A method and system for predicting a dynamic stall flow regime of an airfoil

Abstract

The application discloses a wing dynamic stall flow state prediction method and system, and relates to the flow state prediction field; the method comprises the following steps: calculating a dimensionless pressure coefficient at a current time according to a pressure signal of an upper surface of a wing in a dynamic stall state at the current time; calculating a flow state index at the current time according to the dimensionless pressure coefficient at the current time, and then determining an index normalization value at the current time; determining an actual flow state at the current time according to the index normalization value at the current time and a predicted flow state at the current time; and predicting a flow state at a next time according to the actual flow state at the current time and a peak value detection function, so as to obtain a predicted flow state at the next time; the flow state index is calculated in real time by determining the pressure coefficient, and then the flow state is predicted in real time, so that the prediction of the flow state can be realized simply and efficiently.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of flow state prediction, in particular to a wing dynamic stall flow state prediction method and system. BACKGROUND

[0002] Due to the wide existence of dynamic stall in nature and engineering applications, and the urgent need for flow control, predicting the flow state of dynamic stall has become a significant and long-attention technology. Considering the important influence of the leading edge in the process of airfoil stall, in the early methods, many studies on dynamic stall prediction or modeling focused on seeking indicators of large-scale flow separation near the leading edge. Including using the minimum value of the leading edge surface pressure coefficient, using the corresponding relationship between the leading edge separation and the leading edge flow velocity, using the relationship between the critical leading edge pressure and the critical normal force coefficient, and using the laminar separation of the leading edge to indicate the start of dynamic stall. However, due to the complexity of dynamic stall, the early prediction methods are very limited, and there are many deficiencies in the accuracy of prediction, real-time, parameter universality, generality of different flow stages, and applicability on real wings.

[0003] In recent years, the flow state prediction method of dynamic stall has been further developed, mainly including using flow state indicators for prediction, using reduced order models for prediction, and using machine learning for prediction. The leading edge suction parameter constructed based on the criticality of the leading edge flow can successfully predict the generation of the leading edge vortex, can be obtained based on the real-time measurement of the pressure sensor installed on the leading edge of the wing, and its critical value has weak dependence on kinematic parameters, so it has received extensive attention and is the most commonly used dynamic stall flow state indicator. However, the leading edge suction parameter still has some deficiencies. For example:

[0004] i) The leading edge suction parameter can only predict the generation of the leading edge vortex, but cannot predict the generation of the laminar separation bubble, the shedding of the leading edge vortex, and other events, so it cannot be used to predict the flow stage of dynamic stall.

[0005] ii) In a wide range of parameters, the leading edge suction parameter changes with kinematic parameters, as well as airfoil shape, Reynolds number, etc., which cannot be ignored, so its application range is limited by parameters.

[0006] iii) The leading edge suction parameter only considers the flow situation near the leading edge, and does not consider the influence of the rear separation and other flow phenomena, so its capture ability for more complex flow phenomena is lacking, and it cannot effectively depict many physical phenomena in viscous flow.

[0007] iv) The calculation of the leading edge suction parameter requires the accurate shape of the leading edge of the airfoil to be known, so it cannot be applied to flexible wings with deformable leading edges. In addition to the leading edge suction parameter, the boundary vorticity flux is also a newly proposed flow state indicator, but the calculation of this indicator requires the vorticity near the wing to be known, which is usually difficult to measure directly, so the boundary vorticity flux only has theoretical significance and cannot be practically applied.

[0008] In terms of dynamic stall flow state prediction, in addition to using flow state indicators based on physical relationships, some data-driven low-order models can also be used to achieve this goal. Since it is difficult to directly describe the flow separation in dynamic stall using mathematical models, many researchers have proposed semi-empirical methods that use direct measurements to supplement the flow information in the basic model. These basic models, which can be referred to as dynamic templates, capture many basic physical relationships, but one or more parameters in them need to be determined through measurements. However, these models also have many defects. For example:

[0009] i) These models are usually based on empirical or semi-empirical relationships and the construction process is usually complex, so their applicability is generally limited and they are generally only applicable to specific airfoils and motions.

[0010] ii) The theoretical basis of these models is mostly based on inviscid potential flow theory, so they lack the ability to describe many phenomena related to viscosity, such as flow separation, which can lead to distorted prediction results.

[0011] iii) These models are basically for flat plates or fixed shape airfoils, so they are not suitable for flexible deformable wings, gust flow fields, etc.

[0012] iv) Some of these models require a large amount of computation, which can greatly limit real-time performance.

