Unmanned aerial vehicle space-remaining time reliability evaluation method based on multi-source uncertainty analysis

Abstract

The invention provides an unmanned aerial vehicle airborne time reliability evaluation method based on multi-source uncertainty analysis, and the method comprises the steps: constructing an unmanned aerial vehicle aerodynamic model, carrying out the numerical simulation of the aerodynamic parameters of an unmanned aerial vehicle, obtaining the change conditions of the aerodynamic parameters, such as a lift coefficient and a resistance coefficient, along with the flight state, constructing an unmanned aerial vehicle airborne time model, the method comprises the following steps of: deducing a control equation of the space-remaining time, analyzing the change conditions of the space-remaining time under parameter uncertainty, environment uncertainty and flight state uncertainty, performing a space-remaining time simulation test of the unmanned aerial vehicle under multi-source uncertainty, quantifying the space-remaining time uncertainty of the unmanned aerial vehicle and calculating the distribution of the space-remaining time; and constructing an unmanned aerial vehicle airborne time reliability model, and evaluating the reliability. Based on the influence of the complex environment on the space-leaving time in the flight process of the unmanned aerial vehicle, the reliability model of the space-leaving time is constructed, the reliability of the space-leaving time of the unmanned aerial vehicle is quantitatively evaluated in actual use, and the method has practicability, high accuracy and high engineering applicability.

Description

Technical Field

[0001] The invention belongs to the field of UAV reliability, and in particular relates to a UAV air time reliability assessment method based on multi-source uncertainty analysis. Background Art

[0002] Unmanned aerial vehicle (UAV) technology has the advantages of high manufacturability, low cost and no risk of casualties during service, and has received extensive attention and research in many fields. However, the size and weight restrictions of UAVs and the lack of real-time monitoring increase the potential risks of mission execution. In order to expand the application scope of UAVs and ensure the successful completion of flight missions, the performance of UAVs is crucial, and the flight time is a key performance of UAVs. Since the flight time is affected by multiple sources of uncertainty such as geometric characteristics, aerodynamic performance and atmospheric environment, ensuring its accuracy is an important challenge. Therefore, predicting the aerodynamic performance of UAVs and analyzing the impact of the environment on UAVs are effective ways to evaluate the reliability of UAV time, which has important engineering practical value.

[0003] For UAVs, their aerodynamic performance is related to their design and can be simulated by numerical models. Due to their light weight and lower speed than general aircraft, the atmospheric environment has a serious impact on UAVs. Therefore, the impact of UAV aerodynamic parameters, flight environment, and flight state on the reliability of their air time is mainly considered. In terms of numerical modeling, based on the dynamic equation and kinematic equation, the aerodynamic model of the UAV describes the relationship between its motion parameters such as attitude, speed, and position and aerodynamic parameters. Through the numerical simulation of the aerodynamic model, the aerodynamic parameters of the UAV can be obtained, which is of great significance for the flight control and optimization of the UAV. In addition, on the other hand, there is the impact of the environment on the flight performance of the UAV. The main problem is the influence of wind dynamics on flight performance, especially for small aircraft. The wind direction and intensity will have an exponential impact on the flight time and energy consumption. In the actual flight process, environmental disturbances pose a great threat to flight stability and safety. Therefore, it is necessary to consider environmental disturbances in the simulation process.

[0004] Uncertainty is an important aspect of aircraft simulation, modeling and evaluation. For UAVs, it is important to quantify and analyze the uncertainty of numerical simulation and modeling results, and to consider the uncertain environment in non-deterministic aerodynamic simulations with known aerodynamic parameters. However, current research usually considers the uncertainty of the aircraft at the design stage, or is based only on the uncertainty of a single parameter. There is still a lack of research on the flight performance of UAVs that considers multi-source uncertainty. UAV applications are usually mission- and safety-critical, so the reliability of performing functions within the predetermined mission profile to provide the required services is crucial. On the other hand, the current research on UAV reliability focuses on components of UAV systems, UAV swarms, and UAV network communications. Few studies have studied the reliability of individual UAV flight performance, and there is currently a lack of research on the reliability of flight time.

