancement plate is arranged between the first supersonic flow channel and the second supersonic flow channel, the hybrid enhancement plate separates the first supersonic flow channel from the second supersonic flow channel, a first supersonic airflow flows to the tail end of the hybrid enhancement plate along the length direction of the hybrid enhancement plate above the hybrid enhancement plate, a second supersonic airflow flows to the tail end of the hybrid enhancement plate along the length direction of the hybrid enhancement plate below the hybrid enhancement plate, and the first supersonic airflow and the second supersonic airflow are enhanced and mixed outside the tail end of the hybrid enhancement plate; the controller is connected with the console, and the console is controlled to act on the hybrid reinforcing plate at the tail section of the hybrid reinforcing plate close to the tail end of the hybrid reinforcing plate by sending an excitation function signal to the console, so that the hybrid reinforcing plate vibrates; the speed field measurement system measures the flow field speed at the tail end of the hybrid reinforced plate in real time and transmits the measured data to the computer, the computer is pre-loaded with a machine learning algorithm, an excitation function signal is generated by running the machine learning algorithm and is sent to the controller, and the excitation function signal acts on the hybrid reinforced plate through the console;

s2, initializing a plurality of excitation function signals through a machine learning algorithm pre-loaded in a computer to serve as first generation excitation function signals;

s3, the controller acts the current generation of excitation function signals on the hybrid reinforcing plate through the console, and meanwhile, the speed measuring unit measures the flow field speed outside the tail end of the hybrid reinforcing plate under each excitation function signal in real time and transmits the flow field speed to the computer;

s4, defining a cost function by using the flow field speed, calculating the size of the cost function corresponding to the excitation function signal of the current generation, and evaluating the cost of each excitation function signal of the current generation; then, the current generation excitation function signal is subjected to updating operation through copying, hybridization or/and mutation to generate a new generation excitation function signal;

s5, repeating S3 to S4 until a preset stop condition is reached.

2. The hybrid enhanced closed-loop control method of the supersonic hybrid layer as claimed in claim 1, wherein the velocity field measurement system is a PIV system, and the PIV system is used to take a snapshot of the velocity field of the flow field outside the end of the hybrid enhanced plate and transmit the snapshot to a computer.

3. The supersonic hybrid layer hybrid enhancement closed-loop control method according to claim 1 or 2, wherein in step S5, the machine learning algorithm adopts a linear genetic programming algorithm.

4. The hybrid enhancement closed-loop control method for the supersonic hybrid layer according to claim 1, wherein the form of the excitation function signal b is:

b＝K(s(t),h(t))

where s (t) is the flow field velocity at time t, h (t) is a harmonic signal with respect to time t, and h (t) is defined as the natural frequency f_uOr natural frequency f_uHalf or natural frequency f_uTwice of (c):

h(t)＝[cos(2πf_ut) cos(πf_ut) cos(2πf_ut) cos(2πf_ut) cos(2πf_ut)]。

5. the hybrid enhanced closed-loop control method of the supersonic hybrid layer according to claim 1, wherein in step S4, the pulsation energy of the hybrid layer is evaluated by calculating the pulsation velocity using the flow field velocity, and the pulsation velocity is obtained by moving average calculation, and the formula is as follows:

s′(t)＝s(t)-<s(t)>_τ

where s (t) is the flow field velocity at time t, τ is the moving average time,

the average accumulated pulse energy of each excitation function signal in a certain time T is defined as:

the cost function defining the hybrid enhancement is:

J＝1/K

and calculating to obtain a cost function corresponding to each excitation function signal of the current generation.

6. The hybrid enhancement closed-loop control method for supersonic hybrid layers according to claim 5, wherein in step S5, the preset stop conditions are: when the machine learning algorithm does not generate smaller cost function any more, the excitation function signal corresponding to the minimum cost function at the moment is the optimal excitation function signal for controlling and mixing enhancement of the supersonic mixing layer.

7. The supersonic hybrid layer mixing enhancement closed-loop control method of claim 1, wherein the console employs a VR9500 vibration controller.