[0013] In addition to using low-order flow state models, machine learning and other technologies can also be used to achieve this goal. However, machine learning usually requires large amounts of data for training, and the accuracy of its prediction results depends on the selection of the training set, so there are still many shortcomings in terms of universality, and it is still in the development stage and has a long way to go before it is practical. SUMMARY

[0014] The purpose of the present application is to provide a wing dynamic stall flow state prediction method and system, which can predict the flow state simply and efficiently by determining the pressure coefficient to calculate the flow state indicator in real time.

[0015] To achieve the above purpose, the present application provides the following solutions:

[0016] A method for predicting a dynamic stall flow state of an airfoil, the method comprising:

[0017] obtaining a pressure signal of an upper surface of the airfoil in a dynamic stall state at a current time;

[0018] calculating a dimensionless pressure coefficient at the current time according to the pressure signal at the current time;

[0019] calculating a flow state indicator at the current time according to the dimensionless pressure coefficient at the current time; the flow state indicator comprises a pressure spatial distribution coefficient, a pressure high-order central moment, a negative pressure peak value position, and a modulated negative pressure peak value position;

[0020] determining an indicator normalized value at the current time according to the flow state indicator at the current time and a peak detection function;

[0021] determining an actual flow state at the current time according to the indicator normalized value at the current time and a predicted flow state at the current time; the flow state comprises a laminar separation bubble generation, a leading edge vortex generation, a leading edge vortex position, and a leading edge vortex shedding;

[0022] predicting a flow state at a next time according to the actual flow state at the current time and the peak detection function, to obtain a predicted flow state at the next time.

[0023] Optionally, the obtaining of the pressure signal of the upper surface of the airfoil in the dynamic stall state at the current time specifically comprises:

[0024] obtaining an analog voltage signal of the upper surface of the airfoil in the dynamic stall state at the current time;

[0025] performing analog-digital conversion on the analog voltage signal at the current time to obtain a digital voltage signal at the current time;

[0026] performing filtering processing on the digital voltage signal at the current time to obtain a filtered digital voltage signal at the current time;

[0027] determining the pressure signal at the current time according to a set conversion coefficient and the filtered digital voltage signal at the current time.

[0028] Optionally, the calculation formula of the dimensionless pressure coefficient is:

[0029]

[0030] wherein, p ∞ is a dynamic pressure of a free incoming flow; U ∞ is a velocity of the free incoming flow; p is the pressure signal; and C p is the dimensionless pressure coefficient.

[0031] The embodiment of the present application also provides a wing dynamic stall flow state prediction system, and the system comprises:

[0032] a signal acquisition module, configured to acquire a pressure signal of an upper surface of a wing in a dynamic stall state at a current time;

[0033] a pressure coefficient calculation module, configured to calculate a dimensionless pressure coefficient at the current time according to the pressure signal at the current time;

[0034] a flow state index calculation module, configured to calculate a flow state index at the current time according to the dimensionless pressure coefficient at the current time; the flow state index comprises a pressure spatial distribution coefficient, a pressure high-order central moment, a negative pressure peak value position and a modulated negative pressure peak value position;

[0035] a determination module, configured to determine an index normalized value at the current time according to the flow state index at the current time and a peak detection function;

[0036] a flow state determination module, configured to determine an actual flow state at the current time according to the index normalized value at the current time and a predicted flow state at the current time; the flow state comprises a laminar separation bubble generation, a leading edge vortex generation, a leading edge vortex position and a leading edge vortex shedding;

[0037] a prediction module, configured to predict a flow state at a next time according to the actual flow state at the current time and the peak detection function, so as to obtain a predicted flow state at the next time.

[0038] Optionally, the signal acquisition module specifically comprises:

[0039] a voltage signal acquisition sub-module, configured to acquire an analog voltage signal of the upper surface of the wing in the dynamic stall state at the current time;

[0040] a conversion sub-module, configured to perform analog-digital conversion on the analog voltage signal at the current time, so as to obtain a digital voltage signal at the current time;

[0041] a filtering processing sub-module, configured to perform filtering processing on the digital voltage signal at the current time, so as to obtain a filtered digital voltage signal at the current time;

[0042] a pressure signal determination sub-module, configured to determine the pressure signal at the current time according to a set conversion coefficient and the filtered digital voltage signal at the current time.

[0043] Optionally, a calculation formula of the dimensionless pressure coefficient in the pressure coefficient calculation module is as follows:

[0044]

[0045] wherein, p ∞Dynamic pressure of free stream; U ∞ Velocity of free stream; p is density of free stream; p is pressure signal; C p Dimensionless pressure coefficient.

[0046] Optionally, the voltage signal acquisition submodule comprises a pressure measurement sensor and a data acquisition card.