[0005] In summary, multiple factors introduce uncertainty into the flight time of UAVs, but few studies have comprehensively considered the impact of these factors and further evaluated the reliability of flight flight time under multi-source uncertainty. Therefore, it is extremely important to combine multi-source uncertainty and reliability assessment to establish a UAV flight time reliability assessment method based on multi-source uncertainty analysis to efficiently realize the UAV flight time reliability assessment. Summary of the invention

[0006] In view of the defects in the above-mentioned prior art, the present invention proposes a method for evaluating the reliability of drone air time based on multi-source uncertainty analysis. First, an aerodynamic model of the drone is established, and numerical simulation of the aerodynamic parameters of the drone is performed to obtain the changes of aerodynamic parameters such as the lift coefficient and the drag coefficient with the flight state, and a drone air time model is constructed. The changes of the air time under parameter uncertainty, environmental uncertainty and flight state uncertainty are analyzed, the distribution of the air time is calculated, and a Monte Carlo test of the drone air time simulation under multi-source uncertainty is performed. A reliability model of the drone air time is constructed, and its reliability is evaluated. The present invention takes into account the impact of the complex environment on the drone air time during flight, constructs a reliability model of the air time, and is used to quantitatively evaluate the reliability of the drone air time in actual use. The present invention is practical and has strong engineering applicability.

[0007] Specifically, the present invention provides a method for evaluating the reliability of a UAV's airborne time based on multi-source uncertainty analysis, which comprises the following steps:

[0008] S1. Constructing a UAV aerodynamic model and numerically simulating the aerodynamic parameters of the UAV: ​​using the UAV aerodynamic model to simulate the flight of the UAV, obtain the changes of aerodynamic parameters with the flight state, and obtain the numerical simulation results of aerodynamic parameters;

[0009] S2. Establish the control equation of the UAV's airborne time. The specific method is: according to the numerical simulation results of the UAV parameters and aerodynamic parameters, obtain the propeller efficiency η, the aircraft mass W, the flight airspeed V and the aircraft lift-to-drag ratio K. The control equation of the UAV's airborne time is as follows:

[0010]

[0011] Where K is the lift-to-drag ratio of the aircraft, and the aerodynamic parameters are related to the aircraft mass W and the flight airspeed V. Multi-source uncertainty analysis is performed based on this equation; c f is the engine fuel consumption rate;

[0012] S3. Analyze the change of the time in the air under the uncertainty of parameters, environment and flight status;

[0013] S4. Conduct a Monte Carlo experiment to simulate the UAV’s airborne time under multi-source uncertainty. Calculate the airborne time T of each flight according to the airborne time model under multi-source uncertainty, and perform statistical analysis:

[0014]

[0015] S5. Construct a reliability model for the drone’s airborne time:

[0016] R=Pr(T>T c )(3)

[0017] Among them, R is the reliability model of the UAV’s airborne time, T c The air time is the time required to meet the mission flight time, and T is the air time required for each flight;

[0018] S6. Use the moment integral method to quantify the uncertainty of the drone's airborne time and obtain the distribution of the airborne time T. Evaluate the reliability of the drone's airborne time based on the reliability model of the drone's airborne time: Specifically, it includes the following sub-steps:

[0019] S61. Under the condition that the numerical values ​​of the UAV parameters are determined, the value ranges of the airspeed V and the lift-to-drag ratio K are obtained;

[0020] S62. Determine a random variable, the random variable is:

[0021] x=(V,K)=μ x +R T u; (4)

[0022] Among them, x is the sample value of the random variable, μ x is the average value of the random variable x within the range of values, R is the upper triangular matrix formed by Cholesky decomposition of the variance-covariance matrix of the random variable x, and u is the weighted eigenvalue;

[0023] This step specifically includes the following sub-steps:

[0024] S621. Design a random variable x = (V, K), where x follows a bivariate normal distribution and is converted into two independent Gaussian variables to obtain the mean value μ of the random variable. x , the variance-covariance matrix of the random variables Σ x ;

[0025] S622, for Σ x Perform Cholesky decomposition: R T R=Σ x , where the Cholesky matrix R is:

[0026]

[0027] Among them, r k+1,k+1 is the Cholesky matrix element, k is the order of the random variable Jacobian matrix;

[0028] S623. According to the Cholesky matrix R, the Jacobian matrix J is assembled as follows:

[0029]

[0030] in, is the main diagonal element of the Jacobian matrix, is the Jacobian matrix with diagonal elements on both sides of the main diagonal, r jj is the main diagonal element parameter of the Jacobian matrix, j is the subscript number of the main diagonal element of the Jacobian matrix, j∈1, 2,…, k;