8. Mixed reinforcing closed loop control system of supersonic speed mixing layer, its characterized in that: the device comprises a first supersonic flow channel, a second supersonic flow channel, a controller, a console, a speed field measuring system and a computer, wherein a mixing reinforcing plate is arranged between the first supersonic flow channel and the second supersonic flow channel, the mixing reinforcing plate separates the first supersonic flow channel from the second supersonic flow channel, a first supersonic flow flows to the tail end of the mixing reinforcing plate along the length direction of the mixing reinforcing plate above the mixing reinforcing plate, a second supersonic flow flows to the tail end of the mixing reinforcing plate along the length direction of the mixing reinforcing plate below the mixing reinforcing plate, and the first supersonic flow and the second supersonic flow realize reinforced mixing outside the tail end of the mixing reinforcing plate;

the controller is connected with the console, and the console acts on the hybrid reinforcing plate at the tail section of the hybrid reinforcing plate close to the tail end of the hybrid reinforcing plate by sending an excitation function signal to the console, so that the hybrid reinforcing plate vibrates, and the velocity field measurement system measures the velocity of a flow field outside the tail end of the hybrid reinforcing plate in real time and transmits measured data to the computer;

the computer is pre-loaded with a machine learning algorithm, calculates the size of a cost function corresponding to the current excitation function signal according to the flow field speed, generates a new excitation function signal to the controller, acts on the hybrid enhanced plate through the console until no smaller cost function is generated, and the excitation function signal corresponding to the minimum cost function, namely the supersonic hybrid layer, controls the hybrid enhanced optimal excitation function signal.

9. The system of claim 8, wherein the console employs a VR9500 vibration controller, the velocity field measurement system is a PIV system, and the PIV system is used to take a snapshot of the velocity field of the flow field outside the end of the hybrid enhancement plate and transmit the snapshot to a computer.

10. The supersonic hybrid layer hybrid enhancement closed-loop control system of claim 8, wherein the machine learning algorithm employs a linear genetic programming algorithm.

Technical Field

The invention relates to the technical field of hybrid enhancement control in the field of aerospace, in particular to a supersonic mixed layer hybrid enhancement closed-loop control method and system, which can be used for optimizing the hybrid enhancement of fuel and oxidant in a combined cycle engine.

Background

Mixed layer flow is a typical type of flow in engineering applications, such as the mixing of core fuel gas with bypass air in turbofan engines, such as the mixing of rocket fuel gas with ram air in rocket-based combined cycle engines. Efficient mixing to improve combustion efficiency is one of the key technologies for improving the performance of such engines, and the ways of controlling mixing enhancement can be mainly divided into two categories: passive mixing enhancement and active mixing enhancement.

Passive mixing enhancement is usually achieved by installing some fixing devices in the flow field or changing structural parameters to change the stability characteristics of the shear layer of the flow field. The mode of action can be subdivided into two categories, 1) promotion of flow destabilization in advance, typically as in the case of triangulated perturbation devices and reentrant devices; 2) inducing large-scale flow to the vortex structure, the common devices are a lobe mixer, a slope device, a sawtooth device, a V-shaped device and the like. However, the passive control device cannot be adjusted according to the working condition, and the self-adaptability is poor.

The active mixing enhancement is to apply external energy to a flow field through an exciter, excite flow instability, adjust a vortex structure in the flow field of a mixing layer, promote mixing action among different fluids and further realize mixing enhancement. Active control can be divided into open-loop control and closed-loop control according to whether a feedback mechanism exists in the system or not.

The open-loop control system works according to a preset action mode, does not form feedback control, and is simple and easy to operate. According to different excitation modes, open-loop control mixing enhancement measures such as flat plate disturbance, plasma excitation, synthetic jet and the like can be divided. The frequency, amplitude and location of the excitation have a significant effect on the hybrid layer growth characteristics. But open-loop control systems do not respond in real time to optimize external excitation based on the state of the flow field.