[0047] According to the specific embodiments provided by the present application, the following technical effects are disclosed:

[0048] The present application provides a wing dynamic stall flow state prediction method and system, which first calculates the dimensionless pressure coefficient at the current time according to the obtained pressure signal at the current time, then calculates the flow state index at the current time, and further determines the actual flow state at the current time; according to the actual flow state at the current time and the peak detection function, the flow state at the next time is predicted to obtain the predicted flow state at the next time; since the present application predicts the flow state at the next time through the actual flow state at the current time, it provides a forward-looking prediction for the dynamic stall flow control; and since the present application determines the pressure coefficient to calculate the flow state index in real time, and further predicts the flow state in real time, the problems of lack of universality and distorted prediction results caused by model construction or training are avoided, so that the present application can simply and efficiently realize the prediction of flow state. BRIEF DESCRIPTION OF DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0050] Figure 1 Flow chart of wing dynamic stall flow state prediction method provided by the embodiment of the present application;

[0051] Figure 2 Schematic diagram of flow attachment and dynamic stall provided by the embodiment of the present application;

[0052] Figure 3 Dynamic stall division stage schematic diagram provided by the embodiment of the present application;

[0053] Figure 4 Flow chart of predicting the impending flow event provided by the embodiment of the present application;

[0054] Figure 5 High-frequency pressure sensor distribution schematic diagram provided by the embodiment of the present application;

[0055] Figure 6 A schematic diagram illustrating the changing trend of flow state indicators with convection time under dynamic stall conditions provided in an embodiment of the present invention;

[0056] Figure 7 This is a schematic diagram showing the changes in the negative pressure peak position index and the leading edge vortex core position with convection time during dynamic stall conditions provided in this embodiment of the invention.

[0057] Figure 8 A schematic diagram illustrating the specific process of predicting dynamic stall flow state provided in this embodiment of the invention in practical applications;

[0058] Figure 9 This is a structural diagram of the wing dynamic stall flow state prediction system provided in an embodiment of the present invention.

[0059] Symbol explanation:

[0060] Signal acquisition module-1, pressure coefficient calculation module-2, flow state index calculation module-3, determination module-4, flow state determination module-5, prediction module-6. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] The purpose of this invention is to provide a method and system for predicting the dynamic stall flow state of an airfoil. By determining the pressure coefficient, the flow state index is calculated in real time, thereby predicting the flow state in real time, which can achieve flow state prediction simply and efficiently.

[0063] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0064] Dynamic stall is an unsteady fluid dynamics phenomenon that occurs on the surface of an airfoil or blade, involving many complex interactions, and is caused by rapid motion of the airfoil or sudden changes in the incoming flow conditions. Strictly speaking, dynamic stall refers to the phenomenon of unsteady separation of the fluid boundary layer on the airfoil or blade when the time scale of the airfoil or blade motion is of the same order or faster than the time scale of the flow. On the surface of an aircraft airfoil, a helicopter rotor blade, a wind turbine blade, an engine blade, etc. that undergoes rapid motion or gust disturbance, once dynamic stall occurs, the fluid flow originally attached to the surface of the airfoil or blade will separate from the surface and enter the space to form a shear layer, which further rolls up to form a strong vortex, called a leading edge vortex, also known as a dynamic stall vortex. The schematic diagram of flow attachment and dynamic stall is shown in FIG. 1. Figure 2 . Figure 2 A in FIG. 1 refers to the state of flow attachment; Figure 2 B in FIG. 1 refers to the occurrence of dynamic stall and the generation of a leading edge vortex. Generally, the leading edge vortex will cause stall delay and transient high lift, but its movement will cause rapid changes in the pitch moment, which may lead to negative damping and induce aero-structural coupled flutter. The shedding of the leading edge vortex will further lead to large-scale flow separation on the airfoil surface and the corresponding drop in lift.

[0065] In nature, the flapping flight of birds and the motion of various artificial rotor mechanisms, the wing / airfoil / blade usually performs complex motion with multiple degrees of freedom, even coupled with the deformation of flexible structures, which leads to a very complex mechanism of flow separation and stall. Therefore, in order to simplify the complexity of the research and clarify the independent influence of different motion modes on the evolution mechanism of the leading edge vortex, researchers usually decompose the complex kinematics such as flapping motion into a combination of basic motions, including pitch, horizontal acceleration, lift, rotation, and combinations of these motions. For pitch, lift, horizontal acceleration, etc. motion, due to the quasi-two-dimensional nature of the real flow, these motions are usually simplified as the motion of a two-dimensional airfoil (a section of an airfoil) for research. Through the study of these classic forms of motion, basic knowledge about dynamic stall and the dynamics of the leading edge vortex can be established, thereby laying the foundation for the study of more complex motion. The technical solution of the present invention is also based on the flow around a two-dimensional airfoil.