[0031] S624, solve the Jacobian matrix J eigenvalue ξ=[ξ1,ξ2,…,ξ k ], the corresponding column standard orthogonal eigenvector matrix V is:

[0032]

[0033] Among them, v kk is the element of the Jacobian matrix characteristic matrix;

[0034] S625, extract the first column component of the feature vector and calculate the weight value For each pair of eigenvalues ​​and weights (ξ, w), perform the Cholesky transform to solve for the random parameter x = (V, K);

[0035] S63, assign different weights to random variables according to the eigenvalues, calculate sample values ​​x of different aerodynamic parameters, and cross-form k 2 Group is the aerodynamic parameter test group, k is the order of the random variable Jacobian matrix;

[0036] S64, k 2 The aerodynamic parameter test group is substituted into the flight time calculation model to calculate the UAV flight time and evaluate the reliability of the UAV flight time based on the UAV flight time reliability model.

[0037] Preferably, the specific process of establishing the drone aerodynamic model in step S1 is:

[0038] S11. Use Solidworks to build a clean configuration UAV geometry model;

[0039] S12. Obtain the corresponding aerodynamic parameter values, use the clean configuration UAV geometry model in the simulation software to build a numerical model of computational fluid dynamics, and perform CFD calculations in Ansys Fluent. Use the pressure-based solver to perform steady-state solutions under single precision. Select air as the fluid material, the SST k-omega model as the turbulence model, and include low Reynolds number corrections, set the angle of attack to 0°, and the incoming flow velocity to 20m / s;

[0040] S13, performing grid division;

[0041] S14. Calculate the aerodynamic behavior of the UAV at different angles of attack based on the numerical model of computational fluid dynamics, obtain the change of aerodynamic parameters with the flight state, and obtain the numerical simulation results of aerodynamic parameters.

[0042] Preferably, the specific process of establishing the control equation for the drone's airborne time in step S2 is:

[0043] S21. The flight state of the UAV is fixed straight and level flight condition, and the relationship between the available power and the power required for level flight is established as follows:

[0044]

[0045] S22. According to the fuel consumption calculation formula of a single-engine piston propeller aircraft, the hourly fuel consumption c f,t Equal to the engine fuel consumption rate c f The product of the engine power P is:

[0046] c f,t =c f P (5);

[0047] S23. Consider the amount of fuel consumed in flight per unit time, dQ f The hourly fuel consumption c f,t *dt, accordingly, the mass of the drone is reduced by dW, and we get:

[0048]

[0049] S24. Based on the above three steps, the control equation of the drone's airborne time is obtained, and η and c are selected according to experience. f is a reasonable constant value. For the lift-to-drag ratio K, it is expressed according to the flight states W and V in combination with the aircraft's own aerodynamic characteristics. The aircraft's angle of attack α is an intermediate variable. The calculation formula for the lift-to-drag ratio K is:

[0050] K=K(W,V) (7).

[0051] Preferably, step S3 specifically includes the following sub-steps:

[0052] S31, performing polynomial fitting on the aerodynamic parameters of the UAV and generating residuals in the fitting process, and generating parameter uncertainty in the calculation process;

[0053] S32, analyze the UAV flight environment, perform statistical analysis based on weather station data, and add random fluctuations to simulate environmental uncertainty;

[0054] S33. Perform numerical simulation on the UAV flight state. Based on the correlation between flight states, the uncertainty of aerodynamic parameters and flight environment will propagate during the calculation process, resulting in uncertainty in flight state.

[0055] Preferably, the uncertainty analysis of the UAV in step S3 is quantified using Gaussian distribution, wherein the variance of the Gaussian distribution is quantified by the fitting residual.

[0056] Preferably, the uncertainty analysis of the UAV aerodynamic parameters in step S31 includes angle of attack-lift coefficient polynomial fitting, angle of attack-drag coefficient polynomial fitting, angle of attack-lift-to-drag ratio polynomial fitting, and statistical calculation of the residuals of each polynomial fitting.

[0057] Preferably, the uncertainty analysis of the UAV flight environment and flight status in step S32 and step S33 is quantified by random simulation of wind speed.

[0058] Preferably, the parameters in the drone airborne time calculation model in step S4 are quantified according to different drone models, different environments and different flight missions.

[0059] Preferably, in step S5, the reliability model of the drone's airborne time is established by Gaussian distribution.