The closed-loop control hybrid enhancement technology is particularly important when the state of a flow field changes, can reduce the sensitivity to system change parameters and external disturbance, and has strong robustness. However, for turbulent flow with high-dimensional non-linear characteristics of supersonic mixing layers, the closed-loop control of the turbulent flow by a classical control method still faces great difficulty.

Disclosure of Invention

Aiming at the defects in the prior art, the invention provides a supersonic mixed layer mixing enhancement closed-loop control method and system, which have the characteristics of adaptability to wide-range working conditions, strong robustness, obvious mixing enhancement effect and the like. The mixing enhancement of the supersonic mixing layer in the engine can be realized to the maximum extent.

In order to achieve the technical purpose, the technical scheme provided by the invention is as follows:

the invention provides a supersonic mixed layer mixing enhancement closed-loop control method, which comprises the following steps:

s1, constructing a supersonic mixed layer hybrid enhanced closed-loop control system;

the supersonic mixed layer hybrid enhanced closed-loop control system comprises a first supersonic flow channel, a second supersonic flow channel, a controller, a control console, a speed field measurement system and a computer, wherein a hybrid enhancement plate is arranged between the first supersonic flow channel and the second supersonic flow channel, the hybrid enhancement plate separates the first supersonic flow channel from the second supersonic flow channel, a first supersonic airflow flows to the tail end of the hybrid enhancement plate along the length direction of the hybrid enhancement plate above the hybrid enhancement plate, a second supersonic airflow flows to the tail end of the hybrid enhancement plate along the length direction of the hybrid enhancement plate below the hybrid enhancement plate, and the first supersonic airflow and the second supersonic airflow are enhanced and mixed outside the tail end of the hybrid enhancement plate; the controller is connected with the console, and the console is controlled to act on the hybrid reinforcing plate at the tail section of the hybrid reinforcing plate close to the tail end of the hybrid reinforcing plate by sending an excitation function signal to the console, so that the hybrid reinforcing plate vibrates; the speed field measurement system measures the flow field speed at the tail end of the hybrid reinforced plate in real time and transmits the measured data to the computer, the computer is pre-loaded with a machine learning algorithm, an excitation function signal is generated by running the machine learning algorithm and is sent to the controller, and the excitation function signal acts on the hybrid reinforced plate through the console;

s2, initializing a plurality of excitation function signals through a machine learning algorithm pre-loaded in a computer to serve as first generation excitation function signals;

s3, the controller acts the current generation of excitation function signals on the hybrid reinforcing plate through the console, and meanwhile, the speed measuring unit measures the flow field speed of the tail end of the hybrid reinforcing plate under each excitation function signal in real time and transmits the flow field speed to the computer;

s4, defining a cost function by using the flow field speed, calculating the size of the cost function corresponding to the current generation excitation function signal, evaluating the cost of each excitation function signal of the current generation, and then performing updating operation on the current generation excitation function signal through copying, hybridization or/and variation to generate a new generation excitation function signal;

s5, repeating S3 to S4 until a preset stop condition is reached.

In step S5, S3 to S4 are repeated, and when there is no smaller cost function generated in the machine learning algorithm, it indicates that the algorithm has converged, and at this time, the excitation function signal corresponding to the minimum cost function is the optimal excitation function signal for the supersonic hybrid layer control hybrid enhancement.

As a preferable scheme of the invention, the machine learning algorithm in the invention adopts a linear genetic programming algorithm.

As a preferred embodiment of the present invention, the form of the excitation function signal b in the present invention is:

b＝K(s(t),h(t))

where s (t) is the flow field velocity at time t, h (t) is the harmonic signal with respect to time t, where h (t) is defined as the natural frequency f_uOr natural frequency f_uHalf or nature ofFrequency f_uTwo times, etc.:

h(t)＝[cos(2πf_ut) cos(πf_ut) cos(2πf_ut) cos(2πf_ut) cos(2πf_ut)]

the harmonic signals here contain sine and cosine functions to form phase differences of the excitation function signals, which phase differences can have a significant influence on the hybrid layer.