[0066] General development history and stage division of dynamic stall:

[0067] The prediction of dynamic stall must be based on the understanding of the general development of dynamic stall. During the dynamic stall process, a series of critical flow events will occur, and according to these events, the dynamic stall can be divided into different stages, see Figure 3 Here, the general development of dynamic stall is introduced by taking a pitching airfoil with increasing angle of attack as an example.

[0068] First, the first stage is the attached flow stage. For a pitching airfoil starting from a small angle of attack, when its angle of attack is less than the static stall angle of attack, there is almost no obvious flow separation on the airfoil surface, or the flow separation is only limited in a small area near the trailing edge, while the flow attachment state is maintained in most of the airfoil surface, so this stage is called the attached flow stage.

[0069] The second stage is the initial instability stage. When the angle of attack of the airfoil gradually increases and exceeds the static stall angle of attack, some viscous related unsteady flow phenomena will gradually appear in the flow field, such as small scale flow separation and the first backflow zone appearing at the trailing edge. At the interface between the backflow zone and the free flow area, a shear layer is formed, which controls the subsequent development of dynamic stall. Due to the initial instability of the shear layer, some small scale vortex structures in the shear layer are initially induced, such as K-H vortex. For some working conditions in a certain Reynolds number range, in addition to the trailing edge separation, another important physical feature in the initial instability stage is the laminar separation bubble appearing near the leading edge before the leading edge vortex is generated.

[0070] The third stage is the leading edge vortex convection stage, and the main flow events of this stage include the generation, development and shedding of the leading edge vortex. The leading edge vortex is the most representative flow structure in the dynamic stall process, and with the appearance of the leading edge vortex, the wing suction surface will appear large scale flow separation, and the stall will enter the irreversible stage. The generation of the leading edge vortex is one of the most important flow events in the whole dynamic stall process, which marks the end of the initial instability stage and the beginning of the secondary instability of the leading edge shear layer. In the development process of the leading edge vortex, the separation point of the leading edge shear layer is almost fixed at the leading edge, and the vortex quantity generated by the shear layer will continuously feed the leading edge vortex to make it grow in intensity and spatial scale. With the growth of the leading edge vortex, a secondary vortex quantity is usually introduced to form an anti-rotational vortex quantity area upstream of the leading edge vortex and below the shear layer. With the continuous development of the flow, after the leading edge vortex experiences a period of growth, the leading edge shear layer will be cut off, and the leading edge vortex will start to shed and finally convect over the trailing edge into the wake.

[0071] The fourth stage is the post-stall stage. For many deep dynamic stall conditions, after the initial leading edge vortex shedding, the lift of the airfoil can quickly drop, and then a second leading edge vortex or other large-scale structures can appear, but their strength and ability to induce wall negative pressure are usually not as strong as the initial leading edge vortex. After that, the airfoil surface will be completely stalled, accompanied by spanwise instability and other phenomena, and at this time the flow exhibits three-dimensional characteristics.

[0072] Importance of dynamic stall flow state prediction technology:

[0073] Based on these characteristics of the leading edge vortex, in some flows, such as flapping flight, it is necessary to fully utilize the leading edge vortex to obtain additional lift. In other application scenarios, such as flutter prevention, it is necessary to eliminate the adverse factors brought by the leading edge vortex as much as possible. Therefore, in various flow scenarios with dynamic stall, it is very important to effectively predict the flow state of dynamic stall, especially to monitor the development state of the leading edge vortex. Flow state prediction usually uses limited sensor sampling data to monitor part of the history and current information of the flow field in real time, and then uses these information to construct some characteristic parameters or characteristic models through certain physical rules and mathematical methods, so as to judge the current flow state development degree, predict the flow events that will occur, and make inferences about some unknown spatial or temporal flow field information, see Figure 4 Flow state prediction methods can be used as feedback for closed-loop flow control to provide core information of the flow field for the control system in real time and make some forward-looking predictions, so as to realize intelligent flow control of dynamic stall. At the same time, these prediction methods can also provide a reference for constructing a general framework of dynamic stall development and evolution, and help to reveal the deep flow mechanism of leading edge vortex evolution.

[0074] Embodiment 1

[0075] As shown in Figure 1 , the embodiment of the present application provides a wing dynamic stall flow state prediction method, which comprises:

[0076] Step 100: obtaining a pressure signal of an upper surface of a wing in a dynamic stall state at a current time.

[0077] In the step 100, the step 100 specifically comprises:

[0078] Obtaining an analog voltage signal of the upper surface of the wing in the dynamic stall state at the current time.

[0079] Analog-to-digital conversion is performed on the analog voltage signal at the current time to obtain a digital voltage signal at the current time.