[0060] Preferably, in step S12, calculation is performed in Ansys Fluent, a pressure-based solver is selected, a steady-state solution is performed under single precision, air is selected as the fluid material, the SST k-omega model is selected as the turbulence model, and a low Reynolds number correction is included.

[0061] Compared with the prior art, the beneficial technical effects of the present invention are:

[0062] (1) The reliability assessment method of the drone's airborne time based on multi-source uncertainty analysis proposed in the present invention obtains the drone's flight state parameters as calculation variables for the airborne time according to the numerical simulation results of the drone's parameters and aerodynamic parameters, and can establish the control equation for the drone's airborne time. The parameter uncertainty is calculated based on the polynomial relationship between the aerodynamic parameters lift coefficient, drag coefficient, lift-to-drag ratio and angle of attack; statistical analysis is performed based on the meteorological station data, and random fluctuations are added on this basis to simulate the uncertainty of the drone's flight environment; the drone's flight state is numerically simulated to complete the uncertainty propagation calculation of the aerodynamic parameters and the flight environment, and generate the uncertainty of the flight state. This method analyzes and quantifies the multi-source uncertainty of the airborne time from three aspects, and constructs a drone uncertainty analysis framework, which can analyze and calculate the multi-source uncertainty problem in the flight process.

[0063] (2) The reliability assessment method of drone airborne time based on multi-source uncertainty analysis proposed in the present invention is based on the Monte Carlo test of drone airborne time simulation under multi-source uncertainty, and the airborne time is calculated according to the airborne time model under multi-source uncertainty, and statistical analysis is performed: a reliability model of drone airborne time is constructed, and its reliability is evaluated according to the distribution and probability of airborne time. This method can be used to calculate the reliability of drone airborne time under multi-source uncertainty conditions with high accuracy, providing a relatively simple solution to practical engineering problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Other features, objects and advantages of the present application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings.

[0065] Figure 1 It is a flow chart of the reliability assessment method of UAV air time based on multi-source uncertainty analysis of the present invention;

[0066] Figure 2 is a diagram of a complete model of a UAV in a clean configuration according to an embodiment of the present invention;

[0067] Figure 3 is a schematic diagram of a computational fluid dynamics model in an embodiment of the present invention;

[0068] Figure 4 is a schematic diagram of wind speed variation in a wind field containing uncertainty in an embodiment of the present invention;

[0069] Figure 5 is a corresponding relationship diagram of flight airspeed-wind speed-UAV weight in an embodiment of the present invention;

[0070] Figure 6 is a schematic diagram of probability density distribution of the drone's airborne time in an embodiment of the present invention;

[0071] Figure 7 Schematic diagram of the calculation result of the reliability of the UAV's airborne time in an embodiment of the present invention. DETAILED DESCRIPTION

[0072] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the relevant invention are shown in the accompanying drawings.

[0073] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0074] This paper proposes a reliability assessment method for UAV air time based on multi-source uncertainty analysis. Figure 1 As shown, the method comprises the following steps:

[0075] S1. Construct an aerodynamic model of the UAV and perform numerical simulation on the aerodynamic parameters of the UAV: ​​perform flight simulation of the UAV to obtain the changes of aerodynamic parameters such as lift coefficient and drag coefficient with the flight state. Step S1 specifically includes the following sub-steps:

[0076] S11. Use Solidworks to build a clean configuration UAV geometry model. Figure 2 1 is a model diagram of the UAV machine 100 in a clean configuration according to an embodiment of the present invention.

[0077] S12, obtain the corresponding aerodynamic parameter values, and use the clean configuration of the UAV geometry to build a numerical model of computational fluid dynamics in ANSYS Fluent. Specifically, in step S12, calculations are performed in Ansys Fluent, a pressure-based solver is selected, a steady-state solution is performed under single precision, air is selected as the fluid material, the SST k-omega model is selected as the turbulence model, and a low Reynolds number correction is included, and a 0° angle of attack is set, and the incoming flow velocity is 20 m / s. Figure 3 This is a computational fluid dynamics model in an embodiment of the present invention.

[0078] S13. Ensure the accuracy and efficiency of CFD analysis through proper meshing, so as to better predict the fluid flow behavior. The mesh independence of the model is studied. When the number of meshes is 6×10 6 , the best result can be obtained.

[0079] S14. Calculate the aerodynamic behavior of the UAV at different angles of attack based on the numerical model.