In the invention S4, the pulsation energy of the mixed layer is evaluated by calculating the pulsation velocity by using the flow field velocity, the pulsation velocity is obtained by moving average calculation, and the formula is as follows:

s′(t)＝s(t)-<s(t)>_τ

where s (t) is the flow field velocity at time t, τ is the moving average time,

the average accumulated pulse energy of each excitation function signal in a certain time T is defined as:

to achieve hybrid layer mixing enhancement, i.e., promote instability of the hybrid layer, the corresponding impulse energy K increases, thus defining the cost function of mixing enhancement as:

J＝1/K

and calculating to obtain a cost function corresponding to each excitation function signal of the current generation, and performing updating operations such as copying, mutation, hybridization and the like according to a preset proportion by using a machine learning algorithm according to the size of the cost function corresponding to each excitation function signal of the current generation to generate a new generation of excitation function signals.

The invention provides a supersonic mixing layer mixed enhanced closed-loop control system, which comprises a first supersonic flow channel, a second supersonic flow channel, a controller, a console, a speed field measuring system and a computer, wherein a mixed enhanced plate is arranged between the first supersonic flow channel and the second supersonic flow channel, the mixed enhanced plate separates the first supersonic flow channel from the second supersonic flow channel, a first supersonic airflow flows to the tail end of the mixed enhanced plate above the mixed enhanced plate along the length direction of the mixed enhanced plate, a second supersonic airflow flows to the tail end of the mixed enhanced plate below the mixed enhanced plate along the length direction of the mixed enhanced plate, and the first supersonic airflow and the second supersonic airflow realize enhanced mixing outside the tail end of the mixed enhanced plate;

the controller is connected with the console, and the console acts on the hybrid reinforcing plate at the tail section of the hybrid reinforcing plate close to the tail end of the hybrid reinforcing plate by sending an excitation function signal to the console, so that the hybrid reinforcing plate vibrates, and the velocity field measurement system measures the velocity of a flow field outside the tail end of the hybrid reinforcing plate in real time and transmits measured data to the computer;

the computer is pre-loaded with a machine learning algorithm, calculates the size of a cost function corresponding to the current excitation function signal according to the flow field speed, generates a new excitation function signal to the controller, acts on the hybrid enhanced plate through the console until no smaller cost function is generated, and the excitation function signal corresponding to the minimum cost function, namely the supersonic hybrid layer, controls the hybrid enhanced optimal excitation function signal.

Compared with the prior art, the invention has the advantages that:

1. the supersonic mixed layer hybrid reinforced closed-loop control scheme provided by the invention has high adaptive degree and strong robustness. The invention overcomes the defect that an open-loop control system can not react and optimize external excitation in real time according to the state of a flow field.

2. The invention obtains the optimal excitation function signal through the global search of the machine learning algorithm, can realize the mixing enhancement of the mixing layer to the maximum extent, and avoids the limitation of manually set frequency and limited amplitude.

3. The machine learning is a model-free modeling method based on data driving, complex causal relationship does not need to be considered, and prior knowledge of any fluid mechanics and nonlinear dynamics is not needed when the method is applied to supersonic mixing layer control. Therefore, the method can be popularized to more high-dimensional nonlinear control systems.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments or the prior art of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for the ordinary skill in the art, other drawings can be obtained according to these drawings without creative efforts.

FIG. 1 is a schematic block diagram of a supersonic hybrid layer hybrid enhanced closed-loop control system according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of the operation of a linear genetic programming algorithm employed in one embodiment of the present invention;

the objects, features, and advantages of the present invention will be further explained with reference to the accompanying drawings.

Detailed Description

For the purpose of promoting a clear understanding of the objects, aspects and advantages of the embodiments of the invention, reference will now be made to the drawings and detailed description, wherein there are shown in the drawings and described in detail, various modifications of the embodiments described herein, and other embodiments of the invention will be apparent to those skilled in the art. The exemplary embodiments of the present invention and the description thereof are provided to explain the present invention and not to limit the present invention.