[0080] Filtering the digital voltage signal at the current time to obtain a filtered digital voltage signal at the current time.

[0081] According to the set conversion coefficient and the filtered digital voltage signal at the current moment, the pressure signal at the current moment is determined.

[0082] Specifically, in actual application, appropriate positions can be selected on the airfoil upper surface, and high-frequency pressure measurement sensors can be arranged in sufficient density, such as Figure 5 . Figure 5 The white circles in the figure are high-frequency pressure measurement sensors arranged on the airfoil upper surface. The high-frequency pressure measurement sensors can be arranged in a uniform distribution or in a non-uniform distribution with appropriate encryption at the leading edge, but the distance between two adjacent high-frequency pressure measurement sensors should not be greater than 10% of the chord length of the airfoil. Moderate encryption of sensor distribution is conducive to improving the prediction accuracy. For convenience of use in subsequent formulas, the total number of high-frequency pressure measurement sensors is denoted as N s , and the sensors distributed from the leading edge to the trailing edge are sequentially numbered as 1, 2,..., N s ; the chord-wise x coordinates of these high-frequency pressure measurement sensors are denoted as x1, x2,..., x Ns . Figure 5 The y in the figure is the direction perpendicular to the chord-wise x. For convenience of use, these chord-wise coordinates can be dimensionless, that is, divided by the chord length c of the airfoil, thereby obtaining the dimensionless chord-wise coordinates x1*, x2*,..., x Ns *(x n *=x n / c), wherein the dimensionless coordinate of the leading edge point is 0, and the dimensionless coordinate of the trailing edge point is 1. Wherein, n is the serial number of the sensor.

[0083] During the movement of the airfoil, the real-time measurement values of the arranged pressure measurement sensors are obtained. The signal output end of the high-frequency pressure sensor is connected to a multi-channel signal acquisition card, so as to perform real-time sampling. The sampling frequency of each channel of the acquisition card is not less than 3000 Hz. The acquisition card converts the analog signal into a digital signal, and then transmits it to the computer for processing.

[0084] After obtaining the digital voltage signal delivered by the acquisition card, the computer first needs to convert the voltage signal into a pressure signal by multiplying the voltage signal by a conversion coefficient according to the performance of the high-frequency pressure sensor. Since the original measured pressure data contains a lot of noise, it is necessary to filter and remove the noise first to obtain a noise-free pressure signal.

[0085] Step 200: calculating the dimensionless pressure coefficient at the current moment according to the pressure signal at the current moment.

[0086] Since the pressure sensor usually needs an external pressure measuring tube for pressure measurement, the diameter and length of the pressure measuring tube will have some impact on the measurement results, which need to be corrected. After correction, the dimensionless pressure coefficient C is calculated according to the value p of each point pressure p .

[0087] The calculation formula of the dimensionless pressure coefficient is:

[0088]

[0089] Wherein, p ∞ is the dynamic pressure of the free flow; U ∞ is the velocity of the free flow; p is the pressure signal; C p is the dimensionless pressure coefficient.

[0090] Step 300: calculating the flow state index at the current time according to the dimensionless pressure coefficient at the current time; the flow state index includes: pressure space distribution coefficient, pressure high-order central moment, negative pressure peak position and modulated negative pressure peak position.

[0091] Since the pressure coefficient changes with time during dynamic stall, the obtained indexes are also functions of time t. According to the following formula, it can be known that these flow state indexes can be calculated in real time during the motion of the airfoil, so the flow state can be predicted in real time.

[0092] The calculation formula of the pressure space distribution coefficient SDCP is:

[0093]

[0094] Wherein, SDCP(t) is the pressure space distribution coefficient at time t; N s is the total number of sensors; n is the sensor serial number; x n is the dimensionless chordwise coordinate of the nth sensor; C p，n+1 (t) is the dimensionless pressure coefficient of the n+1th sensor at time t; C p，n (t) is the dimensionless pressure coefficient of the nth sensor at time t.

[0095] The calculation formula of the pressure high-order central moment HCMP is:

[0096]

[0097]

[0098] Wherein, N s is the total number of sensors; n is the sensor serial number; C p，n(t) is the dimensionless pressure coefficient of the nth sensor at time t; is the mean value of the dimensionless pressure coefficient; m is the order of the central moment. The order of the central moment is usually an integer of 3 or more; the HCM P of 3 or more orders can effectively predict the leading edge vortex generation.

[0099] The calculation formula of the negative pressure peak position LPP is:

[0100]

[0101] wherein, N max10％ represents the serial number of the sensor with the strongest negative pressure in the top 10%, N max10％ = N s x 10%.