[0080] S2. Construct a model of the drone's air time and derive the control equation of the air time. The specific method is: according to the numerical simulation results of the drone parameters and aerodynamic parameters, obtain the engine power P, propeller efficiency η, aircraft mass W, flight airspeed V, and aircraft lift-to-drag ratio K. Since the flight state of the drone is a fixed straight and level flight condition, the control equation of the drone's air time can be established:

[0081]

[0082] In the formula, the aerodynamic parameter K is related to the UAV flight state W and V, and the flight state is affected by environmental uncertainty. Subsequent multi-source uncertainty analysis can be performed based on this equation.

[0083] S21. Since the UAV is in a straight and level flight condition, the relationship between the available power and the required power for level flight can be established as follows:

[0084]

[0085] S22. According to the fuel consumption calculation formula of a single-engine piston propeller aircraft, the hourly fuel consumption c f,t Equal to the engine fuel consumption rate c f The product of the engine power P is:

[0086] c f,t =c f P (3).

[0087] S23. Consider the amount of fuel consumed in flight per unit time, dQ f The hourly fuel consumption c f,t *dt, accordingly, the mass of the drone is reduced by dW, so we can get:

[0088]

[0089] S24. Based on the above three steps, the control equation of the drone's airborne time can be obtained. According to experience, η and c are taken as f For a reasonable constant value, the lift-to-drag ratio K can be expressed in combination with the aircraft's own aerodynamic characteristics according to the flight states W and V. The aircraft's angle of attack α is an intermediate variable in this calculation.

[0090] K=x(W,V,α) (5).

[0091] S3, analyze the change of the air time under parameter uncertainty, environmental uncertainty, and flight state uncertainty, where the UAV uncertainty analysis is quantified using Gaussian distribution, and the variance of the Gaussian distribution is quantified by the fitting residual. Specifically, it includes the following sub-steps:

[0092] S31. Uncertainty analysis of UAV aerodynamic parameters including angle of attack α-lift coefficient C L 、Angle of attack α-Drag coefficient C D , angle of attack α-lift-to-drag ratio K polynomial fitting, and statistical calculation of polynomial fitting residuals. Polynomial fitting of UAV aerodynamic parameters will produce parameter uncertainty in the calculation process due to residuals generated in the fitting process.

[0093] A cubic polynomial fitting was performed based on the aerodynamic parameters obtained from the UAV flight simulation in S1. The specific results are shown below:

[0094] C L =0.000124α 3 -0.00248α 2 +0.0870α+0.617 (6)

[0095] C D =0.0000112α 3 +0.00100α 2 +0.00310α+0.0367 (7)

[0096] K=0.00394α 3 -0.172α 2 +0.845α+14.4 (8).

[0097] Residual analysis was performed based on the fitting results, and the fitting residuals of angle of attack-lift coefficient, angle of attack-drag coefficient, and angle of attack-lift-to-drag ratio were 0.0501, 0.00554, and 2.52, respectively. The fitting residuals were used as the variance of the Gaussian distribution to quantify the uncertainties of the lift coefficient, drag coefficient, and lift-to-drag ratio.

[0098] S32. The uncertainty analysis of the UAV flight environment and flight status is quantified through random simulation of wind speed. Specifically, the UAV flight environment is analyzed, statistical analysis is performed based on the weather station data, and random fluctuations are added on this basis to simulate the uncertainty of the environment.

[0099] Based on the wind speed data obtained from the meteorological station, we can statistically analyze its wind speed characteristics and establish a wind field sequence containing turbulence on this basis. Taking into account the strong changes in the wind field during actual flight, the simulation is generated at a time interval of 1 minute in the proportion of 50% light turbulence, 30% moderate turbulence, and 20% heavy turbulence to simulate the actual wind speed fluctuations.

[0100] S33. Perform numerical simulation on the UAV flight state. Since the flight states are interrelated, the uncertainty of aerodynamic parameters and flight environment will propagate during the calculation process, resulting in uncertainty in the flight state.

[0101] The case calculation is performed based on the drone parameters obtained from the air-staying model in S2. The specific parameters are as follows: take-off weight 60kg, empty weight 26kg, wing area 3m 2 , average fuel consumption rate 0.0006kg / w*h, engine-propeller efficiency 60%. According to steps S31 and S32, the aerodynamic parameters and wind field are calculated to obtain the real-time correspondence between flight airspeed-wind speed-UAV weight, and finally quantify the uncertainty of flight state K, W, V. Among them, Figure 4 Graph showing wind speed variation in a wind field containing uncertainty in an embodiment of the present invention. Figure 5 This is a corresponding relationship diagram of flight airspeed-wind speed-UAV weight in an embodiment of the present invention.