An embodiment of the present invention provides a hybrid enhanced closed-loop control method for a supersonic hybrid layer, including the following steps:

s1, constructing a supersonic mixed layer hybrid enhanced closed-loop control system;

referring to fig. 1, the supersonic mixed layer hybrid enhanced closed-loop control system includes a first supersonic flow channel 1, a second supersonic flow channel 2, a hybrid enhancement plate 3, a controller 4, a console 5, a velocity field measurement system 6 and a computer 7, the hybrid enhancement plate 3 is disposed between the first supersonic flow channel 1 and the second supersonic flow channel 2, in this embodiment, the hybrid enhancement plate 3 is a thin flat plate disposed between the first supersonic flow channel 1 and the second supersonic flow channel 2, the hybrid enhancement plate 3 separates the first supersonic flow channel 1 from the second supersonic flow channel 2, a first supersonic flow flows to the end of the hybrid enhancement plate 3 along the length direction of the hybrid enhancement plate 3 above the hybrid enhancement plate 3, a second supersonic flow flows to the end of the hybrid enhancement plate 3 along the length direction of the hybrid enhancement plate 3 below the hybrid enhancement plate 3, and the first supersonic flow and the second supersonic flow enhance the hybrid enhancement outside the end of the hybrid enhancement plate In this embodiment, the gas in the first supersonic gas flow is fuel gas such as methane, and the gas in the second supersonic gas flow is air.

The controller 4 is connected to the console 5, and by sending an excitation function signal b to the console 5, the console 5 acts on the hybrid enhancement plate 3 at the end section of the hybrid enhancement plate 3 near its end, causing the hybrid enhancement plate 3 to vibrate. The console 5 adopts a VR9500 vibration controller.

The velocity field measuring system 6 is a PIV system which monitors the flow field velocity field at the tail end of the hybrid reinforced plate in real time, shoots the flow field velocity field snapshot and feeds the snapshot back to the computer 7. The computer 7 is pre-loaded with a machine learning algorithm, and generates an excitation function signal to the controller by running the machine learning algorithm, and acts on the hybrid enhancement board through the console.

And S2, initializing a plurality of excitation function signals through a machine learning algorithm pre-loaded in a computer to serve as first generation excitation function signals.

The first generation excitation function signals are randomly generated by a method similar to the monte carlo, and the total number of the first generation excitation function signals is set to be N. The form of the excitation function signal b in this embodiment is intended to be:

b＝K(s(t),h(t))

where s (t) is the flow field velocity at time t, h (t) is the harmonic signal related to time t, and h (t) can be defined as the natural frequency f_uOr natural frequency f_uHalf or natural frequency f_uTwo times, etc.:

h(t)＝[cos(2πf_ut) cos(πf_ut) cos(2πf_ut) cos(2πf_ut) cos(2πf_ut)]

the harmonic signals here contain sine and cosine functions to form phase differences of the signals, which phase differences can have a significant influence on the hybrid layer.

And S3, the controller acts the current generation of excitation function signals on the hybrid reinforcing plate through the console, and meanwhile, the speed measuring unit measures the flow field speed of the tail end of the hybrid reinforcing plate under each excitation function signal in real time and transmits the flow field speed to the computer.

S4, defining a cost function by using the flow field speed, calculating the size of the cost function corresponding to the current generation excitation function signal, evaluating the cost of each excitation function signal of the current generation, and then performing updating operation on the current generation excitation function signal through copying, hybridization or/and variation to generate a new generation excitation function signal;

the cost function is used for evaluating the cost corresponding to the excitation function, and the flow state of the mixed layer is closely related to the energy carried by the mixed layer. The invention utilizes the flow field velocity collected by the velocity measuring unit to calculate the pulsation velocity to evaluate the pulsation energy of the mixed layer, the pulsation velocity is obtained by moving average calculation, and the formula is as follows:

s′(t)＝s(t)-<s(t)>_τ

where s (t) is the flow field velocity at time t, τ is the moving average time,

the average accumulated pulse energy of each excitation function signal in a certain time T is defined as:

here T is the time range of flow field velocity measurement during the experiment for each excitation function, which is sufficient to eliminate the occasional error in velocity and ensure convergence of K.