[0102] The calculation formula of the modulated negative pressure peak position MLPP is:

[0103]

[0104]

[0105] C p,RMS represents the root mean square value of the pressure coefficient at a certain point.

[0106] Since the calculation of C p,RMS at time t requires integration of data within a short period of time before t, Δt is the length of the integration time. Δt can be referred to as the time window length for calculating C p,RMS , which is generally selected based on the airfoil chord length c and the incoming flow velocity U ∞ , which can be taken as a unit of convective time length, i.e. Δt = 1*c / U ∞ .

[0107] Step 400: determining the index normalized value of the current time according to the flow state index of the current time and the peak detection function; the index normalized value includes the extreme value, the peak value and the critical value.

[0108] Step 500: determining the actual flow state of the current time according to the index normalized value of the current time and the predicted flow state of the current time; the flow state includes the laminar separation bubble generation, the leading edge vortex generation, the leading edge vortex position and the leading edge vortex shedding.

[0109] Step 600: predicting the flow state of the next time according to the actual flow state of the current time and the peak detection function, to obtain the predicted flow state of the next time.

[0110] Specifically, the extreme value, the peak value and the specific critical value of the flow state index calculated above are distinguished; thereby identifying the specific critical flow event, and further dividing the flow development stage.

[0111] In practical application, the specific value of the index at the critical moment can be used. For example, if it is known that the critical value of a certain index at the moment of leading-edge vortex generation is 10, then when it is calculated that the index is 9 or 9.5, it can be predicted that the leading-edge vortex is about to be generated. This method requires that the working condition be tested in advance to know in advance what the critical value is.

[0112] Another method, i.e. peak detection, can also be used. This method does not depend on the specific value and does not require prior testing. For example, the SDCP has a local peak at t1, so the following function can be used to detect the peak:

[0113]

[0114] SDCP(t) represents the value of the SDCP index at t, SDCP Δt,max (t) represents the maximum value of the SDCP index that has occurred in a sliding time window with a length of Δt up to t (i.e. the maximum value in the time from t-Δt to t. If SDCP(t) at t is the maximum value in this time (in the rising period), then δ(t) should be equal to 0. If SDCP(t) at t has already been in the falling period, then δ(t) will become a value less than 0. Therefore, a threshold value, for example -5%, can be set, and when δ falls to -5%, it can be considered that the peak has fallen by 5%, and thus it can be considered that the peak has passed. This method of identifying the peak may have a slight delay, and the degree of delay depends on the selection of the threshold value. According to this function, the peaks of the SDCP index at t1 and t2 can be identified in turn, so as to predict the generation of the separation bubble and the leading-edge vortex.

[0115] A dynamic stall condition is taken as an example to introduce the application of these flow state indexes. Figure 6This figure illustrates the variation of flow state indices with convection time under a typical dynamic stall condition. The three vertical lines in the figure mark the times t1, t2, and t3, corresponding to the occurrence of three critical flow events: laminar separation bubble formation, leading-edge vortex formation, and leading-edge vortex shedding. The figure shows that the pressure spatial distribution coefficient exhibits a local extremum near the formation of the laminar separation bubble. By identifying this local extremum, the formation of the leading-edge laminar separation bubble can be detected in real time. Similarly, the pressure spatial distribution coefficient has a maximum value at the formation of the leading-edge vortex, thus serving as an indicator of its formation. The higher-order central moment of pressure also has a maximum value at the formation of the leading-edge vortex, thus serving as a supplementary criterion for the pressure spatial distribution coefficient. Furthermore, the positions of the negative pressure peak and the modulated negative pressure peak both have maxima near the shedding time of the leading-edge vortex, thus indicating the shedding time. Since the static stall angle of attack is fixed for specific airfoils and Reynolds numbers, it can be obtained in advance by querying databases. Given the static stall angle of attack and with successful prediction of laminar separation bubble formation, leading-edge vortex formation, and leading-edge vortex shedding, different flow stages of dynamic stall can be successfully delineated. For the specific procedures of predicting dynamic stall flow states in practical applications, please refer to [link to relevant documentation]. Figure 8 .

[0116] In addition to predicting critical flow events, the negative pressure peak location index can also be used to predict the vortex core location of leading-edge vortices during leading-edge vortex convection. Figure 7 The figure shows the changes in the negative pressure peak position and the leading-edge vortex core position over convection time in a typical dynamic stall condition. The vertical line in the figure represents the moment the leading-edge vortex detaches. As can be seen from the figure, before the leading-edge vortex detaches, the negative pressure peak position and the vortex core position show good consistency. Therefore, the negative pressure peak position can effectively predict the spatial position change of the leading-edge vortex.