[0102] S4. Conduct Monte Carlo experiments to simulate the drone’s flight time under multi-source uncertainty. The parameters can be quantified according to different drone models, different environments, and different flight missions. Calculate the flight time of each flight according to the flight time model under multi-source uncertainty. Figure 6 : is a schematic diagram of the probability density distribution of the drone's airborne time in an embodiment of the present invention. Then statistical analysis is performed:

[0103]

[0104] S5. Construct a reliability model for the drone’s time in the air, where the drone’s time in the air reliability model is established through normal distribution. Figure 7 Schematic diagram of the calculation result of the reliability of the drone's airborne time in an embodiment of the present invention. The reliability model of the drone's airborne time is as follows:

[0105] R=Pr(T>T c ) (10).

[0106] S6. Use the moment integral method to quantify the uncertainty of the drone's airborne time and obtain the distribution of the airborne time T. Evaluate the reliability of the drone's airborne time based on the reliability model of the drone's airborne time: Specifically, it includes the following sub-steps:

[0107] S61. Under the condition that the numerical values ​​of the UAV parameters are determined, the value ranges of the airspeed V and the lift-to-drag ratio K are obtained.

[0108] S62: Identify uncertainty parameters, the random parameter vector is:

[0109] x=(V,K)=μ x +R T u; (11)

[0110] Where x is the sample value of the uncertainty parameter; μ xis the average value of the random variable x within the value range; R is the upper triangular matrix formed by Cholesky decomposition of the variance-covariance matrix of the random variable x; u is the weighted eigenvalue, and the specific calculation process includes the following sub-steps:

[0111] S621. Design a random variable x = (V, K), x follows a bivariate normal distribution and can be converted into two independent Gaussian variables. The mean value μ of the random variable is obtained. x , the variance-covariance matrix of the random variables Σ x .

[0112] S622: For Σ x Perform Cholesky decomposition: R T R=Σ x , where the Cholesky matrix R is:

[0113]

[0114] Among them, r k+1,k+1 is the Cholesky matrix element, and k is the order of the Jacobian matrix of the random variable.

[0115] S623: Assemble the Jacobian matrix J according to the Cholesky matrix R as follows:

[0116]

[0117] in, is the main diagonal element of the Jacobian matrix, is the Jacobian matrix with diagonal elements on both sides of the main diagonal, r jj is the main diagonal element parameter of the Jacobian matrix, j is the subscript number of the main diagonal element of the Jacobian matrix, j∈1, 2,…, k.

[0118] S624: Solve the Jacobian matrix J eigenvalue ξ=[ξ1,ξ2,…,ξ k ], the corresponding column standard orthogonal eigenvector matrix V is:

[0119]

[0120] Among them, v kk is the element of the Jacobian matrix characteristic matrix.

[0121] S625: Extract the first column component of the feature vector and calculate the weight value For each pair of eigenvalues ​​and weights (ξ, w), a Cholesky transform is performed to solve for the uncertainty parameter x = (V, K).

[0122] S63: Different weights are assigned according to the eigenvalues, and different aerodynamic parameter sample values ​​x are calculated. The uncertainty distribution of V is N(0,1), and the uncertainty distribution of K is N(0,2.52). Different airspeeds V and lift-to-drag ratios K are selected to form 9 test groups. The final fitting result is that the probability distribution of the drone's airborne time is N(84.52,0.94).

[0123] S64: Substitute the aerodynamic parameter sample values ​​into the air time calculation model, calculate the UAV air time and evaluate the UAV air time reliability according to the reliability model. According to the performance parameter requirements of the case UAV, its mission flight time is 77.8h, and the calculated reliability is 0.99.

[0124] The reliability assessment method of the drone's air time based on multi-source uncertainty analysis proposed in the present invention constructs a drone aerodynamic model, performs numerical simulation on the drone's aerodynamic parameters, obtains the changes of aerodynamic parameters such as lift coefficient and drag coefficient with the flight state, constructs a drone's air time model, derives the control equation of the air time, analyzes the changes of the air time under parameter uncertainty, environmental uncertainty, and flight state uncertainty, calculates the distribution of the air time, conducts a Monte Carlo test of the drone's air time simulation under multi-source uncertainty, constructs a drone's air time reliability model, and evaluates its reliability. The present invention takes into account the impact of the complex environment on the drone's air time during flight, constructs a reliability model of the air time, and is used to quantitatively evaluate the reliability of the drone's air time in actual use. The present invention is practical, highly accurate, and has strong engineering applicability.