To achieve hybrid layer mixing enhancement, i.e., promote instability of the hybrid layer, the corresponding impulse energy K increases, thus defining the cost function of mixing enhancement as:

J＝1/K

and calculating to obtain a cost function corresponding to each excitation function signal of the current generation, and performing updating operations such as copying, mutation, hybridization and the like according to a preset proportion by using a machine learning algorithm according to the size of the cost function corresponding to each excitation function signal of the current generation to generate a new generation of excitation function signals.

S5, repeating S3 to S4 until a preset stop condition is reached.

The new generation of excitation function signals are applied to the hybrid enhancement plate through the console, and the speed measurement unit measures the flow field speed of the tail end of the hybrid enhancement plate under each new generation of excitation function signal in real time and transmits the flow field speed to the computer … … to repeat the process. The preset stop condition in this embodiment indicates that the algorithm has converged when there is no smaller cost function generated in the machine learning algorithm any more, and at this time, the excitation function signal corresponding to the minimum cost function is the optimal excitation function signal for controlling hybrid enhancement by the supersonic hybrid layer.

The machine learning algorithm adopted in one embodiment of the invention is a linear genetic programming algorithm. Fig. 2 is a schematic diagram of a working process of the linear genetic programming algorithm in an embodiment of the present invention, in which a plurality of excitation function signals are initialized randomly as first generation excitation function signals, each excitation function signal is evaluated by a cost function, and then a new generation excitation function signal is generated by operations such as copying, hybridization, and mutation, and the operations are repeated until a preset stop condition is reached.

Referring to fig. 1, the present embodiment provides a supersonic velocity mixing layer hybrid enhanced closed loop control system, referring to fig. 1, the supersonic velocity mixing layer hybrid enhanced closed loop control system includes a first supersonic velocity flow channel 1, a second supersonic velocity flow channel 2, a hybrid enhancing plate 3, a controller 4, a console 5, a velocity field measurement system 6, and a computer 7, a hybrid enhancing plate 3 is disposed between the first supersonic velocity flow channel 1 and the second supersonic velocity flow channel 2, in the present embodiment, the hybrid enhancing plate 3 is a thin flat plate disposed between the first supersonic velocity flow channel 1 and the second supersonic velocity flow channel 2, the hybrid enhancing plate 3 separates the first supersonic velocity flow channel 1 from the second supersonic velocity flow channel 2, a first supersonic velocity air flow flows to a terminal of the hybrid enhancing plate 3 along a length direction of the hybrid enhancing plate 3 above the hybrid enhancing plate 3, a second supersonic velocity air flow flows to a terminal of the hybrid enhancing plate 3 along a length direction of the hybrid enhancing plate 3 below the hybrid enhancing plate 3, the first supersonic air flow and the second supersonic air flow realize enhanced mixing outside the tail end of the mixing enhancement plate, in the embodiment, the gas in the first supersonic air flow is fuel gas such as methane, and the gas in the second supersonic air flow is air.

The controller 4 is connected with the console 5, and by sending an excitation function signal to the console 5, the console 5 acts on the hybrid reinforcing plate at the tail section of the hybrid reinforcing plate close to the tail end of the hybrid reinforcing plate, so that the hybrid reinforcing plate vibrates, and the velocity field measuring system 6 measures the velocity of a flow field outside the tail end of the hybrid reinforcing plate in real time and transmits the measured data to the computer;

the computer 7 is pre-loaded with a machine learning algorithm, calculates the magnitude of a cost function corresponding to the current excitation function signal according to the flow field speed, generates a new excitation function signal to the controller, and acts on the hybrid enhancement plate through the console until no smaller cost function is generated, and the excitation function signal corresponding to the minimum cost function, namely the supersonic hybrid layer, controls the hybrid enhanced optimal excitation function signal.

The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.