[0117] Example 2

[0118] like Figure 9 As shown in the figure, this embodiment of the invention provides a wing dynamic stall flow state prediction system, which includes: a signal acquisition module 1, a pressure coefficient calculation module 2, a flow state index calculation module 3, a determination module 4, a flow state determination module 5, and a prediction module 6.

[0119] Signal acquisition module 1 is used to acquire the pressure signal of the upper airfoil surface of the wing in a dynamic stall state at the current moment.

[0120] Specifically, signal acquisition module 1 includes:

[0121] The voltage signal acquisition submodule is used to acquire the analog voltage signal of the upper surface of the wing in a dynamic stall state at the current moment; the voltage signal acquisition submodule includes a pressure measurement sensor.

[0122] a conversion submodule configured to perform analog-digital conversion on the analog voltage signal at the current time to obtain a digital voltage signal at the current time.

[0123] a filtering processing submodule configured to perform filtering processing on the digital voltage signal at the current time to obtain a filtered digital voltage signal at the current time.

[0124] a pressure signal determination submodule configured to determine a pressure signal at the current time according to a set conversion coefficient and the filtered digital voltage signal at the current time.

[0125] a pressure coefficient calculation module 2 configured to calculate a dimensionless pressure coefficient at the current time according to the pressure signal at the current time.

[0126] wherein the calculation formula of the dimensionless pressure coefficient in the pressure coefficient calculation module 2 is:

[0127]

[0128] wherein p ∞ is the dynamic pressure of the free incoming flow; U ∞ is the velocity of the free incoming flow; p is the density of the free incoming flow; p is the pressure signal; C p is the dimensionless pressure coefficient.

[0129] a flow state index calculation module 3 configured to calculate a flow state index at the current time according to the dimensionless pressure coefficient at the current time; the flow state index includes a pressure spatial distribution coefficient, a pressure high-order central moment, a negative pressure peak value position and a modulated negative pressure peak value position.

[0130] a determination module 4 configured to determine an index normalized value at the current time according to the flow state index at the current time and a peak detection function; the index normalized value includes an extreme value, a peak value and a critical value.

[0131] a flow state determination module 5 configured to determine an actual flow state at the current time according to the index normalized value at the current time and a predicted flow state at the current time; the flow state includes a laminar separation bubble generation, a leading edge vortex generation, a leading edge vortex position and a leading edge vortex shedding.

[0132] a prediction module 6 configured to predict a flow state at a next time according to the actual flow state at the current time and the peak detection function, to obtain a predicted flow state at the next time.

[0133] Embodiment 3

[0134] The embodiment of the present application provides an electronic device, comprising a memory and a processor, the memory is used for storing a computer program, and the processor runs the computer program to enable the electronic device to execute the wing dynamic stall flow state prediction method in the embodiment 1.

[0135] In an embodiment, the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the wing dynamic stall flow state prediction method in the embodiment 1.

[0136] The present application has the advantages of:

[0137] 1. Real-time prediction of the main critical flow events in the dynamic stall development process, such as laminar separation bubble generation, leading edge vortex generation, leading edge vortex convection and leading edge vortex shedding, breaks through the limitation of the original leading edge suction parameter which can only predict the single event of leading edge vortex generation.

[0138] 2. The whole life cycle flow state of the leading edge vortex is monitored for the first time.

[0139] 3. Not limited to the wing shape, so it can be applied to flexible wings and other scenarios, and can also be applied to gust response and other scenarios, greatly expanding the application occasions of the original method.

[0140] 4. Low implementation difficulty, and the calculation amount of the related flow state indicators is very small, so the related state indicators can be used as feedback in closed-loop flow control during the flow process.

[0141] 5. Strong universality, can be applied to various dynamic stall working conditions of different airfoils, different Reynolds numbers and different motion parameters, and improves the shortcomings of narrow applicability and high parameter dependence of the original technology.

[0142] Each embodiment in the specification is described in a progressive manner, and each embodiment focuses on the difference from other embodiments, and the same or similar parts of each embodiment can be referred to each other.

[0143] The principles and implementation modes of the present application are described by applying specific examples in the present application, and the above embodiment description is only used to help understand the method and core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, the specific implementation mode and application range will be changed. In view of the above, the content of the specification should not be understood as a limitation of the present application.