[0125] Finally, it should be noted that the above-described embodiments are only intended to illustrate rather than limit the technical solutions of the present invention. Although the present invention has been described in detail with reference to the above-described embodiments, those skilled in the art should understand that the present invention can still be modified or replaced by equivalents. Any modification or partial replacement that does not depart from the spirit and scope of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for evaluating the reliability of UAV's airborne time based on multi-source uncertainty analysis, characterized in that: It includes the following steps: S1. Constructing a UAV aerodynamic model and numerically simulating the aerodynamic parameters of the UAV: ​​using the UAV aerodynamic model to simulate the flight of the UAV, obtain the changes of aerodynamic parameters with the flight state, and obtain the numerical simulation results of aerodynamic parameters; S2. Establish the control equation of the UAV's airborne time. The specific method is: according to the numerical simulation results of the UAV parameters and aerodynamic parameters, obtain the propeller efficiency η, the aircraft mass W, the flight airspeed V and the aircraft lift-to-drag ratio K, and establish the control equation of the UAV's airborne time as follows: Among them, K is the lift-to-drag ratio of the aircraft, W is the aircraft mass, V is the flight speed, c f is the engine fuel consumption rate, g is the acceleration due to gravity; S3. Analyze the change of the time in the air under the uncertainty of parameters, environment and flight status; S4. Conduct a Monte Carlo experiment to simulate the UAV's airborne time under multi-source uncertainty. Calculate the airborne time T of each flight according to the airborne time model under multi-source uncertainty and perform statistical analysis. The calculation method of the airborne time of each flight is as follows: S5. Construct a reliability model for the drone’s airborne time: R=Pr(T>T c )(3) Among them, R is the reliability model of the UAV’s airborne time, T c is the air time that meets the mission flight time, T is the air time of each flight, and Pr represents the condition satisfied by the reliable model; S6. Use the moment integral method to quantify the uncertainty of the drone's airborne time and obtain the distribution of the airborne time T. Evaluate the reliability of the drone's airborne time based on the reliability model of the drone's airborne time: Specifically, it includes the following sub-steps: S61. Under the condition that the numerical values ​​of the UAV parameters are determined, the value ranges of the airspeed V and the lift-to-drag ratio K are obtained; S62. Determine the random variable as: x=(V,K)=μ x +R T u(4) Among them, x is the sample value of the random variable, μ x is the average value of the random variable x within the range of values, R is the upper triangular matrix formed by Cholesky decomposition of the variance-covariance matrix of the random variable x, and u is the weighted eigenvalue; This step specifically includes the following sub-steps: S621. Design a random variable x = (V, K), where x follows a bivariate normal distribution and is converted into two independent Gaussian variables to obtain the mean value μ of the random variable. x , the variance-covariance matrix of the random variables Σ x ; S622, for Σ x Perform Cholesky decomposition: R T R=Σ x , where the Cholesky matrix R is: Among them, r k+1,k+1 is the Cholesky matrix element, k is the order of the random variable Jacobian matrix; S623. According to the Cholesky matrix R, the Jacobian matrix J is assembled as follows: in, is the main diagonal element of the Jacobian matrix, is the Jacobian matrix with diagonal elements on both sides of the main diagonal, r jj is the main diagonal element parameter of the Jacobian matrix, j is the subscript number of the main diagonal element of the Jacobian matrix, j∈1, 2,…, k; S624, solve the Jacobian matrix J eigenvalue ξ=[ξ1,ξ2,…,ξ k ], the corresponding column standard orthogonal eigenvector matrix V is: Among them, v kk is the element of the Jacobian matrix characteristic matrix; S625, extract the first column component of the feature vector and calculate the weight value For each pair of eigenvalues ​​and weights (ξ, w), perform the Cholesky transform to solve for the random parameter x = (V, K); S63, assign different weights to random variables according to the eigenvalues, calculate sample values ​​x of different aerodynamic parameters, and cross-form k 2 Group is the aerodynamic parameter test group, k is the order of the random variable Jacobian matrix; S64, k 2 The aerodynamic parameter test group is substituted into the airborne time calculation model in step S4 to calculate the UAV airborne time and evaluate the reliability of the UAV airborne time according to the UAV airborne time reliability model in step S5.

2. The method for evaluating the reliability of UAV's airborne time based on multi-source uncertainty analysis according to claim 1 is characterized in that: The specific process of establishing the UAV aerodynamic model in step S1 is as follows: S11. Establish a clean configuration UAV geometric structure model; S12, obtaining corresponding aerodynamic parameter values, and constructing a numerical model of computational fluid dynamics using the clean configuration UAV geometric structure model; S13, performing grid division; S14. Calculate the aerodynamic behavior of the UAV at different angles of attack based on the numerical model of computational fluid dynamics, obtain the change of aerodynamic parameters with the flight state, and obtain the numerical simulation results of aerodynamic parameters.

3. The method for evaluating the reliability of UAV's airborne time based on multi-source uncertainty analysis according to claim 1 is characterized in that: The specific process of establishing the control equation of the drone's airborne time in step S2 is: S21. The flight state of the UAV is a fixed straight and level flight condition. The relationship between the available power and the power required for level flight is established as follows: S22. According to the fuel consumption calculation formula of a single-engine piston propeller aircraft, the hourly fuel consumption c f ,t is equal to the engine fuel consumption rate c f The product of the engine power P is: Cf,t=cfP(6) S23. Consider the amount of fuel consumed in flight per unit time, dQ f The hourly fuel consumption c f ,t*dt, accordingly, the mass of the drone is reduced by dW, and we get: S24. Based on the three steps S21, S22, and S23, the control equation of the drone's airborne time is obtained, and η and c are selected according to experience. f is a reasonable constant value. For the lift-to-drag ratio K, it is expressed according to the flight states W and V in combination with the aircraft's own aerodynamic characteristics. The aircraft's angle of attack α is an intermediate variable. The calculation formula for the lift-to-drag ratio K is: K=K(W,V) (8).

4. The method for evaluating the reliability of UAV's airborne time based on multi-source uncertainty analysis according to claim 1 is characterized in that: Step S3 specifically includes the following sub-steps: S31, performing polynomial fitting on the aerodynamic parameters of the UAV and generating residuals in the fitting process, and generating parameter uncertainty in the calculation process; S32, analyze the UAV flight environment, perform statistical analysis based on weather station data, and add random fluctuations to simulate environmental uncertainty; S33. Perform numerical simulation on the UAV flight state. Based on the correlation between flight states, the uncertainty of aerodynamic parameters and flight environment will propagate during the calculation process, resulting in uncertainty in flight state.

5. The method for evaluating the reliability of UAV's airborne time based on multi-source uncertainty analysis according to claim 4 is characterized in that: In step S3, the uncertainty analysis of the UAV is quantified using Gaussian distribution, where the variance of the Gaussian distribution is quantified by the fitting residual.

6. The method for evaluating the reliability of UAV's airborne time based on multi-source uncertainty analysis according to claim 1 is characterized in that: The uncertainty analysis of the UAV aerodynamic parameters in step S31 includes the angle of attack-lift coefficient polynomial fitting, the angle of attack-drag coefficient polynomial fitting, the angle of attack-lift-drag ratio polynomial fitting, and the statistical calculation of the residuals of each polynomial fitting.

7. The method for evaluating the reliability of UAV's airborne time based on multi-source uncertainty analysis according to claim 1 is characterized in that: The uncertainty analysis of the UAV flight environment and flight status in step S32 and step S33 is quantified by random simulation of wind speed.

8. The method for evaluating the reliability of UAV's airborne time based on multi-source uncertainty analysis according to claim 1 is characterized in that: The parameters in the drone airborne time calculation model in step S4 are quantified according to different drone models, different environments and different flight missions.

9. The method for evaluating the reliability of UAV's airborne time based on multi-source uncertainty analysis according to claim 1 is characterized in that: In step S5, the reliability model of the drone's airborne time is established using Gaussian distribution.

10. The method for evaluating the reliability of UAV's airborne time based on multi-source uncertainty analysis according to claim 1 is characterized in that: In step S12, calculation is performed in Ansys Fluent, a pressure-based solver is selected, a steady-state solution is performed under single precision, air is selected as the fluid material, the SST k-omega model is selected as the turbulence model, and a low Reynolds number correction is included.

Images