Claims

1. A method of predicting a dynamic stall flow regime of an airfoil, the method comprising: The method comprises: acquiring a pressure signal of an upper surface of a wing in a dynamic stall state at a current time; calculating a dimensionless pressure coefficient at the current time according to the pressure signal at the current time; calculating a flow state index at the current time according to the dimensionless pressure coefficient at the current time; the flow state index comprises a pressure spatial distribution coefficient, a pressure high-order central moment, a negative pressure peak position and a modulated negative pressure peak position; determining an index normalized value at the current time according to the flow state index at the current time and a peak detection function; determining an actual flow state at the current time according to the index normalized value at the current time and a predicted flow state at the current time; the flow state comprises a laminar separation bubble generation, a leading edge vortex generation, a leading edge vortex position and a leading edge vortex shedding; predicting a flow state at a next time according to the actual flow state at the current time and the peak detection function, to obtain a predicted flow state at the next time; the wing dynamic stall flow state prediction method realizes monitoring of a flow state in a whole life cycle of a leading edge vortex; an expression of the peak detection function is: ； ； in, For peak detection function; for t Pressure spatial distribution coefficient at any given time; The pressure spatial distribution coefficient is at the cutoff point t At time t, a past time of length t The maximum value that has appeared within the sliding time window; N s This represents the total number of sensors; n For sensor serial number; x n For the first n Dimensionless chordal coordinates of each sensor; For the first n+ One sensor in t The dimensionless pressure coefficient at any given time; For the first n One sensor in t The dimensionless pressure coefficient at any given time.

2. The method of claim 1, wherein, acquiring a pressure signal of an upper surface of a wing in a dynamic stall state at a current time, specifically comprising: acquiring an analog voltage signal of the upper surface of the wing in the dynamic stall state at the current time; performing analog-digital conversion on the analog voltage signal at the current time, to obtain a digital voltage signal at the current time; performing filtering processing on the digital voltage signal at the current time, to obtain a filtered digital voltage signal at the current time; determining the pressure signal at the current time according to a set conversion coefficient and the filtered digital voltage signal at the current time.

3. The method of claim 1, wherein, a calculation formula of the dimensionless pressure coefficient is: wherein, is the dynamic pressure of the free stream; is the velocity of the free stream; is the density of the free stream; is the pressure signal; is the dimensionless pressure coefficient.

4. A system for predicting a dynamic stall flow condition of an airfoil, the system comprising: the system comprises: a signal acquisition module, configured to acquire a pressure signal of an upper surface of a wing in a dynamic stall state at a current time; a pressure coefficient calculation module, configured to calculate a dimensionless pressure coefficient at the current time according to the pressure signal at the current time; a flow state index calculation module, configured to calculate a flow state index at the current time according to the dimensionless pressure coefficient at the current time; the flow state index comprises a pressure spatial distribution coefficient, a pressure high-order central moment, a negative pressure peak position and a modulated negative pressure peak position; a determination module, configured to determine an index normalized value at the current time according to the flow state index at the current time and a peak detection function; a flow state determination module, configured to determine an actual flow state at the current time according to the index normalized value at the current time and a predicted flow state at the current time; the flow state comprises a laminar separation bubble generation, a leading edge vortex generation, a leading edge vortex position and a leading edge vortex shedding; a prediction module, configured to predict a flow state at a next time according to the actual flow state at the current time and the peak detection function, to obtain a predicted flow state at the next time; the wing dynamic stall flow state prediction method realizes monitoring of a flow state in a whole life cycle of a leading edge vortex; an expression of the peak detection function is: ； ； wherein, is a peak detection function; is t a pressure spatial distribution coefficient at time instant is the pressure spatial distribution coefficient at the cut-off time instant t , which is the maximum value occurred in the past sliding time window of length up to time instant N s is the total number of sensors; n is the sensor number; x n is the dimensionless chordwise coordinate of the n sensor; is the dimensionless pressure coefficient of the n+ 1stsensor at time instant t ; and is the dimensionless pressure coefficient of the n sensor at time instant t .

5. The system for predicting dynamic stall flow conditions of claim 4, wherein, the signal acquisition module specifically comprises: a voltage signal acquisition sub-module, configured to acquire an analog voltage signal of an upper surface of a wing in a dynamic stall state at a current time; The conversion submodule is configured to perform analog-digital conversion on the analog voltage signal at the current time to obtain a digital voltage signal at the current time; The filtering processing submodule is configured to perform filtering processing on the digital voltage signal at the current time to obtain a filtered digital voltage signal at the current time; The pressure signal determination submodule is configured to determine a pressure signal at the current time according to a set conversion coefficient and the filtered digital voltage signal at the current time.

6. The system for predicting dynamic stall flow conditions of claim 4, wherein, The calculation formula of the dimensionless pressure coefficient in the pressure coefficient calculation module is: wherein, is the dynamic pressure of the free stream; is the velocity of the free stream; is the density of the free stream; is the pressure signal; is the dimensionless pressure coefficient.

7. The system for predicting dynamic stall flow conditions of claim 5, wherein, The voltage signal acquisition submodule comprises a pressure measurement sensor and a data acquisition card.

Images